EDP Sciences logo
Subscriber Authentication Point
Search
Menu
Home All issues Volume 677 (September 2023) A&A, 677 (2023) L8 Full HTML
Open Access
Issue
A&A
Volume 677, September 2023
Article Number L8
Number of page(s) 8
Section Letters to the Editor
DOI https://doi.org/10.1051/0004-6361/202347293
Published online 05 September 2023

© The Authors 2023

Licence Creative CommonsOpen Access article, published by EDP Sciences, under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

This article is published in open access under the Subscribe to Open model. Subscribe to A&A to support open access publication.

1. Introduction

Stars with masses of up to 8 M⊙ shed most of their hydrogen envelope during the asymptotic giant branch (AGB) phase. As hydrogen- and helium-burning cease in the layers around the core, these stars transition to the early white dwarf sequence where they spend the remainder of their lives as white dwarfs (WDs). Theory predicts that up to one-quarter of these stars experience a ‘final helium shell flash’ during this stage, where the residual helium spontaneously ignites under degenerate conditions (Iben et al. 1983; Blöcker 2001; Herwig 2005), which triggers a so-called ‘born-again event’. During the born-again event, the remaining hydrogen envelope mixes into the stellar core and is burned, leading to a new H-poor and 13C-rich shell ejection. After this, the star undergoes a double loop in the Hertzsprung-Russell diagram (Lawlor & MacDonald 2003; Hajduk et al. 2005) before helium burning stops, and the star re-enters the WD sequence as an H-poor star. The born-again event is thought to be a likely path for the formation of hydrogen-deficient stars such as non-DA white dwarfs, PG 1159, and [WC] stars, as well as some R CrB stars (Werner & Herwig 2006). Because born-again objects evolve on very short timescales, it is extremely rare to observe the process in action. Indeed, this has been achieved for only two cases: V605 Aql, which experienced a final helium shell flash at the beginning of the 20th century, and V4334 Sgr (also known as Sakurai’s object), which was caught undergoing the born-again behaviour in the mid 1990s (Duerbeck et al. 1996). Consequently, the born-again event is still very poorly understood, despite its importance in stellar evolution and evolved stellar populations.

Sakurai’s object, whose assumed distance from the Sun is 3.5 kpc (Hinkle et al. 2020), is the only star to have been monitored with modern facilities throughout most of the born-again process. At the time of discovery, Sakurai’s object had evolved from the early WD sequence to a born-again red giant just a few years earlier. Soon after that, it became completely enshrouded by a thick layer of dust (Kimeswenger et al. 1997). Early optical observations provided the first indications that the morphology of the ejecta was bipolar with a thick obscured disc or torus (Tyne et al. 2000; Kerber et al. 2002; Evans et al. 2006; van Hoof et al. 2007). From MiD-Infrared/VLT Interferometer (MIDI/VLTI) observations it was possible to derive a model of the morphology of the ejecta and determine that it contains a dusty disc seen almost edge-on, efficiently screening the central source (Chesneau et al. 2009). In addition, Gemini Near InfraRed Imager/ALTtitude conjugate Adaptive optics for the InfraRed unit (NIRI/ALTAIR) and Near-Infrared Integral Field Spectrometer (NIFS) images of Sakurai’s object show a pair of lobes expanding with a velocity of ∼300 km s−1, with the northeast and southwest lobes moving away from and toward us, respectively (Hinkle & Joyce 2014; Hinkle et al. 2020). While all these observations point towards some form of bipolar structure, they only give a partial and limited view of the situation. As it is expected that a considerable fraction of the ejected material has cooled down and condensed into molecular gas and dust, observations at longer wavelengths are essential to getting the full picture. In this letter, we present high-angular-resolution Atacama Large Millimeter Array (ALMA) observations, which we use to uncover the morphology of the cool ejecta in the youngest born-again star.

2. Observations

We conducted observations of the continuum and molecular line emissions of Sakurai’s object using ALMA. The project IDs associated with the observations used in this work are 2017.1.00017.S, 2018.1.00088.S (PI: P. van Hoof), and 2018.1.00341.S (PI: D. Tafoya). All observations were performed using the ALMA 12 m array, which included 43-47 antennas. The relevant parameters of the observations are summarised in Table A.1. The data were calibrated and imaged using the ALMA pipeline, which is included in the Common Astronomy Software Application (CASA; McMullin et al. 2007).

For the continuum emission, we included data from all the observations. However, for the molecular line emission, we only used the data from the observations of project 2018.1.00341.S. The continuum emission was obtained separately for each spectral window using line-free channels. The high-angular-resolution continuum image shown in Fig. 1 was made from only the observations carried out on 22 June, 2019. The continuum emission from four spectral windows centred at 224.0 GHz, 226.0 GHz, 240.0 GHz, and 242.0 GHz was combined, resulting in a total bandwidth of 6.73 GHz. The multi-scale algorithm was used in the cleaning process with the CASA task ‘tclean’. The robust parameter of the Briggs weighting scheme and the pixel size of the image were set to 0.5 and 0.004 arcsecs, respectively. The resulting root mean square (rms) of the continuum image is approximately 44 μJy beam−1, with a beam size of 0.021 × 0.019 arcsec (P.A.1 −84.1°).

thumbnail Fig. 1.

ALMA continuum emission of Sakurai’s object at 233 GHz. The emitting regions consist of a bright compact central component and faint extended structures elongated in the northeast-southwest directions. A Gaussian fit to the central component is shown as a grey ellipse. The black cross indicates the continuum peak position at (J2000) RA = 17h52m32 . s $ .\!\!^{\,\rm s} $6990 ± 0 . s $ .\!\!^{\,\rm s} $0002, Dec = −17°41′7.​​″915 ± 0.​​ ″003. The rms noise level of the image is 44 μJy beam−1. The horizontal bar indicating the linear scale of the image assumes a distance of 3.5 kpc to the source. The synthesised beam of the ALMA observations is shown in the bottom-left corner and its parameters are: θbeam = 21 × 19 milliarcsec, P.A. = −84.1°.

The spectral data include emission of the H12CN(J = 4 → 3), H13CN(J  =  4 → 3), and CO(J  =  3 → 2) lines. The spectral setup of these observations consists of four spectral windows, each of 1.875 GHz in width, centred at frequencies 343.009 GHz, 344.967 GHz, 355.009 GHz, and 356.897 GHz. Each spectral window has 1917 channels with a width of 976.562 kHz. The channel maps were created manually using a robust parameter of 0.0 and a pixel size of 0.06 arcsecs. The typical rms noise in the individual 0.8 km s−1 wide channels is 3 mJy beam−1, with a beam size of approximately 0.33 × 0.30 arcsec (P.A. ∼ 73°).

3. Results and discussion

3.1. Continuum emission: The dust disc of Sakurai’s object

The circumstellar dust of Sakurai’s object has been extensively studied at infrared (IR) wavelengths (e.g., Geballe et al. 2002; Tyne et al. 2002; Käufl et al. 2003; Chesneau et al. 2009; Hinkle & Joyce 2014; Hinkle et al. 2020; Evans et al. 2020, 2022, and references therein). In contrast, observations of the dust at submillimetre (mm) or longer wavelengths remain limited, with only a few instances reported (e.g., Evans et al. 2004). From an analysis of the evolution of the circumstellar dust in Sakurai’s object over an ∼20 year period, Evans et al. (2020) find that, overall, the mid-IR dust emission can be adequately modelled as originating from a blackbody that has cooled from Tdust = 1200 K in 1998 to Tdust = 180 K in 2016. Additionally, the dust mass has increased from Mdust ∼ 10−9 M⊙ to Mdust ∼ 10−5 M⊙ in the same period. While the mid-IR emission indicates a decline in dust temperature, the increase in continuum emission in the near-IR suggests the formation of new hot dust (Hinkle et al. 2020; Evans et al. 2020, 2022). Therefore, a multi-temperature model to fit both cooling, expanding dust and a hotter near-IR excess is required, and observations at mm and submillimetre (submm) wavelengths are invaluable in order to constrain physical parameters.

Figure 1 shows an image of the 1.3 mm continuum emission of Sakurai’s object with an unprecedented angular-resolution of 20 milliarcsec (70 AU). The image reveals an emitting region consisting of a bright compact central component and faint extended structures elongated along the northeast–southwest direction. The total flux at 233 GHz is 25 mJy and the peak of the emission is located at (J2000) RA = 17h52m32 . s $ .\!\!^{\,\rm s} $6990 ± 0 . s $ .\!\!^{\,\rm s} $0002, Dec = −17°41′7.​​″915 ± 0.​​ ″003. A Gaussian fit to the bright compact central component, shown as a grey ellipse in Fig. 1, gives a deconvolved size of 44 ± 3 × 27 ± 2 milliarcsec (154 ± 10 × 95 ± 7 AU) with a P.A. of 90° ± 6°. The peak brightness temperature of this component is TB = 217 K. The faint extended emission lies along P.A. ≈ 60° with a total length of approximately 230 milliarcsec (800 AU). Both the major axis of the bright compact central component and the axis of elongation of the faint extended structures are rotated compared to the bipolar outflow seen in near-IR images, P.A. = 21° ± 5° (Hinkle et al. 2020), suggesting that the ALMA continuum is not tracing this bipolar outflow. Moreover, the bright compact central component is oriented nearly perpendicularly to the bipolar outflow. Furthermore, its brightness temperature is in excellent agreement with the temperature (T ≈ 200 K) of the cold dust disc hinted at by mid-IR observations (Chesneau et al. 2009; Hinkle et al. 2020; Evans et al. 2020). This strongly suggests that the mm continuum emission of the bright compact central component is produced by such a cold dust disc2, implying that our ALMA image is the first direct image of the disc.

Early analysis of IR observations indicated that the circumstellar dust of Sakurai’s object primarily consists of graphitic carbon (Eyres et al. 1998; Tyne et al. 2002). Subsequent studies using Spitzer Space Telescope and James Clerk Maxwell Telescope (JCMT) data suggested that the graphitic carbon fraction decreased, while the amorphous carbon fraction increased (Evans et al. 2004, 2006). Chesneau et al. (2009) noted the absence of spectral features in the MIDI IR spectrum, which these authors interpreted as indicative of the dominance of amorphous carbon grains in the dust composition. Furthermore, Evans et al. (2020) reported evidence for weak 6–7 μm absorption, which these latter authors attribute to hydrogenated amorphous carbon formed in material ejected by Sakurai’s object at early stages after the born-again event.

While all of this evidence appears to suggest that the dust is predominantly composed of amorphous carbon, there remains a need for additional constraints on the properties of the dust grains. Some of these constraints can be obtained from the analysis of data taken within the far-IR to mm spectral range. In particular, at mm and submm wavelengths, valuable insights into the characteristics of the emitting dust can be gained by assuming optically thin emission and fitting the observed flux densities to a power law of the form Sν ∝ ν2 + β, where β represents the dust opacity index. For amorphous carbon, β ≈ 0.7 − 1, whereas for graphitic carbon, β ≈ 2 (e.g., Mennella et al. 1995, 1998). From our ALMA observations, we obtained the continuum flux densities of Sakurai’s object for the frequency range 212–357 GHz. The fluxes were measured by integrating a region containing emission exceeding three times the rms noise level of the image, and they are listed in Table A.2 and shown in Fig. 2. A power-law fit to the data gives an opacity index of β = 0.53 ± 0.03. It is noteworthy that, up to this point, this value represents the most precisely determined dust opacity index for this source within the mm and submm wavelengths range. This value is close to that expected for the sample BE of amorphous carbon analogues studied by Mennella et al. (1998), albeit slightly smaller. The reason for this could be, as Evans et al. (2004) points out, that the emission is produced by large grains and/or there has been an increase in the population of hotter grains (which have flatter emissivities).

thumbnail Fig. 2.

Spectral energy distribution of the ALMA continuum emission of Sakurai’s object. The dashed line indicates a fit using a power law function, as described in the main text. The error bars indicate the nominal uncertainty in the absolute calibration of the ALMA observations of 5%.

Assuming isothermal dust with a temperature Tdust = 217 K (i.e., assuming that the peak continuum emission is optically thick) and that most of the dust emission is optically thin, the mass of the dust can be obtained using the following expression:

M dust = S ν D 2 κ ν B ν ( T dust ) , $$ \begin{aligned} M_{\rm dust}=\frac{S_{\nu }\,D^{2}}{\kappa _{\nu }\,B_{\nu }(T_{\rm dust})}, \end{aligned} $$

where Sν is the flux density of the continuum emission, D is the distance to the source, κν = κ0(ν/ν0)β is the dust absorption coefficient, and Bν(T) is the Planck function. Dust absorption coefficient of amorphous carbon in the literature range from κ300 GHz ∼ 0.9 cm2 g−1 up to κ300 GHz ∼ 90 cm2 g−1 (Draine & Lee 1984; Ossenkopf & Henning 1994; Mennella et al. 1998; Suh 2000). Using the dust opacity index obtained from our ALMA observations (β = 0.53 ± 0.03, see Fig. 2), the calculated mass of amorphous carbon dust in Sakurai’s object is therefore 6 × 10−3 M⊙ and 6 × 10−5 M⊙ for κ300 GHz ∼ 0.9 cm2 g−1 and κ300 GHz ∼ 90 cm2 g−1, respectively. Estimations of the dust mass based on mid-IR observations range from 1.8 to 6 × 10−5 M⊙ (Chesneau et al. 2009; Hinkle et al. 2020; Evans et al. 2020). Therefore, in order to obtain a dust mass consistent with these previous estimations, it is necessary to consider amorphous carbon dust with a relatively large absorption coefficient in the submm wavelength regime.

3.2. Line emission: An expanding disc and a fast bipolar outflow in Sakurai’s object

Carbon-rich molecules (CN, C2, CO) were detected in the ejecta of Sakurai’s object as early as one year after it was discovered to be undergoing the born-again event (Eyres et al. 1998). Subsequently, H-containing species (e.g., HCN and C2H2) were also found in the ejected material (Evans et al. 2006). All of these molecular species were detected against the dust continuum emission, probing primarily the molecular gas along the line-of-sight direction. More recently, Tafoya et al. (2017) detected emission of H12CN(J = 4 → 3), H13CN(J = 4 → 3), and H12CN(J = 2 → 1) lines for the first time, and van Hoof et al. (2018) presented semi-resolved ALMA images of the line of emission of CN, CO, and HC3N. While these observations suggest that the CO molecules are situated in proximity to the dusty region (Eyres et al. 2004; Worters et al. 2009; van Hoof et al. 2018), the HCN and CN are believed to reside in bipolar lobes (Evans et al. 2006; van Hoof et al. 2018). However, due to the limited angular resolution of previous observations, it has not been possible to discern the specific regions from which the molecular emission originates.

To further investigate the molecular component of the ejecta around Sakurai’s object, we conducted ALMA observations targeting the H12CN(J = 4 → 3), H13CN(J = 4 → 3), and CO(J = 3 → 2) line emissions. These observations provided an angular resolution of approximately 300 milliarcsec (∼1000 AU) and a spectral resolution of 0.8 km s−1. Figure B.1 shows the spectra obtained from the ALMA observations, alongside the spectra acquired using the Atacama Pathfinder Experiment (APEX) in 2013 and 2016 (Tafoya et al. 2017). Due to the lower S/N of the APEX spectra, accurately measuring any changes in the brightness of the H12CN(J = 4 → 3) and H13CN(J = 4 → 3) lines between the APEX and ALMA observations is challenging. However, the spectrum in the right panel of Fig. B.1 clearly shows that the CO(J = 3 → 2) line observed in the ALMA data was either not detected or was significantly weaker in the APEX observations. This could be the result of either the recent formation of significant amounts of CO molecules or changes in the physical conditions of the emitting gas. We note that although CO molecules were found in the newly formed ejecta of Sakurai’s object (Eyres et al. 1998; Pavlenko et al. 2004), they were detected through absorption features in the near-IR spectra, which is sensitive to even small amounts of CO. Consequently, the cause of the observed enhancement in the CO(J = 3 → 2) line intensity remains uncertain.

The widths of the molecular lines are notably broad. The velocity offset range covered by the H12CN(J = 4 → 3) line extends from −350 km s−1 to 300 km s−1, with the redshifted portion even exceeding the bandwidth of our spectral window. In contrast, the H13CN(J = 4 → 3) and CO(J = 3 → 2) lines are blended, preventing us from spatially separating their emissions. Furthermore, the blueshifted part of the CO(J = 3 → 2) line also extends beyond the limits of our spectral window. Despite these limitations, the significantly higher spectral resolution of the ALMA observations enables the identification of distinct velocity components for each molecule. Additionally, the achieved high S/N allows precise determination of the centroid of the molecular emission in each velocity channel, surpassing the size of the synthesised beam (Condon 1997). Specifically, the uncertainties of the centroid positions are obtained as σ centroid = ( σ θ 2 + σ BP 2 ) 1 / 2 $ \sigma_{\mathrm{centroid}}=({\sigma_{\theta}^{2}+\sigma_{\mathrm{BP}}^{2}})^{1/2} $, where σθ ≈ θbeam/(2 S/N) and σBP = θbeam(σϕ/360°). In these expressions, θbeam is the beam size and σϕ is the phase noise of the bandpass calibrator (e.g., Zhang et al. 2017). For our observations, θbeam = 300 milliarcsecs and σϕ = 0.08°.

Figure 3 presents the spatial distribution and a position–velocity (PV) diagram of the H12CN(J = 4 → 3) line emission. The emission exhibits a clear north–south velocity gradient, with the blueshifted and redshifted components positioned to the south and north of the peak of the continuum emission, respectively. The spatial distributions and PV diagrams of the H13CN(J = 4 → 3) and CO(J = 3 → 2) lines show similar characteristics to those of the H12CN(J = 4 → 3) line, but their emitting region appears more compact. This aligns perfectly with the direction and velocity gradient of the near-IR outflow (Hinkle et al. 2020), confirming that the H12CN(J = 4 → 3) line emission indeed traces the bipolar outflow of Sakurai’s object.

thumbnail Fig. 3.

Spatial-kinematical distribution of the H12CN(J = 4 → 3) line emission in Sakurai’s object. Left: spatial distribution of the centroid positions of the H12CN(J  =  4  →  3) line emission at each individual velocity channel. Right: position–velocity diagram of the H12CN(J = 4 → 3) line emission centroids along the Declination axis. The vertical and horizontal dashed lines indicate the velocity offset and Declination offset, respectively, over which an inversion of the velocity gradient is observed. The error bars correspond to the uncertainties of the centroid positions. The axis indicating the linear scale assumes a distance of 3.5 kpc to the source.

Upon closer examination of the PV diagram, it becomes evident that there is a distinctive change in the velocity gradient. In the central region, where velocity offsets range from −50 to 50 km s−1 (see right panel of Fig. 3), the slope is inverted. The horizontal dashed lines in the PV diagram indicate that this emission extends over approximately 50 milliarcsec (∼175 AU), which is comparable to the size of the dust disc observed in the ALMA continuum image. This particular pattern of an inverted velocity gradient is commonly observed in systems that exhibit both a bipolar outflow and an expanding equatorial disc (e.g., Alcolea et al. 2007). Thus, it is likely that the inverted velocity gradient is produced by the expansion of the equatorial disc seen in the mm continuum image (see Fig. 1).

For an expanding circular thin disc of radius R, whose polar axis has an inclination i with respect to the line of sight, that is i = 0° and i = 90° correspond to the disc seen face-on and edge-on, respectively, the projected minor axis in angular units, θdisc, and its line-of-sight velocity, vlos, are given by: θdisc = 2R cos i/D and vlos = vexp sin i, respectively, where D is the distance to the source and vexp is the expansion velocity of the disc. Assuming a constant expansion velocity, the radius of the disc can be calculated as R = vexp τ, where τ is the time since the ejection of the molecular gas. Therefore, the inclination of the disc can be obtained from the following expression:

tan i = 3.0 [ v los 50 km s − 1 ] [ τ 25 years ] [ θ disc 50 milliarcsec ] − 1 [ D 3.5 kpc ] − 1 . $$ \begin{aligned} \mathrm{tan}\,i= 3.0 \left[\frac{v_{\rm los}}{\mathrm{50\,km\,s^{-1}}}\right]\left[\frac{\tau }{\mathrm{25\,years}}\right]\left[\frac{\theta _{\rm disc}}{\mathrm{50\,milliarcsec}}\right]^{-1}\left[\frac{D}{\mathrm{3.5\,kpc}}\right]^{-1}. \end{aligned} $$

From our observations, we find that vlos = 50 km s−1 and θd = 50 milliarcsec, and considering that the ejection of the material occurred around 25 years ago, the resulting inclination of the disc is i = 72°. This result confirms that the equatorial disc is seen almost edge-on, completely obscuring the central star (Chesneau et al. 2009; Hinkle et al. 2020). The de-projected expansion velocity and the radius of the disc are vexp = 53 km s−1 and R = 277 AU, respectively.

We note that the inclination of the disc derived from our ALMA observations is opposite to that deduced from the VLTI observations (Chesneau et al. 2009). While the VLTI observations suggest that the northern side of the disc is tipped towards us, which would mean that the northern edge of the expanding disc is redshifted, the ALMA observations yield the opposite for the inner 50 milliarcsec: blueshifted emission towards the north and redshifted emission towards the south (cf. right panel of Fig. 3). Furthermore, the inclination obtained from the VLTI observations implies that the northern part of the outflow should be redshifted, which is opposite to the velocity gradient seen from the near-IR and ALMA observations. We therefore conclude that the southern side of the disc is the side tipped towards us.

The derived inclination provides an estimate of the de-projected expansion velocity of the bipolar outflow detected through the H12CN(J = 4 → 3) emission, which is approximately 1000 km s−1, assuming a fairly collimated structure. Such high expansion velocities for the bipolar outflow of Sakurai’s object are not unexpected. For instance, when correcting for the inclination, the expansion velocity of the He I 10830 Å emission line observed with the Gemini NIFS shows a similar value. Also, Kerber et al. (2002) concluded that the bipolar outflow traced by the [N II] λ6583 line emission exhibited an expansion velocity of about 800 km s−1. It is worth noting that such a high expansion velocity implies that the observed lobes should be more than four times larger than observed, assuming that the molecular bipolar outflow formed at the same time as the rest of the material was ejected. This could indicate that the H12CN(J = 4 → 3) emission does not fully trace the entirety of the molecular outflow but only the regions where the excitation conditions required to observe HCN molecules are optimal. On the other hand, we cannot dismiss the possibility that the outflow traced by the H12CN(J = 4 → 3) emission has a relatively wide opening angle. In such a case, the expansion velocity of the molecular material would be lower, which could explain the smaller observed size of the molecular lobes.

The left panel of Fig. 4 shows the continuum emission superimposed on the spatial distribution of the H12CN(J  =  4  →  3) line emission. Towards the north, the continuum emission shows two protrusions that appear to be tracing the base of an hourglass-shaped outflow (indicated by the dashed lines). On the southern side, the continuum emission exhibits another such protrusion. A sketch of the inferred geometry and orientation of the material around Sakurai’s object is shown in the right panel of Fig. 4. As suggested by the protrusions in the continuum emission, the bipolar outflow lobes have an opening angle of around 60° and their walls seem to be denser than the regions close to the polar axis. In addition, as mentioned above, the spatio-kinematical configuration of the molecular emission is the same as that derived from observations of the He I 10830Å and the [C I] 9850Å lines (Hinkle et al. 2020). It is therefore possible that the expanding atomic gas forms a more collimated outflow that entrains the surrounding molecular material around it, forging the molecular outflow.

thumbnail Fig. 4.

Bipolar outflow and expanding disc around Sakurai’s object. Left: H12CN(J = 4 → 3) line emission and 233 GHz continuum emission. The dashed lines delineate the hourglass morphology suggested by the continuum emission. Right: hourglass model for the spatial-kinematical distribution of the bipolar outflow in Sakurai’s object. The right and top axes indicate the linear scale assuming a distance of 3.5 kpc to the source.

4. Concluding remarks

It is remarkable that most known objects that are thought to have experienced a born-again event (A30, A78, V605 Aql, Sakurai’s object) have a strikingly similar morphology, namely an expanding equatorial disc and a bipolar outflow (Toalá et al. 2015; Hinkle et al. 2020; Tafoya et al. 2022; Rodríguez-González et al. 2022). The formation of this particular morphology has most often been attributed to a stellar/substellar binary companion. For example A30, a common-envelope phase following the born-again event has been proposed to explain the observed physical structure and abundances of the ejecta around the central star (Rodríguez-González et al. 2022). Tafoya et al. (2022) proposed a similar common-envelope scenario to account for the expanding equatorial structure and bipolar outflow seen in their ALMA observations of V605 Aql. In contrast, standard models of the born-again event are based on single-star evolution that would result in a simple, spherical morphology of the ejecta. This has so far been observed in only one case, HuBi 1 (Toalá et al. 2021; Rechy-García et al. 2020). It is difficult to explain the more complex morphology observed for the other observed objects without invoking common envelope evolution, which would necessitate a paradigm shift because close-binary interactions would certainly influence the physics of the born-again process.

Our ALMA observations reveal, in unprecedented detail, the morpho-kinematical structure of the material surrounding the youngest born-again star, allowing us to study the initial stages of the ejection process. Our results set stringent observational constraints for future models, which will need to be consistent not only with the observed chemistry, but also the physical structure and intriguing morphology observed in the different stages of the born-again process.

1

The position angle (P.A.), measured north through east, relative to the north celestial pole.

2

The structure referred to as a ‘disc’ may also exhibit a torus-like morphology, but for consistency with the nomenclature in the literature, we refer only to discs unless indicating otherwise.

Acknowledgments

This paper makes use of the following ALMA data: ADS/JAO.ALMA # 2017.1. 00017.S, 2018.1.00088.S and 2018.1.00341.S. ALMA is a partnership of ESO (representing its member states), NSF (USA) and NINS (Japan), together with NRC (Canada), MOST and ASIAA (Taiwan), and KASI (Republic of Korea), in cooperation with the Republic of Chile. The Joint ALMA Observatory is operated by ESO, AUI/NRAO and NAOJ. D.T. acknowledges support from Onsala Space Observatory for the provisioning of its facilities support. The Onsala Space Observatory national research infrastructure is funded through Swedish Research Council grant No 2017-00648. M.H. thanks the Ministry of Education and Science of the Republic of Poland for support and granting funds for the Polish contribution to the International LOFAR Telescope (arrangement no. 2021/WK/02) and for maintenance of the LOFAR PL-612 Baldy station (MSHE decision no. 28/530020/SPUB/SP/2022). The authors also thank the anonymous referee for constructive comments and suggestions that helped to improve the manuscript.

References

  1. Alcolea, J., Neri, R., & Bujarrabal, V. 2007, A&A, 468, L41 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  2. Blöcker, T. 2001, Ap&SS, 275, 1 [CrossRef] [Google Scholar]
  3. Chesneau, O., Clayton, G. C., Lykou, F., et al. 2009, A&A, 493, L17 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  4. Condon, J. J. 1997, PASP, 109, 166 [NASA ADS] [CrossRef] [Google Scholar]
  5. Draine, B. T., & Lee, H. M. 1984, ApJ, 285, 89 [NASA ADS] [CrossRef] [Google Scholar]
  6. Duerbeck, H. W., Pollacco, D., Verbunt, F., et al. 1996, IAU Circ., 6328 [Google Scholar]
  7. Evans, A., Geballe, T. R., Tyne, V. H., et al. 2004, MNRAS, 353, L41 [NASA ADS] [CrossRef] [Google Scholar]
  8. Evans, A., Tyne, V. H., van Loon, J. T., et al. 2006, MNRAS, 373, L75 [NASA ADS] [CrossRef] [Google Scholar]
  9. Evans, A., Gehrz, R. D., Woodward, C. E., et al. 2020, MNRAS, 493, 1277 [NASA ADS] [CrossRef] [Google Scholar]
  10. Evans, A., Banerjee, D. P. K., Geballe, T. R., et al. 2022, MNRAS, 511, 713 [NASA ADS] [CrossRef] [Google Scholar]
  11. Eyres, S. P. S., Evans, A., Geballe, T. R., Salama, A., & Smalley, B. 1998, MNRAS, 298, L37 [NASA ADS] [CrossRef] [Google Scholar]
  12. Eyres, S. P. S., Geballe, T. R., Tyne, V. H., et al. 2004, MNRAS, 350, L9 [NASA ADS] [CrossRef] [Google Scholar]
  13. Geballe, T. R., Evans, A., Smalley, B., Tyne, V. H., & Eyres, S. P. S. 2002, Ap&SS, 279, 39 [NASA ADS] [CrossRef] [Google Scholar]
  14. Hajduk, M., Zijlstra, A. A., Herwig, F., et al. 2005, Science, 308, 231 [NASA ADS] [CrossRef] [Google Scholar]
  15. Herwig, F. 2005, ARA&A, 43, 435 [NASA ADS] [CrossRef] [Google Scholar]
  16. Hinkle, K. H., & Joyce, R. R. 2014, ApJ, 785, 146 [NASA ADS] [CrossRef] [Google Scholar]
  17. Hinkle, K. H., Joyce, R. R., Matheson, T., Lacy, J. H., & Richter, M. J. 2020, ApJ, 904, 34 [NASA ADS] [CrossRef] [Google Scholar]
  18. Iben, I., Jr, Kaler, J. B., Truran, J. W., & Renzini, A. 1983, ApJ, 264, 605 [NASA ADS] [CrossRef] [Google Scholar]
  19. Käufl, H. U., Koller, J., & Kerber, F. 2003, A&A, 406, 981 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  20. Kerber, F., Pirzkal, N., De Marco, O., et al. 2002, ApJ, 581, L39 [NASA ADS] [CrossRef] [Google Scholar]
  21. Kimeswenger, S., Gratl, H., Kerber, F., et al. 1997, IAU Circ., 6608, 1 [NASA ADS] [Google Scholar]
  22. Lawlor, T. M., & MacDonald, J. 2003, ApJ, 583, 913 [NASA ADS] [CrossRef] [Google Scholar]
  23. McMullin, J. P., Waters, B., Schiebel, D., Young, W., & Golap, K. 2007, in Astronomical Data Analysis Software and Systems XVI, eds. R. A. Shaw, F. Hill, & D. J. Bell, ASP Conf. Ser., 376, 127 [Google Scholar]
  24. Mennella, V., Colangeli, L., & Bussoletti, E. 1995, A&A, 295, 165 [NASA ADS] [Google Scholar]
  25. Mennella, V., Brucato, J. R., Colangeli, L., et al. 1998, ApJ, 496, 1058 [NASA ADS] [CrossRef] [Google Scholar]
  26. Ossenkopf, V., & Henning, T. 1994, A&A, 291, 943 [NASA ADS] [Google Scholar]
  27. Pavlenko, Y. V., Geballe, T. R., Evans, A., et al. 2004, A&A, 417, L39 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  28. Rechy-García, J. S., Guerrero, M. A., Santamaría, E., et al. 2020, ApJ, 903, L4 [CrossRef] [Google Scholar]
  29. Rodríguez-González, J. B., Santamaría, E., Toalá, J. A., et al. 2022, MNRAS, 514, 4794 [CrossRef] [Google Scholar]
  30. Suh, K.-W. 2000, MNRAS, 315, 740 [NASA ADS] [CrossRef] [Google Scholar]
  31. Tafoya, D., Toalá, J. A., Vlemmings, W. H. T., et al. 2017, A&A, 600, A23 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  32. Tafoya, D., Toalá, J. A., Unnikrishnan, R., et al. 2022, ApJ, 925, L4 [NASA ADS] [CrossRef] [Google Scholar]
  33. Toalá, J. A., Guerrero, M. A., Todt, H., et al. 2015, ApJ, 799, 67 [Google Scholar]
  34. Toalá, J. A., Lora, V., Montoro-Molina, B., Guerrero, M. A., & Esquivel, A. 2021, MNRAS, 505, 3883 [CrossRef] [Google Scholar]
  35. Tyne, V. H., Eyres, S. P. S., Geballe, T. R., et al. 2000, MNRAS, 315, 595 [NASA ADS] [CrossRef] [Google Scholar]
  36. Tyne, V. H., Evans, A., Geballe, T. R., et al. 2002, MNRAS, 334, 875 [Google Scholar]
  37. van Hoof, P. A. M., Hajduk, M., Zijlstra, A. A., et al. 2007, A&A, 471, L9 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  38. van Hoof, P. A. M., Kimeswenger, S., Van de Steene, G., et al. 2018, Galaxies, 6, 79 [NASA ADS] [CrossRef] [Google Scholar]
  39. Werner, K., & Herwig, F. 2006, PASP, 118, 183 [Google Scholar]
  40. Worters, H. L., Rushton, M. T., Eyres, S. P. S., Geballe, T. R., & Evans, A. 2009, MNRAS, 393, 108 [NASA ADS] [CrossRef] [Google Scholar]
  41. Zhang, Q., Claus, B., Watson, L., & Moran, J. 2017, ApJ, 837, 53 [NASA ADS] [CrossRef] [Google Scholar]

Appendix A: ALMA Observations

Table A.1.

Parameters of the ALMA observations

Table A.2.

Frequency and flux density measurements

Appendix B: Additional Figure

Fig. B.1 shows Sakurai’s object H12CN(J = 4→3), H13CN(J = 4→3) and CO(J = 3→2) line emission observed by ALMA and APEX.

thumbnail Fig. B.1.

Sakurai’s Object H12CN(J = 4→3), H13CN(J = 4→3) and CO(J = 3→2) line emission. The green line is the spectrum of the emission line detected previously with APEX. The blue line is the spectrum observed with ALMA. The spectral resolutions of the APEX and ALMA observations are 20 km s−1 and 0.8 km s−1, respectively. The bottom axis indicates the local standard of rest velocity and the top axis indicates the velocity offset from the systemic velocity assumed to be 125 km s−1.

All Tables

Table A.1.

Parameters of the ALMA observations

In the text
Table A.2.

Frequency and flux density measurements

In the text

All Figures

thumbnail Fig. 1.

ALMA continuum emission of Sakurai’s object at 233 GHz. The emitting regions consist of a bright compact central component and faint extended structures elongated in the northeast-southwest directions. A Gaussian fit to the central component is shown as a grey ellipse. The black cross indicates the continuum peak position at (J2000) RA = 17h52m32 . s $ .\!\!^{\,\rm s} $6990 ± 0 . s $ .\!\!^{\,\rm s} $0002, Dec = −17°41′7.​​″915 ± 0.​​ ″003. The rms noise level of the image is 44 μJy beam−1. The horizontal bar indicating the linear scale of the image assumes a distance of 3.5 kpc to the source. The synthesised beam of the ALMA observations is shown in the bottom-left corner and its parameters are: θbeam = 21 × 19 milliarcsec, P.A. = −84.1°.
In the text
thumbnail Fig. 2.

Spectral energy distribution of the ALMA continuum emission of Sakurai’s object. The dashed line indicates a fit using a power law function, as described in the main text. The error bars indicate the nominal uncertainty in the absolute calibration of the ALMA observations of 5%.
In the text
thumbnail Fig. 3.

Spatial-kinematical distribution of the H12CN(J = 4 → 3) line emission in Sakurai’s object. Left: spatial distribution of the centroid positions of the H12CN(J  =  4  →  3) line emission at each individual velocity channel. Right: position–velocity diagram of the H12CN(J = 4 → 3) line emission centroids along the Declination axis. The vertical and horizontal dashed lines indicate the velocity offset and Declination offset, respectively, over which an inversion of the velocity gradient is observed. The error bars correspond to the uncertainties of the centroid positions. The axis indicating the linear scale assumes a distance of 3.5 kpc to the source.
In the text
thumbnail Fig. 4.

Bipolar outflow and expanding disc around Sakurai’s object. Left: H12CN(J = 4 → 3) line emission and 233 GHz continuum emission. The dashed lines delineate the hourglass morphology suggested by the continuum emission. Right: hourglass model for the spatial-kinematical distribution of the bipolar outflow in Sakurai’s object. The right and top axes indicate the linear scale assuming a distance of 3.5 kpc to the source.
In the text
thumbnail Fig. B.1.

Sakurai’s Object H12CN(J = 4→3), H13CN(J = 4→3) and CO(J = 3→2) line emission. The green line is the spectrum of the emission line detected previously with APEX. The blue line is the spectrum observed with ALMA. The spectral resolutions of the APEX and ALMA observations are 20 km s−1 and 0.8 km s−1, respectively. The bottom axis indicates the local standard of rest velocity and the top axis indicates the velocity offset from the systemic velocity assumed to be 125 km s−1.
In the text

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.