A helium-burning white dwarf binary as a supersoft X-ray source

Type Ia supernovae are cosmic distance indicators1,2, and the main source of iron in the Universe3,4, but their formation paths are still debated. Several dozen supersoft X-ray sources, in which a white dwarf accretes hydrogen-rich matter from a non-degenerate donor star, have been observed5 and suggested as Type Ia supernovae progenitors6–9. However, observational evidence for hydrogen, which is expected to be stripped off the donor star during the supernova explosion10, is lacking. Helium-accreting white dwarfs, which would circumvent this problem, have been predicted for more than 30 years (refs. 7,11,12), including their appearance as supersoft X-ray sources, but have so far escaped detection. Here we report a supersoft X-ray source with an accretion disk whose optical spectrum is completely dominated by helium, suggesting that the donor star is hydrogen-free. We interpret the luminous and supersoft X-rays as resulting from helium burning near the surface of the accreting white dwarf. The properties of our system provide evidence for extended pathways towards Chandrasekhar-mass explosions based on helium accretion, in particular for stable burning in white dwarfs at lower accretion rates than expected so far. This may allow us to recover the population of the sub-energetic so-called Type Iax supernovae, up to 30% of all Type Ia supernovae13, within this scenario.

The optical spectrum is unique, in that it shows predominantly He I and He II emission lines (Fig. 2).There are no indications for Balmer lines (see inserts in Fig. 2), no absorption lines typical for a main-sequence star, and no indications either for C or O as seen in Wolf-Rayet stars.The only other emission lines we identify (ED Fig. 2) are 7 lines of NII (5001.5, 5666.6,5679.6,6482.0,6610.6 Å) and SiII (6347.1,6371.4Å).While such lines are typically seen in AM CVn stars, several facts argue against such an interpretation.We find no evidence in the extracted 2D long-slit spectrum of any extended nebulous emission.The strong continuum emission argues against an HII-like region of a (He-rich) planetary nebula.
High-resolution optical spectra taken at 3 epochs with the HRS (High Resolution Spectrograph) at SALT reveal a double-peaked profile of all lines (Fig. 3), thus demonstrating their origin in an accretion disk.With the theoretical maximum intensity of an accretion disk line profile coming from the area of ≈0.95 of its maximum Doppler velocity, and assuming Keplerian rotation, we infer a projected velocity of the outer disk of v K × sin(i) ≈ 60 km s −1 .This suggests that the disk is seen at a low inclination angle, close to face-on.Interestingly, the He II lines have a similar profile.The FWZI (full width at zero intensity) in the He I lines suggests a maximum projected velocity of 120±10 km/s, with that of the He II 4686 line clearly being different, about 200±20 km/s.
The accretion disk is not only the origin of the emission lines, but also of the UV-optical-NIR continuum emission, as indicated by its luminosity and spectral slope; the accreting white dwarf and the donor are both hidden under this disk flux.Optical photometry shows periodic variations by a factor of 1.3, with little color variation (see ED Fig. 3).A Lomb-Scargle periodogram shows the largest power at a period of 1.1635 days, and a secondary lower-power peak at 2.327 days.The folded light curve for this longer period has a lower variance and a clear odd-even asymmetry.Phase-resolved spectroscopy is certainly needed to firmly establish which one is the true orbital period.
The helium-dominated accretion disk has two consequences: First, the donor star must be in an evolutionary phase where all the hydrogen is lost.An intriguing option is a helium star donor, with the nitrogen lines providing evidence for CNO-processed matter from the donor.Secondly, we interpret the high X-ray luminosity as due to steady He burning in a shell on the white dwarf (accretor) surface.Similar to the steady H-shell burning in the canonical supersoft X-ray sources 19;20 , models of accreting white dwarfs predict a narrow range of accretion rates, with a canonical value of ∼10 −6 M /yr, at which He-shell burning is steady 11;12;7;8;21;22;23 .If the accretion rate is higher, the accreted material puffs up and forms an envelope around the WD which becomes similar to a red giant, likely leading to common envelope evolution.If the accretion rate is lower, burning in the accreted He-layer is unstable, i.e., first starting to oscillate and then leading to He shell flashes that increase the luminosity temporarily by factors of 10 or more, on timescales which depend on various parameters 24;25 .Even lower accretion rates result in explosive helium burning.
While the measured X-ray temperature is exactly in the range expected for steady He shell burning, our measured luminosity is about ten times smaller than expected for accretion at the canonical rate.At the same time, the historical X-ray light curve, from Einstein (1979) and EXOSAT limits (1984-1986) to the ROSAT detection in 1992, and the XMM-Newton and eROSITA detections since 2019, suggests that the luminosity of [HP99] 159 is stable to within a factor of 5 (relative to the XMM-Newton value) for nearly 50 years (ED Fig. 5).This indicates the possibility that helium accretion at rates well below the canonical one (i.e., ≈ 10 −6 M /yr) can still lead to stable helium burning.
Stable burning at low accretion rates has been suggested for the case that the accreting white dwarf is rapidly rotating 24;26 .In corresponding models, stable burning is found 21 down to 5×10 −7 M /yr, and even for 3×10 −7 M /yr when allowing for fluctuations of the burning rate of a factor three.In the latter situation, the X-ray luminosity at any given time may be up to a factor of three smaller, or larger, than the value deduced from a given accretion rate assuming strictly stationary burning.If [HP99] 159 were currently near a luminosity minimum, which is more likely than it being near a maximum, its helium accretion rate could indeed be as high as 3×10 −7 M /yr.While we can not exclude that the burning rate of [HP99] 159 is oscillating with a growing amplitude, leading to instability, the expected short timescale of the evolution renders this unlikely.
A lower than the canonical burning rate is consistent with our optical spectra.If the accretion rate in [HP99] 159 would be significantly higher, a wind from the white dwarf is expected 27 .This would manifest itself with emission lines, broadened by the wind velocity (of order thousands km/s).Such broadened lines are not detected.
We have the following constraints on the mass of the He star: For initial He-star masses above ≈1 M , long-term stable evolution has been found 9 .The maximum possible initial mass depends on the assumptions concerning the wind.The present mass could be smaller than that.A rough upper bound on the present mass could be derived using the constraint that its luminosity is obviously smaller than that of the accretion disk.A helium star luminosity below ≈ 1000 L implies 28 that the current mass of the helium star is smaller than ≈ 2 M .
An orbital period of (1.16 d) 2.32 d suggests that the He star fills its Roche lobe radius of ≈(3) 4 R , being about a factor of 10 larger than on the He main sequence.In this picture, as long as the mass of the He star donor is larger than that of the white dwarf accretor, mass transfer will proceed on the thermal timescale (≈10 5 ...10 6 yrs), reducing the separation of the stars.Indeed, for He stars in the 0.8-2 M range (corresponding to initial masses on the mainsequence of 4-8 M ), this thermal timescale mass transfer 29 (during their sub-giant or giant phases) is predicted to reach rates of order 10 −7 ...10 −5 M /yr, allowing for stable He burning.After mass ratio inversion, the mass transfer rate drops and the binary widens.This may lead to the weak He-shell flash regime, consistent with [HP99] 159.
Various scenarios of white dwarfs accreting matter from a helium star companion have been suggested to lead to Type Ia supernovae.At the lowest accretion rates, helium can pile up on the white dwarf and lead to a sub-Chandrasekhar mass explosion after a critical amount of mass has been accumulated.However, in [HP99] 159 the X-ray emission implies continuous burning of the accreted matter, and consequently a continuous growth of the white dwarf mass.For this case, it has been suggested that the white dwarf undergoes a Type Ia supernova explosion once the Chandrasekhar mass is reached.A standard Type Ia explosion may strip 2. ..5% of the mass of the helium star 30 , of which no signature has been observed so far.However, it has been suggested that Chandrasekhar mass WDs may undergo sub-energetic deflagrations 31 , leading to subluminous so-called Type Iax supernovae, which are expected to strip off about ten times less mass from their helium donors 32 .Weak helium lines have been observed in the spectra of two Type Iax supernovae 13 , and a helium donor star has been proposed for the Type Iax SN 2012Z based on deep pre-explosion imaging 33 .The recent detection of helium in the bright Type Ia SN 2020eyj 34 indicates that helium donors may also sometimes trigger energetic white dwarf explosions.
While we do not know whether [HP99] 159 will evolve into a Type Ia supernova, its properties provide evidence for the pathway towards Chandrasekhar mass explosions based on helium accretion being wider than thought before.Its X-ray luminosity of ∼ 1800 L corresponds to a stationary helium accretion rate of 1.5 10 −7 M /yr, for which many models currently predict unstable burning 26 .However, [HP99] 159 appears to be relatively stable within the last 50 yr.Stable burning for lower accretion rates, as perhaps enabled by rapid rotation 21 , may allow lower mass donors to push their companion WDs to the Chandrasekhar mass.This may allow us to recover the SN Iax population within this scenario, which makes up about 30% of all Type Ia supernovae 13 .Folding our constraint on the radius of the WD in [HP99] 159 with a WD massradius relation 35 , we find a current WD mass of 1.20 +0.18 −0.40 M , implying that [HP99] 159 could undergo a Type Iax supernova explosion in the future.
When we assume that ≈10% of all Type Ia supernovae in our Galaxy (≈10 −3 per year 9 ) follow the path of helium accretion leading to Type Iax explosions, and adopting a lifetime of 3×10 5 yrs (assuming 0.3 M need to be transferred at 10 −6 M /yr), we predict about 30 helium accreting supersoft X-ray sources presently in the Milky Way.Scaling with the star formation rate would yield a handful of systems in the LMC.The detection and study of more of these sources will likely allow us to tighten the constraints on the single degenerate progenitor channel for Type Ia supernovae.

Methods
Optical photometry: SkyMapper: The optical brightness, measured by SkyMapper 36 (not simultaneously) is g = 15.82±0.02mag, r = 16.04±0.02mag, i = 16.41±0.01mag, z = 16.59±0.04mag, and after correcting for the Galactic and LMC reddening of E(B-V) = 0.105 mag (see below) results in an absolute V-band magnitude of M V = −2.8mag (assuming a LMC distance 16 of 50 kpc).This is about 5 mag (or a factor 2.5 5 = 100) brighter than typical disks in high-accretion rate nova-like cataclysmic variables 37 , and still 15-40x brighter for a face-on disk.OGLE: The region of our X-ray source was monitored regularly in the V and I bands with the Optical Gravitational Lensing Experiment (OGLE) 41;42 at a cadence of 1-3 days.Photometric calibration is done via zeropoint measurements in photometric nights, and color-terms have been used for both filters when transforming to the standard V I system.The long-term lightcurve over the 2010-2020 period shows variations by a factor 1.3 and little color variation (see ED Fig. 3).A Lomb-Scargle periodogram identifies a period of P = 1.1635 days with the largest power (panels (a) and (b) of Fig. 4), in agreement with P = 1.163471 days listed in the EROS-2 catalog of LMC periodic variables (EROS-ID lm0454n2690) 45 .Two other strong peaks at longer periods are aliases (see ED Tab. 1).A much smaller peak is seen at 2.327 days (see below).MACHO: The source was also covered by the MACHO project 43 that monitored the brightnesses of 60 million stars in the Large and Small Magellanic Clouds, and the Galactic bulge between 1992-1999.A visual (4500-6300 Å) and a red filter (6300-7600 Å) were used, the magnitudes of which were transformed to the standard Kron-Cousins V and R system, respectively, using previously determined color-terms 44 .TESS: The Transiting Exoplanet Survey Satellite 46 (TESS) is an all-sky transit survey to detect Earth-sized planets orbiting nearby M dwarfs.It continuously observes a given region of the sky for at least 27 days.For sources down to white light magnitudes ≈ 16 mag, TESS achieves ≈1% photometric precision in single 10 min.exposures.However, its large plate scale (21 px −1 ) means that care must be taken wrt.to blended sources.
[HP99] 159 was observed during all of TESS Sectors 27-39 (except Sector 33), i.e. from 2020 July to 2021 June.The analysis of [HP99] 159 is complicated by a 13 mag star at 12 distance.Yet, the 1.16 d period found in OGLE data (which resolves these two stars) is clearly visible in a Lomb-Scargle periodogram of the TESS data (Fig. 4) as the strongest peak by far.There is a signal at 2.3268 d, exactly twice of the OGLE period, at a significance of 3σ.While this is marginal, the folded (and re-binned) light curve reveals a clear odd-even effect with smaller variance that leads us to believe that this is the true period, and the strong peak at 1.16 d is likely the first harmonic of this period.The small amplitude difference, at the 0.2%-level, would explain that this is only marginally seen in the TESS periodogram.This period is also seen in the OGLE periodogram, demonstrating that it is a real feature.The phenomenon of asymmetrical maxima and minima, known in some detached binaries 47 , is unique in interacting binaries, and is especially puzzling given our inferred near face-on geometry.
With the TESS light curve 48 , we also did an independent, more sensitive search at even shorter periods that are inaccessible to OGLE.The TESS light curve was pre-whitened of the 1.16 d period and 25 of its harmonics, and the Fourier transform of the "cleaned" data was calculated.There are no indications for a shorter period down to ≈3 hrs (Fig. 4).There is also no signal at 0.538 days.This would be the fundamental period if the 1.16 d period still were an alias with the 1-3 days observing cadence of OGLE.On the other hand, two additional periodicities are found, at P 1 = 2.635 h and P 2 = 1.32 h, with significances at the 4σ level 1 .Given the non-Poissonian nature of the light curve after pre-whitening, we do not consider these two periods, which are not related harmonically, to be significant enough for further investigation.Swift/UVOT: A 1061 second Swift observation was obtained on Aug. 9, 2022, starting at 23:15 UT.While not detected in X-rays (as expected, Fig. 5), we detect [HP99] 159 in all filters of the ultraviolet-optical telescope (UVOT), at AB magnitudes as follows: UVW2 = 15.29±0.04mag, UVW1 = 15.33±0.04mag, U = 15.44±0.04mag, B = 15.73±0.04mag, V = 15.93±0.05mag, where the error is the quadratic sum of statistical and systematic error.When added to the (non-simultaneous) measurements on the longer-wavelengths bands (ED Fig. 1), the spectral energy distribution is still well described by a straight powerlaw, extending from 0.2-8 µm, without any sign of the He donor.SED modelling and extinction correction: The recent reddening map 38 of the LMC returns a much smaller reddening than previous estimates.In addition, it provides a combined reddening value for the Galactic foreground and the median LMC-intrinsic value, together with a spread due to variation within the LMC.Instead of trying a somewhat arbitrary extinction correction, we instead forward-fold a powerlaw model to all the photometry from Swift/UVOT, SkyMapper, 2MASS and Spitzer.We fit for the powerlaw slope extinguished by a combination of Milky Way and LMC dust.The powerlaw model fit is very good, and does not require a more complicated spectral model (ED Fig. 1)., and E(B-V) values of 0.01±0.01for Milky Way and 0.14±0.01for LMC dust.The latter is somewhat larger than the E(B-V)=0.11mag provided by the LMC reddening map 38 (composed of E(I-V)=0.08mag to the center of the LMC and an additional E(I-V)=0.06mag towards the far end of the LMC).More importantly, the slope of the spectral energy distribution is different from that expected for a standard accretion disk F ν ∝ ν 1/3 (ED Fig. 1).This is very similar to the SEDs of other supersoft X-ray sources like CAL 83 39 .The flatter slope has been interpreted as due to reprocessing of the high-luminosity soft X-rays, making the emission ≈100-1000 times larger than the accretion luminosity 40 .

Optical spectroscopy:
Optical spectroscopy of our source was undertaken on the Southern African Large Telescope (SALT).On 14 August 2020 a 1200 s long-slit exposure was obtained using the Robert Stobie Spectrograph (RSS) 49 in the 4070-7100 Å range (Fig. 2).Three further exposures (2020 Sep.16, Oct. 06 and 07), using the High Resolution Spectrograph (HRS) 50 , covered the 3700-5500 Å and 5500-8900 Å wavelength ranges.The primary reduction, which includes overscan correction, bias subtraction and gain correction, were carried out with the SALT science pipeline 51 .

X-ray analysis:
XMM-Newton: 4XMM J052015.1-654426 was covered serendipitously in a 29 ks XMM-Newton observation (ObsId 0841320101, PI: Pierre Maggi) on 2019 September 16/17.The EPIC instruments were operating in full-frame mode, with thin and medium filters for the pn and MOS detectors, respectively.We used the XMM-Newton data analysis software SAS version 20.0.0 to process these data.Good time intervals were identified following the method described at https://www.cosmos.esa.int/web/xmm-newton/sas-thread-epic-filterbackground.A whole field-of-view lightcurve for single-pixel events with 10000 < P I < 12000 is created and visually inspected for periods of flaring.A quiescent rate of less than 0.46 cts/s is determined and a GTI file satisfying this condition is created and used to filter the observation.After this filtering and given the off-axis position (8.7 arcmin) of [HP99] 159, its resulting vignetted exposure was ≈11.5 ks.The events used for the spectral analysis were filtered with the following expression using the SAS task evselect: '(PATTERN == 0) && (PI in [150 : 15000]) && (FLAG == 0)'.The SAS task especget was used to extract (source and background) events from a circular region with radius 60 centered on the position RA (2000.0)= 5 h 20 m 15. s 4, Decl.(2000.0)= −65 • 44 32 , as well as to calculate RMF and ARF for these events.The same was done with a circular region with radius 110 centered on the position RA = 5 h 20 m 15. s 5, Decl.= −65 • 41 11 , to be used as the background only, after excising two point sources in that region.In order to estimate the spectral parameters of the source, a Bayesian approach was implemented via 3ML 52;53 .The analysis was restricted to the 0.2−2.3keV energy band.The background and source contribution to the detected photons were modelled and folded through the appropriate responses to calculate posterior distributions of the spectral parameters.The source was modelled as an absorbed blackbody, using the 3ML models TbAbs*Blackbody (no separate abundances are used for the foreground Galactic and the LMC-intrinsic absorption).The background was modelled as a combination of instrumental background (read noise and fluorescence lines) and astrophysical background (Fig. 4), as follows: (i) a Gaussian line with normalisation, line energy and width left free to account for the low-energy noise introduced by the read-out electronics, (ii) a Gaussian line with line energy and width fixed representing the Al-K fluorescent line near 1.5 keV, which is excited by particles in the camera body, (iii) an unabsorbed APEC model with temperature left free to vary around 0.11 keV accounting for the hot gas of the local bubble, (iv) an APEC model with temperature allowed to vary around 0.22 keV absorbed by the average Galactic hydrogen column in the direction of the source, describing the contribution from the Galactic halo, and (v) a powerlaw with fixed slope of −1.41, absorbed by the combined hydrogen column of the Galaxy and the LMC in the direction of the source, arising from unresolved AGN.The contribution of the particle background is negligible in our spectral range.The photons in the source extraction region were modelled by adding the source spectrum and the background spectrum, scaled by the ratio of the extraction areas.During the fit of the data, the parameters describing the background models were linked.We obtain the following best-fit values (errors at the 1σ level): kT = 45 ± 3 eV, N H = (2.7 ± 0.4) × 10 21 cm −2 , and an unabsorbed bolometric luminosity of 6.8 +7.0 −3.5 × 10 36 erg/s, see Fig. 1.This implies an emission radius of 3700 +3900 −1900 km, consistent with a white dwarf radius.Apart from the possibility of the flux-oscillations due to the accretion rate being slightly below the burning rate, two other factors may contribute to the discrepancy of the measured vs. expected X-ray luminosity.First, due to the accretion of pure helium, the burning proceeds via the triple-α process 54 , with logT (K) ≈ 8.4 and ρ ≈ 1000 g cm −3 at the burning depth, leading to elevated levels of carbon and oxygen.Convective envelope mixing and subsequent wind ejection of CO-rich matter could lead to noticeable local X-ray absorption in the emission volume.Secondly, Non-LTE model atmospheres (as frequently used for the supersoft phase in postnova) usually give a higher peak intensity 55 than blackbody models (at the same temperature).Both effects, if taken into account in future work with improved data, would likely result in a higher X-ray luminosity (and WD radius) than that estimated above.eROSITA: [HP99] 159 = eRASSU J052015.3-654429 was detected by eROSITA 56 in each of the survey scans.Until the end of 2021, eROSITA scanned the source during five epochs as summarised in Table .2. The X-ray position was determined from the combined four eRASS surveys to be RA (2000.0)= 05 h 20 m 15. s 52 and Decl.(2000.0)= −65 • 44 28. 9 with a 1σ statistical uncertainty of 0. 6.The positional error is usually dominated by systematic uncertainties 57 which presently amount to 5 in pointed and 1 in scanning observations.Due to the unprecedented energy resolution (about 56 eV at 0.28 keV), eROSITA data are particularly sensitive to temperature changes of the source.Thus, we decided to perform spectral fitting despite the low number of counts.The spectral analysis was done using the five detectors with the on-chip aluminium filter (telescope modules 1, 2, 3, 4, and 6), avoiding the light leak in the other two detectors 56 .The eSASS 57 users version 211214 was used to process the data.Only single-pixel events without any rejection or information flag set were selected, using the eSASS task evtool.With the eSASS task srctool, a circular source region with a radius 100 , centered on the coordinates RA(2000) = 5:20:16.6,Decl.(2000) =−65:44:27 was defined to select source events.A background region of the same size and shape centered on RA(2000) = 05 h 21 m 09.s 4, Decl.(2000) = −65 • 46 00 was defined, so as to lie at the same ecliptic longitude as the source region, and hence in the scanning direction of eROSITA.The corresponding ARF and RMF files were created by the same eSASS task.Spectra were constructed by combining all events within the respective regions for each of the 5 epochs of observation.An absorbed blackbody was fitted to each of the spectra using 3ML.The priors of the free parameters were chosen based on the XMM-Newton fit results.For the absorbing column a Gaussian centered at µ = 2.7×10 21 cm −2 and with a width of σ = 0.4×10 21 cm −2 was used.The prior on kT was a Gaussian with µ = 45 eV and σ = 4 eV, truncated at zero, and the prior on the normalization was a log-normal distribution with µ = log(400) and σ = 1.For the eROSITA data, the background was not modeled due to the low number of counts; rather the data was binned to have at least 1 background photon in every bin and a profile Poisson likelihood was used.For the five epochs we obtain best-fit temperatures of kT 1 = 42 +3 −2 eV, kT 2 = 44 +3 −2 eV, kT 3 = 42 +3 −2 eV, kT 4 = 42 ± 2 eV, and kT 5 = 43 ± 2 eV.The corresponding fluxes are listed in ED Tab. 2, and shown in ED Fig. 5 together with the fluxes (or limits) of the other X-ray missions.ROSAT: [HP99] 159 was originally identified 14 in a 8.3 ks ROSAT/PSPC pointed observation (ID: 500053p) of April 1992.We have re-analyzed this observation, and find the source with a vignetting-corrected count-rate of 0.005±0.001PSPC cts/s (40±8 source counts).A blackbody fit with free parameters leads to kT = 38±15 eV, N H = (0.9 +3.2 −0.3 )×10 21 cm −2 and an unabsorbed bolometric luminosity of 1.3 +41.7 −1.0 × 10 36 erg/s.A fit with a fixed, XMM-derived temperature of 45 eV is statistically indistinguishable (due to the very small number of counts and the low energy resolution), and results in an absorption-corrected bolometric luminosity of 1.7 +41 −1.0 ×10 36 erg/s, consistent within the errors of the free fit.A fit with fixed, XMM-derived temperature and N H is substantially worse.
[HP99] 159 was not detected during the ROSAT all-sky survey, with a PSPC count rate upper limit of <0.012 cts/s.Using the best-fit spectral model of the above ROSAT pointed observation leads to a luminosity limit of < 2.5 × 10 36 erg/s, while using the XMM-derived spectral parameters leads to < 3.2×10 37 erg/s.For consistency with the Einstein and EXOSAT upper limits we choose to plot the latter value in Fig. 5.

Arguments against an AM CVn interpretation:
The He-dominated accretion disk and the NII and SiII lines (ED Fig. 2) allow the possibility of an AM CVn nature of [HP99] 159.However, a number of reasons argue against this interpretation: (i) AM CVn objects have luminosities 58 in the range of 10 30 ...10 32 erg/s.For this to be applicable to [HP99] 159, it would need to be at a distance of order 100 pc.(ii) This is incompatible with the Gaia data, which suggest a minimum distance of 8-12 kpc.(iii) Similarly, all AM CVn stars have large proper motion 58 , of order 0. 5/yr, due to their vicinity.This is a factor 100 larger than that of [HP99] 159.(iv) Finally, and most convincing, the velocity shift of all the strong lines clearly indicates LMC membership.At that distance, an AM CVn system is incompatible with the parameters we observe.

Comparison to known similar systems:
To our knowledge, the only other 'known' system of this kind was the progenitor of the He nova V445 Pup 59 .A pre-outburst luminosity of log (L/L ) = 4.34±0.36would be compatible with a 1.2-1.3M star burning helium in a shell 60 .No optical spectrum exists of the progenitor; the post-outburst spectra are H-deficient, with the strongest lines being CII and FeII 61 .Based on photographic plates taken before the outburst, an optical modulation by a factor of 1.25 and a period of 0.650654 (10) days was found, and interpreted as orbital variation of a common-envelope binary 62 .There are three possibilities for the X-ray non-detection: (i) the flux oscillations during burning with phases of low luminosity 25 , or (ii) the substantial Galactic foreground absorption in the case that the X-ray spectrum was similarly soft as [HP99] 159, or (iii) an only slightly lower temperature as compared to [HP99] 159 which would shift the emission below the X-ray detection window.Thus, the progenitor of the He nova V445 Pup could have been an object similar to [HP99] 159.

Figure 1 |Figure 2 |
Figure 1 | X-ray temperature and luminosity constraints of [HP99] 159.A 3ML fit of an absorbed blackbody model to the XMM-Newton spectrum (a), with a simultaneous fit of the background with linked parameters (see Methods) provides a good fit.Purple symbols and line show the source+background data and model, red the source only, and green the total background.The individual background components are shown in Extended Data (ED) Fig. 4. Panel (b) shows the posterior distribution in the temperature vs. hydrogen column density plane of the spectral fit (a).Each dot represents one model realization.The colour coding represents the unabsorbed bolometric luminosity assuming a distance of 50 kpc.1σ, 2σ and 3σ confidence contours are overplotted in green.Vertical dashed lines mark the Galactic foreground absorption and the sum of Galactic and total LMC absorption.

Figure 3 |
Figure 3 | Double-peak shape of optical emission lines.Flux-normalized optical spectra of different He lines (as labelled on the top right of each panel) taken with the SALT/HRS spectrograph at three different epochs: 15 Sep 2020 (dashed), 5 Oct 2020 (dotted) and 6 Oct 2020 (solid).The peak separation of all major lines is similar, incl.that of He II.The vertical dotted lines indicate a peak separation of ±60 km s −1 .The relative variation of blue/red peaks is usually explained as the orbiting hot spot created by the impact of the accretion stream on the outer edge of the accretion disk.

Figure 4 |
Figure 4 | Optical light variation.Lomb-Scargle periodograms derived from the OGLE I-band data (a) and the TESS data (b).For the interpretation of the OGLE peaks see ED Tab. 1.The TESS data are folded with a period of P = 1.1635 days (c) and P = 2.327 days (d), corresponding to ephemeres of 2459036.2885297858(14)+ 1.1635*N and 2459036.288557(8)+ 2.327*N (Barycentric Julian Date), respectively.The error bars for a given point represent the rms error of the individual 10 min data points that went into that one bin.While the Lomb-Scargle periodograms show the highest power at P = 1.1635 days, the folded light curve for the longer period (d) has a lower variance and a clear odd-even asymmetry.Lacking phase-resolved spectroscopy for a definite proof, we tentatively identify P = 2.327 days as the orbital period of [HP99] 159.
(a) Count rate is in the 0.2-2 keV band, and luminosity is foreground-absorption corrected.(b) eROSITA test scans from 2019-12-08 to 2019-12-11 are designated as eRASS0.The source position was covered by these test scans and in the early phase of eRASS1.The ≈2 week visibility of [HP99] 159 for eROSITA starts at the end of each formal eRASS survey and extends into the start of the following eRASS: we group this into 'epochs' of continuous coverage.(c) Net exposure per telescope after correcting for vignetting, averaged over the 5 telescope modules used.(d) Net source count rate after correcting for vignetting, summed over the 5 telescope modules used.

Table 1 |
Peaks in the Lomb-Scargle periodogram of the OGLE data.Notes: (i) 1/(f II -f III ) = 360.8daysindicates that these two frequencies are one year aliases of each other.(ii) 1 -f III = 0.859494 ≈ f I , suggesting that f I and f III are one day aliases.(iii)Possible shorter periods are: 1/(1+f III ) = 0.87680 days, or 1/(1+f I ) = 0.53778 days.Extended Data Table 2 | X-ray Observations of [HP99] 159.