aa r X i v : . [ a s t r o - ph . S R ] S e p The Symbiotic Stars U LISSE M UNARI
When Merrill and Humason (1932) discovered CI Cyg and AX Per, the first known sym-biotic stars (hereafter SySts), they were puzzled (in line with the wisdom of the time, noteasily contemplating stellar binarity) by the co-existence in the ’same’ star of features be-longing to distant corners of the HR diagram: the TiO bands typical of the coolest M giants,the HeII 4686 ˚A seen only in the hottest O-type stars, and an emission line spectrum match-ing that of planetary nebulae (hereafter PN). All these features stands out prominently inthe spectrum of CI Cyg shown in Figure 6.1 together with its light-curve displaying a largeassortment of different types of variability, with the spectral appearance changing in pace(a brighter state usually comes with bluer colors and a lower ionization).A great incentive to the study of SySts was provided in the 1980ies by the first confer-ence (Friedjung and Viotti, 1982) and monograph (Kenyon, 1986) devoted entirely to them,the first catalog and spectral atlas of known SySts by Allen (1984), and the first simple ge-ometrical modeling of their ionization front (Seaquist et al., 1984). Allen offered a cleanclassification criterium for SySts: a binary star, combining a red giant (RG) and a compan-ion hot enough to sustain HeII (or higher ionization) emission lines.
The spectral atlas byMunari and Zwitter (2002), shows how the majority of SySts meeting this criterium displayin their spectra emission lines of at least the NeV, OVI or FeVII ionization stages, requiringa minimum photo-ionization temperature of 130,000 K (Murset and Nussbaumer, 1994).For sometime a main sequence star accreting at a furious rate (10 − − − M ⊙ yr − ) wasadvocated as the hot component of several symbiotic stars (Kenyon and Webbink, 1984;Mikołajewska and Kenyon, 1992), but this scenario was later abandoned in favor of a WD.Accretion onto a main-sequence star must apply instead to pre -SySts systems like 17 Lepwhich are in the first phase of mass transfer (Blind et al., 2011), when the AGB progenitorof the future WD transfers mass to a main sequence companion. As indicated by satel-lite observations (eg. Munari and Buson, 1994), accretion alone cannot sustain the extremeluminosities (10 –10 L ⊙ ) encountered in most SySts (Murset et al., 1991), and - givenits high efficiency - nuclear burning at the WD surface must be invoked: burning 1 gr ofhydrogen provides infact 6 × erg, while accreting the same 1 gr on a 1.3 M ⊙ WD lib-
Reprinted from
The Impact of Binaries on Stellar Evolution , Beccari G. & Boffin H.M.J. (Eds.), c (cid:13) Ulisse Munari
Figure 6.1 Photometric and spectroscopic behavior of the prototype symbiotic star CI Cyg.
Upperpanels: the arrows mark the time for the two spectra shown, and the thick bars the passage at superiorconjunction of the WD in the 855 days orbit around the M giant companion. The letter ’ a ’ indicatesan eclipse of the outbursting WD, while ’ b ’ and ’ c ’ minima in the irradiation-modulated light-curve. α , β and γ mark three separate outbursts of different amplitude, shape and duration. The fact thatthe B − I color evolution follows the B light-curve indicates how the M giant keeps stable while theactivity resides entirely with the WD and the circumstellar gas ionized by it. Bottom panel: spectrumat outburst peak (2008 Sep 8) and at irradiation minimum (2015 Aug 13). Note the huge diversityin ionization degree (HeII, [ NeV ] , [ FeVII ] ), integrated flux of emission lines, and emission in theBalmer continuum (for this and all the following figures spectroscopy is provided by Asiago andVarese telescopes, and photometry by ANS Collaboration). erates 6 × erg, 6 × erg on a 0.5 M ⊙ WD, and only 2 × erg on a main-sequencestar like the Sun.Homogeneity did not last for long however, and by the time of the next catalog of SySts(Belczy´nski et al., 2000), systems with an excitation lower than HeII, and evidence fromX-rays for a neutron star (NS) rather than a WD, begun percolating through the broaden-ing classification criteria. Currently (Mukai et al., 2016), any binary where a WD or NS he Symbiotic Stars accretes enough material from a RG companion such that this interaction can be detectedat some wavelength is called a SySt. The number of known SySts is rapidly expanding( ∼ The amount of material burnt at the surface of the WD as been traditionally consideredequal to that continously accreted from the RG (eg. Kenyon, 1986), so that neither thenuclear burning switches-off (accretion too low), nor the envelope expands to red-giantdimension (accretion too high). The WD in burning SySts is however radiating close to itsEddington limit and accretion could fall quite shorter than required to replenish what theWD burnt. Either nuclear burning conditions are therefore met only temporarely (followedby an accreting-only interval to refuel the envelope of the WD) or the Eddington limit iscircumvent by discrete accretion episodes like that of a massive disk dumping mass ontothe WD during low-amplitude outbursts. The burning WD emits profusely super-soft X-rays, and indeed several SySts have been detected as SSS-sources, including C-1 in theDraco dwarf galaxy (Luna et al., 2013). Most however are not detected as SSS-sources,the super-soft X-rays being absorbed locally by the abundant circumstellar gas (a situationsimilar to the early evolution of novae, before the ejecta dilute and turn optically thin tosuper-soft X-rays from the burning WD at their center). The low impact made by accretion Ulisse Munari in burning SySts is confirmed by the widespread absence of both flickering and signatureof magnetic-driven accretion (Sokoloski, 2003; Zamanov et al., 2017).The SySts (mostly those of the burning type) are the only known class of astronomicalobjects known to show OVI 1032, 1036 ˚A Raman-scattered by neutral hydrogen into a pairof broad emission features at 6825, 7088 ˚A (Schmid, 1989). A so far unique exceptionseems to be the Sanduleak’s Star in LMC, which partnership with SySts has been ques-tioned (Angeloni et al., 2014). Other Raman-scattered lines from HeII 940, 972 and 1025˚A and NeVII 973 ˚A have been proposed for SySts (Lee et al., 2014). The simultaneouscoexistence of OVI and neutral hydrogen can be offered only by the very extended wind(10 au) of a RG orbited within a few photospheric radii by a burning WD.To accomodate a RG, the orbital separation in a SySt is measured in astronomical unitsrather then solar radii of cataclysmic variables. The orbital periods range mostly from 1 to 4yrs, with a maximum at 2/3 yrs and an M5III spectral type for the RG. Most burning SyStsin our Galaxy seems to belong to the metal-rich Bulge population (Whitelock and Munari,1992, Gaia will soon tell) and are O-rich (M spectral types) as opposed to C-rich (Car-bon spectral types) in the Magellanic clouds, a fact related to the lower metallicity ofSMC/LMC and its impact on the amount of Carbon brought up by the third dredge-upon the AGB. Chemical abundances has been recently derived for several of the brighestSySts from near-IR spectra. They should be treated with caution given the adopted over-simplifications (thin, plane-parallel, static, LTE atmospheres) contrasting with the hugecomplication of real RGB/AGB atmospheres (non-LTE, shocked, wind-supported, macro-turbolent, and hugely 3D extended). In about 15% of known SySts, the RG is a Mira, of anaverage M7III spectral type, and with a pulsation period usually quite longer than for fieldMiras (Whitelock, 2003). To accomodate the Mira well within its Roche lobe so to allow aregular pulsation, the orbital period must be P ≥
20 yrs. At such a wide orbital separation,no appreciable interaction would have occurred prior to the Mira evolutionary stage, withthe binary system classified as an isolated, field RG. The Miras in SySts frequently comewith warm dust (D-type SySts, as opposed to S-type with no detectable dust, and D’-typewith cooler dust), which is believed to be preferentially located in the shadow cone cre-ated by the Mira itself, and possibly causing periodic obscurations along the orbital motion(Whitelock 2003), even if an alternative location in the collision zone between the windsfrom the Mira and the WD has been proposed (Hinkle et al., 2013), at least for the Mirasin symbiotic novae.The mass transfer from the RG to the WD of SySts can occur via either capture fromwind or Roche-lobe overflow. SySts with Roche-lobe filling RG present ellipsoidal dis-torted light-curves at I -band or longer wavelengths, as illustrated in Figure 6.2, with anamplitude depending on orbital inclination and spectral type. For SySts with accreting-only WD, the ellipsoidal modulation dominates all the way down to bluest wavelengths(T CrB in Figure 6.2), while for burning SySts the irradiation by the WD of the facingside of the RG is responsible for the sinusoidal modulation dominating the bluest pho-tometric bands. The amplitude of this modulation is proportional to the temperature andluminosity of the burning WD (cf. IV Vir and higher ionization LT Del in Figure 6.2). he Symbiotic Stars Bottom panel: low resolution optical spectra of SU Lyn (red) reveals nothing differentfrom those of field normal giant of the same spectral type (black). The presence of an accretingWD is betrayed by the extra flux at far-blue ( λ ≤ Upper panel:
High resolution observations show feeble, structured, and highly variableemission in H α (and other lines as well), again disclosing the accreting nature of SU Lyn. The strong orbital dependence of the emission in the Balmer continuum (primarily respon-sible for the huge amplitude seen in the U band) suggests that an important fraction ofthe irradiation effect resides in the ionized gas between the WD and the RG (Proga et al.,1998), but at least in some cases a direct increase in the surface temperature of the irra- Ulisse Munari diated side of the RG has been documented (Chakrabarty and Roche, 1997; Munari et al.,2016). There are SySts showing no ellipsoidal distortion of the I or JHK light-curveswhile presenting deep eclipses of the WD during outbursts (eg. FG Ser). Their RG mustresides well within the Roche lobe, and the WDs have therefore to accrete from the wind,as it is the case for SySts containing a pulsating Mira variable. Hydrodynamic simulations(Mohamed and Podsiadlowski, 2012) show that the wind can be confined within the RG’sRoche lobe and strongly focused toward the binary orbital plane. Such a wind Roche-lobeoverflow (WRLOF) can be so efficient to allow the WD to accrete ∼
50% of the RG’smass loss and not just the few % typical of a Bondi-Hoyle-Littleton dynamical cross-section. The WRLOF, as other means of boosting the efficiency of mass accretion fromwind (Bisikalo et al., 2006; Skopal and Carikov´a, 2015; Pan et al., 2015), would also of-fer a way out to the evolutionary paradox posed by the yellow
SySts. They are a smallgroup, including both burning and accreting-only cases. Their RG are G/K-type giants,with Halo kinematics and low metallicity ([Fe/H] ≤− s -type elements (mostnotably barium), which are normally brought to surface during third dredge-up at the tipof the AGB. They lack however the presence of unstable Tc isotopes and are less luminousthan the tip of the AGB, indicating and extrinsic origin of the s -type elements, i.e. pollu-tion from the progenitor of the current WD companion (Jorissen et al., 2005; Pereira et al.,2017). Some of these SySts are rotating at a significant fraction of their rotational break-upvelocities ( V rot sin i ≥
100 km s − ), suggesting a massive transfer of both mass and angularmomentum from the progenitor of the current WD. The distribution of orbital periods andeccentricities of Barium SySts require that dynamically unstable mass transfer by Rochelobe overflow (and the resulting common-envelope phase with its orbit shrinking and cir-cularization), is avoided and massive transfer of mass and angular momentum be achievedvia wind. All-sky surveys as well as pointed X-ray observations (with the
Swift satellite in particular)are discovering a population of optically unconspicous RG that emits in hard X-rays, a factrequiring them to pair in a binary system with a WD or a NS. Their relatively low X-ray( ≈ ⊙ ) and UV ( ≈ ⊙ ) luminosities currently limits the serendipitous discoveryto systems within ∼ he Symbiotic Stars Table 6.1
Different types of X-ray emission observed in Symbiotic Stars,their likely origin and some of the best known examples in each class.
Type Description Examples α Super-soft, photon energies ≤ β Soft, photon energies ≤ δ Hard, absorbed, with thermal emission SU Lyn, 4 Dra, T CrBdetectable at ≥ β / δ characteristics of both β and δ type NQ Gem, CH Cygsimultaneously present, from colliding MWC 560winds and disk/WD boundary layer γ Absorbed NS accretor, pulsed by NS GX 1+4, V934 Herspin, optically thick Comptonized plasma 4U 1954+31 is well epitomized by SU Lyn, a V ∼ Swift -BAT sources, Mukai et al. (2016) noted that SU Lyn lied within the error box of oneof them. Follow-up observations were organized with
Swift (to refine the astrometric po-sition of the BAT source and better characterize its UV and X-ray emission) and with theAsiago spectrographs (to investigate if SU Lyn optical spectra could betray peculiaritiessupporting a physical association with the
Swift -BAT source). Some results are summa-rized in Figure 6.3. While the optical spectrum of SU Lyn is identical to that of a normalM6III giant, its bluest part ( λ ≤ Swift
UVOT telescope and the soft and hard X-ray emission observed by
Swift
XRT and BAT instruments. Only high-resolution Echelle spectra can reveal a feeble,structured and quite variable emission in H α .Similar tortuous paths affect the discovery of SySts hosting a NS, usually named sym-biotic X-ray binaries (or SyXBs; Masetti et al., 2007). SyXBs are quite rare: among the ∼
200 low-mass X-ray binaries (LMXBs) known in the Galaxy, only ∼
10 SyXBs cases arecurrently known. Observationally, these systems are characterized by appreciable X-rayemission ( ∼ –10 erg s − ) positionally associated with a RG star which spectroscop-ically does not show any abnormal features, with the possible exception of a continuumexcess in the blue and ultraviolet ranges (similar to what illustrated in Figure 6.3 for SULyn). The X-ray emission is pulsed (periods from 10 to 10 seconds, 4U 1954+319 be-ing the slowest at P spin ∼ Ulisse Munari a strong spin-up of − × − hr hr − during outbursts and a spin-down of 2.1 × − hrhr − in quiescence. The level of X-ray emission can vary up to four orders of magnitude,suggesting accretion from an inhomogeneous stellar wind and possibly coupled with anhighly elliptical orbit of the accretor. A notable outlier is GX 1+4, which emits in X-raysup 10 erg s − , with P spin ∼
120 s (Chakrabarty and Roche 1997), and an optical spectrum(Munari and Zwitter 2002) quite similar to those of burning SySts. A geometrically-thinand optically-thick accretion disk heavily irradiated by the hard X-rays from the centralNS would provide the UV-source needed to photo-ionize the RG wind: plasma diagnosticshows in fact that the gas in GX 1+4 is ionized by thermal UV radiation ( T ph ∼ . Normal SySts frequently enjoy outburts, that usual come in trains of a few individualepisodes separated by longer periods spent at quiescence. The example of CI Cyg in Fig-ure 6.1 is indicative: three different maxima ( α , β , γ ), of declining strength and duration,separated in time by ∼ B band, and de-clines toward the red. The color evolution of CI Cyg in Figure 6.1 well illustrates how theoutburst status is barely detectable in the I -band, where the flux is dominated by the RGat all phases. This type of frequent and multiple outbursts is called Z-type or Z-And typefrom Z Andromedae, a prototype SySt. Jets (frequently seen bi-polar in high-res spectra)have been observed in about 12 symbiotic stars (for 1/3 of them also spatially resolved;cf. the spectacular images for R Aqr by Schmid et al. (2017)), and in most cases they areassociated to Z-And outbursts. The projected jet velocities are of the order of 1000 − − , equivalent to the escape velocity from the region closest to the central WD, withthe noteworthy exception of MWC 560 where velocities V ej ≥ − were observed(Tomov et al., 1990). In response to the great differences seen in the jets from one ob-ject to another, a variety of launching mechanisms have been proposed and modelled (eg.Stute and Sahai, 2007; Skopal et al., 2009; Tomov et al., 2011). Whereas RGs in SySts ro- he Symbiotic Stars a is an accretion event, b an expansion of the burning shell once the accreted material reaches it. The arrows points to timesfor the representative spectra in the bottom panel. F λ (H α ) is the integrated flux of H α (in units of10 − erg cm − s − ), and T the photo-ionization temperature of the hot source. tate faster than field RGs and appear synchronized with orbital period (P rot ≃ P orb ), thosein systems emitting jets seem to rotate faster at P rot < P orb (Zamanov and Stoyanov, 2012).As of the causes of Z-And type outburst, a great variety of different mechanisms have Ulisse Munari been invoked (eg. Bisikalo et al., 2006; Tomov et al., 2011; Skopal et al., 2011; Ramsay et al.,2016; de Val-Borro et al., 2017) like a sudden increase in the mass-transfer rate from theRG, either triggered by intrinsic variability of the RG or its passage at periastron; theformation of an optically thick, cool, disk-shaped zone around the WD equator as a con-sequence of enhanced wind from the WD; an enhanced wind from the WD which leadsinstead to the disruption of the inner part of the accretion disk with the formation of hollowcones around the WD axis of rotation and thus to the appearance of collimated outflows;changes in the kinematical regime of colliding winds from the WD and the RG, etc. Acoordinated effort between X-ray, ultraviolet, optical, and radio observations to follow indetail and over all the relevant phases the Z-And outbursts of at least a few SySts seemsrequired to rise firm constraints useful in guiding future modeling efforts.Basically, the explanations for Z-And type outbursts tend to cluster into two broad cat-egories: ( a ) release of potential energy from extra-accreted matter, or ( b ) shift to longerwavelengths of the emission from WD burning shell of the WD, expanding in radius asresponse to an increase in the mass accretion rate. Both modes could occur in successionin the same object, as illustrated in Figure 6.4 by the 2016/17 outburst of StH α . A very few ( ∼
10) Symbiotic Novae (SyN) have been seen to erupt in historical times inour Galaxy, while others were already in outburst when discovered as SySts. They shouldnot be confused with the novae erupting in symbiotic binaries described below. The out-bursts of SyN last about a century, are of large amplitude and only one eruption has beenrecorded, with the possible exception of BF Cyg that shortly after returning to quiescencefrom the outburst initiated in 1894 (Leibowitz and Formiggini, 2006), it started a new SyNcycle in 2006 and currently is still at maximum brightness.The typical photometric and spectroscopic evolution of SyN are illustrated in Figure 6.5.A large rise in brightness takes the star, in about ∼ he Symbiotic Stars V -band light-curve of the symbiotic nova V4368 Sgr (= Wakuda’s object) is givenin the middle panel, with marked by arrows the epochs of the sample spectra shown in the bottompanel. For comparison, the upper panel displays the light-curve of V1016 Cyg, another SyN. TNR of a classical nova and the consequent violent mass ejection. To adjust to the largenuclear luminosity produced at its base, the nondegenerate envelope expands to supergiantdimensions. The absence of massive ejection retains most of the mass in the WD envelope,which keeps burning under stable conditions for a long time, in excess of the ∼ century aSyN takes to return to quiescence. AG Peg has only recently returned to pre-SyN brightnessafter the SyN outburst initiated around 1850: now it is a normal-burning SySt that experi-ences normal Z-And type outbursts (Tomov et al., 2016; Ramsay et al., 2016). If accretion Ulisse Munari
Figure 6.6 The peculiar spectral evolution of a nova erupting within a symbiotic binary (NwSySt) iswell illustrated by the 2010 outburst of V407 Cyg.
Bottom: portion of an Echelle spectrum obtained+2.3 days past optical maximum, with the simultaneous presence of very sharp (FWHM ≤
20 km s − )and broad lines (FWHM ∼ − ). Sharp lines are produced by recombination within the flash-ionized wind of the red giant. Broad lines comes from the expanding ejecta of the nova. Top: sampleof H α profiles at various epochs showing the quick disappearance of the narrow component, theviolent deceleration of the broad one, and the persistent presence of a narrow absorption originatingin the outer wind of the RG unperturbed by the nova outburst. cannot keep pace with hydrogen depletion by nuclear burning, sooner or later the shell onthe WD in AG Peg will slim under the critical value, the burning will stop, and the star willmove back to the anonymity typical of accreting-only SySts. After quietly accreting for anappropriately long interval, it will be ready for the next symbiotic cycle to be initiated bya new century-long SyN outburst. . Under proper balance between mass loss and gain, the WDs of symbiotic stars can grow he Symbiotic Stars in mass toward the Chandrashekar limit, with the bright prospect of concluding their lifewith a spectacular explosion as type Ia supernovae (Munari and Renzini, 1992). Approach-ing that limit, the WDs become so massive that normal nova explosions could occur onsuch a short time scale that more than one has been observed in historical times. Mostfamous recurrent novae among SySts are RS Oph (7 outbursts), V745 Sco (3), T CrB (2),and V3890 Sgr (also 2 outbursts).A nova erupting within a SySt (NwSySt) evolves quite differently from a classical one.When the TNR culminates with an intense UV flash (Starrfield et al., 2016), the RGwind absorbs most/all of it, get ionized, and soon start glowing under recombination. TheNwSySt is taken almost instantaneously to peak optical brightness, whereas in classicalnovae the UV flash disperses in the surrounding emptiness and goes unnoticed hours/daysbefore the nova is discovered, and well before peak brightness is attained at the time ofmaximum expansion for the pseudo-photosphere of the optically thick ejecta. Given thelarge electron density in the wind of the RG at the distance it is orbited by the WD (10 –10 cm − ), the recombination from the UV flash proceeds rapidly in NwSySts ( e -folding time3–6 days). The recombining wind is not kinematically perturbed, consequently the linesit emits remain very sharp (FWHM ≃
20 km s − ). In the meantime, material is ejected athigh velocity from the central nova (FWHM of thousands km s − ), producing very wideemission lines (of the He/N nova type) co-existing with the narrow ones (Figure 6.6, bot-tom panel). The fast ejecta ram onto the pre-existing wind of the RG, and upon sweepingit up they are violently decelerated causing a rapid narrowing of the broad-lines profiles( e -folding time of a few days; Figure 6.6 top panel). γ -rays in the GeV range are thenproduced as a consequence of the violent shock. The best documented NwSySt eruptionis probably that of 2010 for the symbiotic Mira V407 Cyg (Munari et al., 2011; Pan et al.,2015). References
Akras, S., Guzman-Ramirez, L., Leal-Ferreira, M., Ramos-Larios, G. 2018. A new symbi-otic stars catalogue using 2MASS and WISE. I.
MNRAS , submitted.Allen, D. A. 1984. A catalogue of symbiotic stars.
Proc. Astron. Soc. Australia , , 369–421.Angeloni, R., and 12 colleagues 2014. Symbiotic stars in OGLE data - I. Large MagellanicCloud systems. MNRAS , , 35–48.Belczy´nski, K., Mikołajewska, J., Munari, U., Ivison, R. J. and Friedjung, M. 2000. Acatalogue of symbiotic stars. A&AS , , 407–435.Bisikalo, D. V., Boyarchuk, A. A., Kilpio, E. Y., Tomov, N. A., Tomova, M. T. 2006. Astudy of the outburst development in the classical symbiotic star Z And within thecolliding-winds model. ARep , , 722–732.Blind, N., Boffin, H. M. J., Berger, J.-P., Le Bouquin, J.-B., M´erand, A., Lazareff, B., Zins,G. 2011. An incisive look at the symbiotic star SS Leporis. Milli-arcsecond imagingwith PIONIER/VLTI. A&A , , A55.Chakrabarty, D., Roche, P. 1997. The Symbiotic Neutron Star Binary GX 1+4/V2116 Ophi-uchi. ApJ , , 254–271. Ulisse Munari
Corradi, R. L. M., and 21 colleagues 2010. IPHAS and the symbiotic stars . II. New dis-coveries and a sample of the most common mimics.
A&A , , A41.de Val-Borro, M., Karovska, M., Sasselov, D. D., Stone, J. M. 2017. Three-dimensionalhydrodynamical models of wind and outburst-related accretion in symbiotic systems. MNRAS , , 3408–3417.Friedjung, M., Viotti, R. 1982. The nature of symbiotic stars. IAU Coll. 70. Ap&SpS. Libr. , . Dordrecht: Reidel.Hinkle, K. H., Fekel, F. C., Joyce, R. R., Wood, P. 2013. Infrared Spectroscopy of Symbi-otic Stars. IX. D-type Symbiotic Novae. ApJ , , 28.Jorissen, A., Zaˇcs, L., Udry, S., Lindgren, H., Musaev, F. A. 2005. On metal-deficientbarium stars and their link with yellow symbiotic stars. A&A , , 1135–1148.Kenyon, S. J. 1986. The symbiotic stars. Cambridge: Cambridge University Press.Kenyon, S. J., Webbink, R. F. 1984. The nature of symbiotic stars. ApJ , , 252–283.Lee, H.-W., Heo, J.-E., Lee, B.-C. 2014. Raman-scattered Ne VII 973 at 4881 ˚A in thesymbiotic star V1016 Cygni. MNRAS , , 1956–1962.Leibowitz, E. M., Formiggini, L. 2006. Multiperiodic variations in the last 104-yr lightcurve of the symbiotic star BF Cyg. MNRAS , , 675–681Luna, G. J. M., Sokoloski, J. L., Mukai, K., Nelson, T. 2013. Symbiotic stars in X-rays. A&A , , A6.Marcu, D. M., and 11 colleagues. 2011. The 5 hr Pulse Period and Broadband Spectrum ofthe Symbiotic X-Ray Binary 3A 1954+319. ApJL , , L11.Masetti, N., and 10 colleagues. 2007. IGR J16194-2810: a new symbiotic X-ray binary. A&A , , 331–337Merrill, P. W., Humason, M. L. 1932. A Bright Line of Ionized Helium, λ PASP , , 56–57.Mikołajewska, J., Kenyon, S. J. 1992. On the nature of the symbiotic binary AX Persei. AJ , , 579-592.Mikołajewska, J., Caldwell, N., Shara, M. M. 2014. First detection and characterization ofsymbiotic stars in M31. MNRAS , , 586–599.Mikołajewska, J., Shara, M. M., Caldwell, N., Iłkiewicz, K., Zurek, D. 2017. A survey ofthe Local Group of galaxies for symbiotic binary stars. I. MNRAS , , 1699–1710.Miszalski, B., Mikołajewska, J., Udalski, A. 2013. Symbiotic stars and other H α emission-line stars towards the Galactic bulge. MNRAS , , 3186–3217.Mohamed, S., Podsiadlowski, P. 2012. Mass Transfer in Mira Binaries. BalA , , 88–96.Mukai, K., and 8 colleagues. 2016. SU Lyncis, a hard X-ray bright M giant: clues point toa large hidden population of symbiotic stars. MNRAS , , L1–L5.Munari, U., Buson, L. M. 1994. The ultraviolet spectra of the symbiotic stars MWC 960,FN Sgr, SS 29 and Draco C-1. A&A , , 87–94.Munari, U., Renzini, A. 1992. Are symbiotic stars the precursors of type Ia supernovae?. ApJLett , , L87–L90.Munari, U., Zwitter, T. 2002. A multi-epoch spectrophotometric atlas of symbiotic stars. A&A , , 188–196.Munari, U., and 8 colleagues. 2011. The 2010 nova outburst of the symbiotic Mira V407Cyg. MNRAS , , L52–L56.Munari, U., Dallaporta, S., Cherini, G. 2016. The 2015 super-active state of recurrent novaT CrB and the long term evolution after the 1946 outburst. New Astronomy , , 7–15.Murset, U., Nussbaumer, H. 1994. Temperatures and luminosities of symbiotic novae. A&A , , 586–604. he Symbiotic Stars Murset, U., Nussbaumer, H., Schmid, H. M., Vogel, M. 1991. Temperature and luminosityof hot components in symbiotic stars.
A&A , , 458–474.Murset, U., Wolff, B., Jordan, S. 1997. X-ray properties of symbiotic stars. A&A , ,201–210.Nu˜nez, N. E., Nelson, T., Mukai, K., Sokoloski, J. L., Luna, G. J. M. 2016. Symbiotic Starsin X-Rays. ApJ , , 23.Pan, K.-C., Ricker, P. M., Taam, R. E. 2015. Simulations of the Symbiotic Recurrent NovaV407 CYG. I. Accretion and Shock Evolutions. ApJ , , 27.Pereira, C. B., Baella, N. O., Drake, N. A., Miranda, L. F., Roig, F. 2017. High-resolutionOptical Spectroscopic Observations of Four Symbiotic Stars: AS 255, MWC 960, RWHya, and StH α ApJ , , 50.Proga, D., Kenyon, S. J., Raymond, J. C. 1998. Illumination in Symbiotic Binary Stars:Non-LTE Photoionization Models. II. WIND Case. ApJ , , 339–356.Ramsay, G., Sokoloski, J. L., Luna, G. J. M., Nu˜nez, N. E. 2016. Swift observations of the2015 outburst of AG Peg. MNRAS , , 3599–3606.Rodr´ıguez-Flores, E. R., and 8 colleagues. 2014. IPHAS and the symbiotic stars. III. A&A , , A49.Schmid, H. M. 1989. Identification of the emission bands at 6830, 7088 A. A&A , ,L31–L34.Schmid, H. M., and 42 colleagues 2017. SPHERE/ZIMPOL observations of R Aqr. I.Imaging of the stellar binary and the innermost jet clouds. A&A , , A53.Seaquist, E. R., Taylor, A. R., Button, S. 1984. A radio survey of symbiotic stars. TheAstrophysical Journal 284, 202-210.Skopal, A., Carikov´a, Z. 2015. Wind mass transfer in S-type symbiotic binaries. I. Focusingby the wind compression model. A&A , , A8.Skopal, A., Pribulla, T., Budaj, J., Vittone, A. A., Errico, L., Wolf, M., Otsuka, M.,Chrastina, M., Mikul´aˇsek, Z. 2009. Transient Jets in Z And. ApJ , , 1222–1235.Skopal, A., and 14 colleagues. 2011. Formation of a disk structure in the symbiotic binaryAX Persei during its 2007-10 precursor-type activity. A&A , , A27.Sokoloski, J. L. 2003. Rapid variability as a diagnostic of accretion and nuclear burning insymbiotic stars and supersoft X-ray sources. ASPC , , 202–217Starrfield, S., Iliadis, C., Hix, W. R. 2016. The Thermonuclear Runaway and the ClassicalNova Outburst. PASP , , 051001.Stute, M., Sahai, R. 2007. Hydrodynamical Simulations of the Jet in the Symbiotic StarMWC 560. III. Application to X-Ray Jets in Symbiotic Stars. ApJ , , 698–706.Tomov, N. A., Bisikalo, D. V., Tomova, M. T., Kil’Pio, E. Y. 2011. Interpretation of theLine Spectrum of Classical Symbiotic Stars. AIPCS , , 35–44.Tomov, T., Kolev, D., Zamanov, R., Georgiev, L., Antov, A. 1990. MWC560 - A uniqueastrophysical object. Nature , , 637.Tomov, T. V., Stoyanov, K. A., Zamanov, R. K. 2016. AG Pegasi: now a classical symbioticstar in outburst ?. MNRAS , , 4435–4441.Whitelock, P. A. 2003. A Comparison of Symbiotic and Normal Miras (invited reviewtalks). ASPC , , 41–56Whitelock, P. A., Munari, U. 1992. Photometric properties of symbiotic stars and the natureof the cool component. A&A , , 171–180.Zamanov, R. K., Stoyanov, K. A. 2012. Rotation of the red giants and white dwarfs insymbiotic binary stars. Bulg.A.J , , 41–52Zamanov, R. K., and 11 colleagues 2017. Discovery of optical flickering from the symbi-otic star EF Aquilae. AN ,338