R Coronae Borealis Star

A star that hides behind its own smoke

Most variable stars dim and brighten because something about the star itself changes — it pulsates, an eclipsing companion passes in front, spots rotate across the face. An R Coronae Borealis (RCB) star does something stranger and far more dramatic: it manufactures soot, ejects it into the space directly between us and the photosphere, and disappears. For weeks the star can sit serenely at maximum light, indistinguishable from an ordinary bright supergiant. Then, without warning, its brightness collapses — sometimes by a factor of several thousand — over just a few weeks. Months later it crawls back to where it started. The luminosity of the star barely changed; we simply could not see it through the smoke.

The behaviour was first noticed by the English astronomer Edward Pigott in 1795, when the otherwise sixth-magnitude star R Coronae Borealis abruptly faded below naked-eye visibility. The class is sometimes summarised, only half-jokingly, as "the opposite of a nova": where a nova suddenly brightens, an RCB star suddenly vanishes. The defining light curve is a flat maximum punctuated by sharp, irregular, deep declines with slow recoveries — a signature unlike anything produced by pulsation or eclipses.

What an RCB star actually is

An RCB star is a luminous supergiant, with luminosity of order 10⁴ L☉ and an effective temperature typically between about 5000 K and 7000 K (a handful of hotter members reach toward 20000 K). Its radius is several tens to roughly a hundred solar radii. By these gross properties it resembles a yellow supergiant. What sets it apart is its bizarre chemistry. Where the Sun is about 90% hydrogen by number, an RCB atmosphere is hydrogen-deficient by factors of 10⁴ to 10⁶: hydrogen is essentially absent. The star is instead dominated by helium, with carbon enhanced to a few percent, and shows anomalies in nitrogen, oxygen, fluorine, and the heavy s-process elements.

This combination — supergiant luminosity, no hydrogen, abundant carbon and helium — places RCB stars in a tiny, exotic corner of the Hertzsprung-Russell diagram alongside the hotter extreme helium stars. They are not on any normal evolutionary track. Only about 150 are confirmed in the entire Milky Way, plus a few dozen in the Magellanic Clouds, making them among the rarest classified stars known.

The dust-formation mechanism

The deep minima are produced by carbon dust, not by intrinsic dimming. Because the atmosphere is carbon-rich and hydrogen-poor, almost all of the carbon is free to condense rather than being locked into carbon monoxide (in a hydrogen-rich star, oxygen would tie up the carbon). When the star ejects a parcel of gas — driven by pulsation and radiation pressure — and that parcel cools below roughly 1500 K, carbon condenses into amorphous-carbon soot grains only about 5 nanometres in radius. Such grains are spectacularly opaque per unit mass, so a freshly condensed cloud a few stellar radii out can block the photosphere almost completely if it happens to lie along our sightline.

The amount of dust required is astonishingly small. The opacity of a screen of grains is

τ = κ Σ        and        Δm = 1.086 τ   (magnitudes of extinction)

where κ is the mass absorption coefficient of the soot (of order 10⁴ cm² g⁻¹ for nanometre carbon grains at optical wavelengths) and Σ is the dust column density along the line of sight. To produce a 9-magnitude drop we need τ ≈ 8, which requires only Σ ≈ τ/κ ≈ 8 × 10⁻⁴ g cm⁻². Spread over a cloud the size of a few stellar radii, that is a dust mass of order 10⁻¹⁰ to 10⁻⁹ M☉ — comparable to or less than the mass of a large asteroid. A near-total stellar blackout is achieved with a wisp of carbon weighing as much as a small rock.

Anatomy of a deep minimum

The light curve of a deep minimum has a characteristic asymmetry: a fast plunge and a slow recovery. The decline is set by how quickly an opaque cloud can grow and drift across the disc — typically 20 to 50 days from maximum to the bottom, occasionally faster. The recovery is governed by the much slower process of pushing the cloud outward with radiation pressure, letting it expand, geometrically dilute, and finally clear the sightline; this takes months and sometimes one to three years. During recovery the spectrum often shows narrow emission lines from gas excited above the dust, then broad emission, then the photospheric absorption spectrum reasserting itself.

Phase	Timescale	What is happening	Observable
Maximum light	months – years	Clear line of sight to photosphere	Steady, near-constant brightness
Dust condensation	days	Ejected gas cools below ~1500 K; soot nucleates	Onset of fade
Decline	~20 – 50 days	Opaque cloud grows across the disc	Drop of 1 – 9 mag, near-vertical
Deep minimum	weeks – months	Photosphere fully obscured	Sharp emission lines from outer gas
Recovery	months – years	Radiation pressure disperses the cloud	Slow, irregular brightening

Superimposed on the deep minima is a low-amplitude pulsation of a few tenths of a magnitude on a period of roughly 40 to 100 days. These pulsations are thought to help trigger the mass-loss episodes that seed the dust, though dust forms in only a fraction of pulsation cycles and only sometimes ends up on our sightline — which is why the deep minima are irregular and unpredictable.

Where RCB stars come from

The peculiar chemistry — no hydrogen, surplus helium and carbon — demands an unusual history. Two models compete, and the evidence now strongly favours the first.

• Double-degenerate (white-dwarf merger) model. The favoured channel. A carbon-oxygen white dwarf and a helium white dwarf in a close binary spiral together by gravitational-wave emission and merge. The helium white dwarf is tidally shredded into a disc and accreted onto the carbon-oxygen core; helium burns explosively and then steadily on the surface. The merged object briefly inflates into a hydrogen-free, helium-and-carbon supergiant — an RCB star — for a few thousand years before contracting to a hot extreme-helium star and ultimately a white dwarf.
• Final-flash (born-again) model. A single post-asymptotic-giant-branch star, already on its way to becoming a white dwarf, experiences a very late thermal pulse — a final helium-shell flash — that ingests and burns its thin residual hydrogen envelope. The star balloons back to giant dimensions hydrogen-poor. Sakurai's Object (V4334 Sagittarii), observed in real time in 1996, is the textbook example of this born-again scenario, and a few RCB stars may form this way.

The decisive evidence is isotopic. RCB atmospheres show oxygen-18 enhanced to ¹⁶O/¹⁸O ratios near unity, versus the solar value of about 500, plus enhanced fluorine and, in some stars, lithium. Oxygen-18 is made copiously when helium burns on a nitrogen-14 seed — precisely the conditions when a helium white dwarf is dredged onto and burned atop a carbon-oxygen white dwarf. A single-star final flash does not naturally produce so much oxygen-18. The fingerprint therefore points to white-dwarf mergers as the dominant origin.

Real numbers

Property	Typical value	Comparison / note
Luminosity	~10⁴ L☉	Supergiant; like a yellow supergiant
Effective temperature	5000 – 7000 K	Hot RCBs (e.g. DY Cen) reach ~20000 K
Mass	~0.8 – 0.9 M☉	Merger of a ~0.6 + ~0.3 M☉ WD pair
Radius	~50 – 100 R☉	Inflated post-merger envelope
Hydrogen abundance	10⁻⁴ – 10⁻⁶ of solar	Sun is ~90% H by number
Carbon (by mass)	a few percent	Free to condense; not locked in CO
¹⁶O/¹⁸O ratio	~1	Solar value ~500 — merger signature
Dust grain size	~5 nm (amorphous carbon)	Soot; extremely opaque per gram
Dust mass per cloud	~10⁻¹⁰ – 10⁻⁹ M☉	Comparable to a large asteroid
Deep-minimum depth	1 – 9 mag	Factor ~2.5 to several thousand
Decline time	~20 – 50 days	Recovery: months to a few years
RCB lifetime	~10³ – 10⁴ yr	Brief phase; explains rarity
Known in Milky Way	~150	Plus a few dozen in Magellanic Clouds

The brevity of the RCB phase — only a few thousand years out of a stellar lifetime — combined with the rarity of white-dwarf mergers explains why so few exist at any one time. Surveys such as ASAS, OGLE, Gaia, and WISE (which finds RCBs efficiently via their warm circumstellar dust shells in the infrared) have roughly tripled the known sample in the past two decades, but the total remains small.

Famous examples

• R Coronae Borealis (R CrB). The naked-eye prototype, about 4000 light-years away in the constellation Corona Borealis. Normally near magnitude 6; it can drop below magnitude 14. Its famous 2007–2014 deep minimum was the longest and deepest on record, lasting essentially seven years before a full recovery.
• RY Sagittarii. The brightest member of the class, normally around magnitude 6 and the southern-hemisphere analogue of R CrB. A bright, well-monitored target whose declines have been studied for over a century.
• V854 Centauri. An unusually active RCB that fades frequently, with a somewhat higher hydrogen content than typical — a useful outlier for testing the formation models.
• DY Persei. The prototype of a cooler subclass (DYPer stars) that bridge RCB stars and ordinary carbon stars; their declines are slower and shallower.
• Sakurai's Object (V4334 Sagittarii). A born-again star caught undergoing a final helium-shell flash in 1996, evolving on a human timescale and developing RCB-like dust obscuration — the clearest case for the final-flash channel.

How we observe and confirm them

An RCB classification rests on three legs. First, the light curve: a flat maximum interrupted by abrupt, irregular, deep declines with slow recoveries — a pattern amateur observers and surveys like the AAVSO have monitored for over a century. Second, the spectrum at maximum light: strong carbon features (the Swan bands of C₂ and CN, the C I lines) together with weak or absent hydrogen Balmer lines and absent or very weak ¹³C bands. Third, an infrared excess from a warm (600–900 K) circumstellar dust shell, the cumulative debris of many past ejections, which WISE and Spitzer detect as a characteristic mid-infrared glow even when the star is at maximum in the optical.

During a decline, spectroscopy traces the obscuration directly: as the photosphere is blocked, sharp emission lines from gas above the dust appear, then broad emission, providing a tomographic view of the expanding cloud. Polarimetry of the light scattered by the asymmetric dust confirms that the obscuration is patchy and directional rather than a uniform shell — consistent with discrete puffs ejected in random directions.

Common misconceptions and edge cases

• "The star itself is dimming." No. The intrinsic luminosity is nearly constant through a deep minimum; what changes is the line-of-sight extinction. An observer at a different angle, looking through clear sky, would see the same star unchanged at maximum light while we watch it vanish.
• "RCB stars are just cool carbon stars." Ordinary carbon stars and Mira variables are hydrogen-rich AGB giants that pulsate smoothly. RCB stars are hydrogen-deficient supergiants whose deep minima are caused by dust, not pulsation, and are irregular rather than periodic.
• "It's like an eclipse." An eclipse is by an orbiting companion and is strictly periodic and predictable. RCB minima are produced by the star's own freshly minted dust, ejected at random phases and in random directions, so timing and depth are unpredictable.
• "Every pulsation makes a minimum." Dust condenses in only a fraction of cycles, and even then only obscures us if the puff lands on our particular sightline. Most ejected clouds miss us entirely, which is why minima are sparse and aperiodic.
• "All RCB stars are cool and yellow." A hot subclass (e.g. DY Centauri, near 20000 K) and a cool subclass (DYPer stars) bracket the canonical 5000–7000 K members, reflecting different evolutionary stages along the post-merger contraction track.
• "They are unrelated to gravitational-wave sources." As sub-Chandrasekhar white-dwarf merger products, RCB stars are siblings of the same double-degenerate population whose more massive members may produce Type Ia supernovae; studying RCB statistics constrains the cosmic rate of white-dwarf mergers.