CRL 618: Carbon-Rich Pre-Planetary Nebula
- CRL 618 is a carbon-rich pre-planetary nebula characterized by a rapid transition from the asymptotic giant branch to an ionized planetary nebula.
- Observations reveal multiple collimated lobes, fast molecular outflows, and diverse kinematic components indicating episodic mass ejections.
- Detailed spectral and polarimetric studies uncover complex chemistry, evolving shock dynamics, and binary interaction influences on its shaping.
CRL 618 is a carbon-rich pre-planetary nebula, and also an object described as a post-AGB star that has started to ionize its ejecta, placing it in the brief transition between the asymptotic giant branch and the planetary-nebula phase. Its observational phenomenology combines a compact and rapidly expanding H II region, multiple highly collimated optical and molecular lobes, a dense equatorial core or torus embedded in a more extended AGB halo, rich carbon-chain and other molecular chemistry, and ordered magnetic fields traced by dust polarization (Lee et al., 2013, Tafoya et al., 2013, Balick et al., 2014).
1. Evolutionary status, distance, and characteristic timescales
CRL 618 is consistently identified as a very young object in the post-AGB to planetary-nebula transition. The ionization of the circumstellar envelope is inferred to have begun around yr, a date that has been taken as the start of the planetary-nebula phase on the basis of the secular rise of the radio continuum flux and the angular size of the compact H II region (Tafoya et al., 2013). Independent radio expansion-parallax work at 43 GHz yields an angular expansion rate of mas yr along the major axis and a geometric distance of kpc; within errors, this agrees with the value of pc derived from the expansion of the optical lobes (Cerrigone et al., 2024).
Several ages coexist because they refer to different physical components. The dense central core mapped at 350 GHz has a dynamical age of yr and is interpreted as a recent enhanced heavy mass-loss episode that ends the AGB phase (Lee et al., 2013). The optical lobes and fast molecular outflows at their tips have expansion ages of order $100$ yr, while additional compact outflows near the source have ages of yr, implying at least two recent dynamical episodes (Huang et al., 2016). High- CO spectroscopy further indicates that the fast bipolar outflow has kinematic ages of –0 yr, consistent with very recent acceleration by shocks (Soria-Ruiz et al., 2013). Mid-infrared molecular spectroscopy characterizes CRL 618 as a young 1 PPN moving rapidly from the AGB into its ionized phase (Fonfría et al., 2010).
A frequent misconception is that these age estimates are mutually inconsistent. They are instead attached to distinct layers and tracers: the dense core records the terminal AGB mass-loss history, the bipolar lobes and bullets trace post-AGB shaping events, and the ionized core traces the onset of the planetary-nebula phase (Lee et al., 2013, Tafoya et al., 2013, Huang et al., 2016).
2. Nebular morphology and kinematics
High-resolution optical, infrared, and submillimeter observations reveal a strongly structured nebula. HST imaging shows five eastern lobes, E1–E5, with projected lengths from 2 for the smaller lobes up to 3 for the two major lobes E1 and E2; proper-motion measurements give transverse speeds 4–5 km s6 (Huang et al., 2016). CO 7–2 and HCN 8–3 imaging at up to 9 resolution spatially resolves the fast molecular-outflow region into multiple outflows oriented along different optical lobes, and also detects a pair of equatorial outflows inside the dense torus (Lee et al., 2013).
Herschel/HIFI-based modeling separates the molecular envelope into four components: a diffuse spherical halo, a pair of double empty shells (“ellipses”), a dense central core, and a fast bipolar outflow (Soria-Ruiz et al., 2013). Their kinematics span a wide range. The dense central core expands slowly, with 0 km s1, giving 2 km s3 near the center and a strong velocity gradient (Soria-Ruiz et al., 2013). The fast bipolar outflow extends to 4 cm and reaches velocities from 5 up to 6 km s7, dominating the high-8 CO wings out to 9 km s0 (Bujarrabal et al., 2010, Soria-Ruiz et al., 2013).
At sub-arcsecond scales, the central dense core mapped at 350 GHz shows that most of the detected emission arises within 1 AU 2 from the central star. Its expansion accelerates roughly linearly from 3 km s4 in the innermost part to 5 km s6 at 630 AU, with a best-fit gradient 7 km s8 arcsec9 in 0 and intercept 1 km s2 (Lee et al., 2013).
Two related kinematic patterns recur in the literature. First, the position-velocity cuts of the outflows show a linear increase of velocity with distance, 3, described as a linear “Hubble-flow” relation (Huang et al., 2016). Second, the low-velocity molecular emission traces cavity walls, while the high-velocity emission concentrates near the lobe tips (Lee et al., 2013, Huang et al., 2016). These patterns are central to current shaping models.
3. Central stars, ionized core, and secular change
Optical spectroscopy identifies an unsettled central engine. The nuclear optical spectrum combines very broad C III and C IV emission, a forest of [Fe III] lines, He I 4, narrow H I lines, and a luminous near-infrared continuum coincident with a cool post-AGB star. This combination is interpreted as characteristic of an active symbiotic system, and the broad carbon lines match a WC8-type companion by the Crowther et al. criteria (Balick et al., 2014). In the lobes, Balmer, He I, and [O III] are largely dust-scattered nuclear emission, whereas low-ionization lines such as [N I], [O I], [S II], and [N II] peak in shocked knots (Balick et al., 2014).
The compact H II region has been increasing in both flux density and size over several decades. At 22.49 GHz, seven VLA epochs from 1982.48 to 2007.52 show a monotonic flux rise from 5 mJy to 6 mJy, while the major axis grows from 7 to 8 and the minor axis from 9 to 0 (Tafoya et al., 2013). The fitted electron-density law is
1
and the frequency dependence of the major axis, 2, is interpreted as the consequence of this gradient (Tafoya et al., 2013).
Millimeter radio recombination lines probe even smaller radii. In 2015, H 3, H 4, H 5, and H 6 were detected with FWHM values from 7 km s8 down to 9 km s$100$0; relative to 1987, these lines are stronger by factors of $100$1–$100$2 and broader by $100$3–$100$4, indicating that the ionized wind now has larger expansion velocities and mass-loss rate than $100$5 years earlier (Contreras et al., 2017). A non-LTE model of the ionized core adopts $100$6 K, $100$7, and an expansion law rising from $100$8 km s$100$9 near 45 au to 0 km s1 at large 2, with 3 of the mm emission coming from 4 au (Contreras et al., 2017).
The optical line ratios evolve secularly. Across the nebula, 5 and 6 have been steadily rising, 7 has been steadily decreasing, 8 remains nearly constant, and low-ionization line ratios formed in the shocked knots have been in decline in different ways at various locations (Balick et al., 2014). These trends support a picture in which the ionizing output of the central source has increased while the broad 9 component associated with the symbiotic star has faded.
4. Molecular envelope, excitation, and circumstellar chemistry
CRL 618 has a rich molecular inventory extending from centimeter to far-infrared wavelengths. Mid-infrared TEXES/IRTF spectroscopy in the ranges 0–1 and 2–3 cm4 identified more than 170 ro-vibrational lines of C5H6, HCN, C7H8, and C9H0, but no unambiguous signature of C1H2 (Fonfría et al., 2010). The derived total column densities are
3
4
with rotational temperatures ranging from 5 to 6 K and vibrational temperatures from 7 to 8 K, demonstrating that the inner envelope is strongly out of local thermodynamic equilibrium (Fonfría et al., 2010). The abundance ladder 9, together with the non-detection of C00H01, is interpreted as the outcome of a UV-driven radical-neutral network in which stellar ultraviolet photons fragment C02H03 to C04H, which then reacts with C05H06 to build longer chains until photodestruction outpaces growth (Fonfría et al., 2010).
High-07 CO data add a complementary thermal picture. Herschel/HIFI observations detect 08CO and 09CO 10–5, 11–9, and 12–15, with wide profiles and spectacular line wings (Bujarrabal et al., 2010, Soria-Ruiz et al., 2013). Reanalysis with a spatio-kinematical model indicates that the fast bipolar outflow is hotter than previously estimated, with 13 K at 14 cm and 15 K at the outer edge; the dense central core has 16 K and very low expansion velocity (Soria-Ruiz et al., 2013). Earlier analysis already concluded that gas flowing at 17 km s18 must have 19 K and that the high temperature is evidence that the fast outflow has been accelerated by a shock and has not yet cooled down (Bujarrabal et al., 2010). Within a comparative Herschel/HIFI sample, CRL 618 stands out as having a particularly warm fast wind with characteristic 20 K (Bujarrabal et al., 2011).
The dense core also shows a complex carbon- and nitrogen-bearing chemistry at 350 GHz. More than 100 isolated or nearly isolated transitions were detected, dominated by HC21N and its 22C isotopologues, HCN and its isotopologues, and CH23CHCN toward the innermost part (Lee et al., 2013). The isotopic ratios in the core are 24 and 25, both lower than solar values (Lee et al., 2013).
At 20–25 GHz, the detected molecular spectrum is restricted to NH26 inversion absorption. Two-component rotational-diagram analysis gives 27 K and 28 K for the broad components and 29 K for the narrow absorption clump; no cyanopolyyne lines were detected at the survey sensitivity (Zhang et al., 2020). This suggests strong chemical stratification and, plausibly, efficient destruction or ionization of long carbon-chain species under the harder UV field and shocks.
A rare transient phenomenon is the 4765 MHz OH maser. Arecibo detected a narrow line with FWHM 30 km s31, peak flux density 32 mJy, and 33 km s34, about 35 km s36 blueshifted from the systemic velocity; the line was absent in later epochs and is therefore interpreted as a variable maser (Strack et al., 2019). Proposed origins include shock-induced H37O dissociation after sublimation of icy objects and an oxygen-rich circumstellar region associated with a binary companion (Strack et al., 2019).
5. Magnetic fields and polarization structure
Dust polarimetry shows that CRL 618 hosts an ordered magnetic field, but the inferred geometry depends on angular scale. SMA observations at 38–39 GHz detected dust continuum polarized emission above 40 with peak polarized intensity 41 mJy beam42 43, mean polarized intensity 44 mJy beam45 46, and a mean fractional polarization 47 (Sabin et al., 2013). The field vectors, obtained by rotating the E-vectors by 48, form a coherent, slightly curved pattern opening to the east, with measured mean position angle 49 (Sabin et al., 2013). The CO 50–2 outflow axis is 51, so the field lies parallel to the outflow major axis within 52 (Sabin et al., 2013). No equatorial toroidal component is detected at that resolution; the field is described as pure poloidal or dipole-like (Sabin et al., 2013).
The polarization fraction decreases monotonically toward regions of high Stokes 53, an anti-correlation described approximately as 54 (Sabin et al., 2013). The paper discusses beam-averaging of tangled field lines, loss of grain alignment in the dense core, and varying grain properties as possible causes.
JCMT POL-2 observations at 55 probe much larger scales, with a native beam FWHM of 56, corresponding to 57 au at 58 kpc (Pattle et al., 29 Jul 2025). Over the ten central pixels, the mean magnetic-field position angle is 59 east of north; in the north/east periphery it is 60, and in the south it is 61 (Pattle et al., 29 Jul 2025). The measured polarization fraction spans 62–63, with a high-SNR mean of 64 in the central region (Pattle et al., 29 Jul 2025). The central 65 is within 66 of the mean bullet-outflow axis at 67, and within 68 of the northeast extreme position angle, suggesting association with one cavity wall (Pattle et al., 29 Jul 2025).
Taken together, the polarimetric studies do not support a single, scale-independent toroidal geometry. At SMA resolution, CRL 618 exemplifies a well-ordered poloidal field in a young source; at JCMT resolution, polarized emission may preferentially arise from material in the walls of the dust cavity opened by the bullets (Sabin et al., 2013, Pattle et al., 29 Jul 2025). A plausible implication is that the field morphology is being reorganized by outflow-envelope interaction rather than sampling a static large-scale configuration.
6. Shaping mechanisms and dynamical interpretations
Two main classes of shaping model have been advanced for CRL 618’s multipolar structure. One is the multi-directional bullet model. In this framework, five dense clumps are ejected in two epochs, with bullets 1–2 at 69 yr and bullets 3–5 at 70 yr, interacting with a toroidal dense core and a spherical AGB halo (Huang et al., 2016). Each bullet is modeled as a 71 AU 72 AU cylinder with 73 K; bullets 1–2 have 74, bullets 3–5 have 75, and the total bullet mass is 76 (Huang et al., 2016). ZEUS-3D simulations including molecular and atomic cooling and simplified H77 chemistry reproduce five collimated lobes, a U-shaped low-velocity cavity wall, high-velocity CO emission at all five lobe tips, and the observed linear 78-79 relation in PV diagrams (Huang et al., 2016).
The corresponding observational basis is strong. SMA CO 80–2 and HCN 81–3 imaging resolves six distinct fast molecular outflow components aligned with the optical lobes. Their tip radial velocities range from 82 to 83 km s84, their dynamical ages from 85 to 86 yr, and their estimated masses from 87 to 88 (Lee et al., 2013). The outflows fall into two age groups: inner bullets around 89 yr and outer bullets around 90 yr, supporting two separate, roughly simultaneous ejection events in multiple directions (Lee et al., 2013).
A second class is the asymmetric precessing jet model. Three-dimensional hydrodynamical simulations of a precessing jet launched from an orbiting companion, with alternation in which lobe is launched at each periastron passage, reproduce a four-finger, mirror-symmetric pattern comparable to HST images (Velazquez et al., 2014). In the best-fit model, the secondary is on an eccentric orbit with 91, orbital period 92 yr, the disk axis precesses with half-angle 93 and period 94 yr, and the jet speed decreases as 95 with a periodic modulation (Velazquez et al., 2014). After 96 yr, the simulated fingers have proper motions of 97–98 km s99 and kinematic ages of 00–01 yr, again matching the observationally inferred century-scale ages (Velazquez et al., 2014).
The proposed launch mechanisms reflect this modeling diversity. For bullets, the literature discusses magneto-rotational explosions near the end of the AGB and nova-like explosions around a binary companion, the latter motivated in part by the unusually low 02 and 03 isotope ratios in the core (Huang et al., 2016, Lee et al., 2013). For the precessing-jet scenario, the key ingredient is the interaction between a Roche-lobe-filling companion, an accretion disk, orbital motion, and precession (Velazquez et al., 2014).
Rather than collapsing these models into a single mechanism, the current literature treats CRL 618 as a system in which binary interaction, episodic ejection, shocks, dense equatorial structure, and magnetic fields all plausibly participate in shaping. This is consistent with the broader empirical picture of a rapidly evolving source whose morphology, chemistry, and ionization state are all changing on humanly accessible timescales (Balick et al., 2014, Zijlstra, 2015).
