sign in sign up
<<
Home , Giant star, Red giant

NEWS AND VIEWS ------~------

are fused into a nucleus) occurs Flashing giants forge fluorine in such a thin shell that as energy is released, the temperature rises much G. J. Mathews faster than the can expand and cool. This leads to a thermonuclear runaway FLlIORINE is one of the few chemical stitute the only available information on which is arrested only by the develop­ elements whose stellar origin is still the abundance of fluorine outside the ment of a deep zone which being debated_ There arc two reasons . and they provide strong dredges material from the thermonuclear for this. For one. the only stable fluorine evidence that fluorine is produced deep burning regions all the way to the sur­ isotope. 19F. is fragile :lI1d easily des­ within the interiors of giant face of the star. This process of episodic troyed by proton capture reactions at during episodic flashes of thermonuclear -burning shell flashes implies that stellar temperatures. Most elements of burning. as these stars evolve, their surfaces be­ similar mass to fluorine are produced stars are identified accord­ come enriched in the products of ther­ during the late stages of thermonuclear ing to their spectral class (or redness). In monuclear burning. burning in massive stars. However, order of increasing redness (or decreas­ It is therefore significant that these fluorine is so quickly destroyed that it ing surface temperature). the types of measurements have revealed fluorine tends to be bypassed by normal stellar red giant stars studied in this work were abundances as much as 30 times those in evolution. The second reason is that. those denoted K. Ba, M, MS, S, Sc, N the . This trend can be seen in the although clues to the origin of many and J. These different spectral types can graph, taken from Jorissen et al. I, which elements can be obtained from stellar roughly be identified with ages of red shows the logarithm of the ratio of surface abundances. it has been difficult giant stars as they evolve along what is fluorine to , [FlO], (relative to to find fluorine in stars. owing to a lack called the asymptotic giant branch2 (see the mean ratio for K and M red giants) of useful atomic fluorine lines in the box). This is a phase in which stars as a function of the carbon to oxygen visible region of the spectrum. This lat­ become progressively more luminous ratio, CIO. The various symbols repre­ ter difficulty was overcome recently by and red as thermonuclear burning occurs sent different spectral types. For exam­ Jorissen. Smith and Lamberti. who in shells outside of an inert carbon­ ple, the open squares show K-stars obtained fluorine abundances on the sur­ oxygen core while the outer envelopes whose fluorine abundance is not much faces of various red giant stars from the become extended. There is a peculiarity different from that of the Sun. The more 2 3 infrared rotation-vibration lines of the in the structure of these stars . . Helium evolved M stars ( open circles) show HF molecule. Their measurements con- burning (by which three helium nuclei increases in fluorine and carbon abun­ dances relative to oxygen as do S (closed circles) N (crosses) and J stars (open Asymptotic-giant-branch stars triangles). The trend of increasing car­ bon relative to oxygen with spectral type Top, structure of a red as it is an expected outcome of the dredge-up evolves along the asymptotic giant process which brings the carbon ashes of branch. The outer convective envelope helium burning to the stellar surface. extends to hundreds of times the diam• The correlation of increasing fluorine eter of the Sun and continuously mixes along with increasing carbon strongly material from its base to the surface. suggests that fluorine is also being manu­ Below this, a -burning shell factured by the star. produces helium (predominantly via the An important question, however, is 'eNO' cycle for stars more massive than the Sun). The radius of this shell is how fluorine could be produced and actually only a few times the radius of survive at the temperature of the helium­ the and would barely be visible if shell flashes. The authors have investi­ this figure were drawn to scale. This gated several possibilities and conclude radius, however, encompasses a mass that the most likely reaction sequence comparable to that of the Sun. Slightly involves the capture of an a-particle below the hydrogen-burning shell (a (helium nucleus) by 14N to give radio­ fraction of an Earth radius and a active 18F, which decays to give 180; thousandth of a away) is a proton capture by this then leads shell burning helium into carbon. All of indirectly to 19F - denoted this rests on an inert core of mostly 14N(a,y)18F(r)180(p,a)15N(a,y)19F. The carbon plus some oxygen formed during protons are produced by a pair of reac­ helium burning. tions involving BC and 14N. The fact that Below, evolution of the helium- and there are only a few protons present and hydrogen-burning shells. Periodically that the circulation removes fluorine (every 30-100 thousand years, depend• from the high-temperature regions saves ing upon the mass of the star) the the fluorine from destruction by proton helium shell ignites. Because the star capture. The fact that fluorine has not can not expand fast enough to dissipate been destroyed by the 19F(a,pf2Ne reac­ the heat generated, a thermonuclear runaway ensues. This is arrested by the tion also implies a constraint on the development of a convective shell (teardrop shape in this figure). The hydrogen• temperature and neutron source reaction burning shell then temporarily ceases and the outer convective zone briefly dips into for the slow neutron-capture (s-process) the region below the hydrogen-burning shell. This process dredges up material from nucleosynthesis of heavy elements4--6. At the helium-burning regions to the stellar surface. The observations of fluorine temperatures high enough to produce enhancement along with carbon on the surface of evolved red giants strongly neutrons via 22Ne(a,nf5Mg in helium­ suggests that fluorine is produced with carbon as a byproduct of helium burning and shell flashes, fluorine would be des­ subsequently dredged to the surface. G.J. M. troyed. Thus, 22Ne can not be the neu­ tron source for the s-process. This is 18 NATURE· VOL 359 . 3 SEPTEMBER 1992 © 1992 Nature Publishing Group