Standard stellar evolution theory predicts for low-mass stars in globular clusters only minor changes in surface composition during the red giant phase. However, observations over the past two decades have shown a large variation in the abundances of C, N, O, Na, Mg, and Al among globular cluster red giant stars (see R. P. Kraft, C. Sneden, G. H. Smith, M. D. Shetrone, G. E. Langer, & C. A. Pilachowski, AJ, 115, 1500 [1998]). Moreover, it has been found that the abundances of O and Na are anticorrelated in a number of globular clusters, as are the abundances of Mg and Al. These correlations are thought to be the result of deep mixing of material from the surface of the star into regions where the temperature is high enough to allow proton captures to take place. Additional observations of M13 red giant branch stars show an anticorrelation between 24Mg and Al and that the abundance sum 25Mg]26Mg is approximately constant for large variations in Al (M. D. Shetrone, AJ, 112, 2639 [1996]). Production of Al and destruction of 24Mg depend on the 24Mg(p, c)25Al reaction rate at stellar temperatures 0.04 (where T9D T9 109 K). However, the commonly used rate (G. R. Caughlan & W. A. Fowler, At. Data Nucl. Data Tables, 40, 283 [1998]) is too low to account for the observations within the present mixing models. In the stellar temperature range of the T9 0.02E0.055, 24Mg]p reaction rate is determined by contributions from the lowest lying resonance at keV and the direct E R 223 capture process into Ðnal bound states of 25Al. Direct proton capture on 24Mg has been studied extensively by H. P. Trautvetter, & C. Rolfs (Nucl. Phys., A242, 519 [1975]). The strength of the keV resonance has E R 223 also been measured previously. Based on this experimental information, the reaction rate at low stellar temperatures has been calculated by Caughlan & Fowler (1988). It was pointed out recently by C. S. Zaidins & G. E. Langer (PASP, 109, 252 [1997]) that at the temperatures important for globular cluster red giant stars, the Gamow window is located far below the energy of the keV resoE R 223 nance. Consequently, the contribution from the low-energy wing of this resonance has to be taken into account. Accurate calculation of the wing contribution requires additional information regarding resonance parameters (such as the proton partial width c-ray partial width and the p , !c, total width !). However, only an experimental upper limit of ! 32 eV has been reported in the published literature (M. Uhrmacher, K. Pampus, F. J. Bergmeister, D. Purschke, & K. P. Lieb, Nucl. Instrum. Meth., B9, 234 [1985]). It was estimated by Zaidins & Langer (1997) that the 24Mg]p reaction rate at can increase by a factor of 32 due T9 0.04 to the contribution of the resonance wing if the experimental upper limit for is adopted. The main goal of this thesis was to accurately determine the properties of the E R 223 keV resonance in order to improve the reaction rate estimate for 24Mg(p, c)25Al. Resonance strengths uc, c-ray branching ratios and !c/!, mean lifetimes were measured for the keV resoE R 223 nance in 24Mg(p, c)25Al at the Triangle Universities Nuclear Laboratory. The measured resonance strength of uc 12.7^ 0.9 meV and branching ratio of !c/! 0.91 ^ 0.04 allow for the determination of and !. The !p, !c, measurement of the mean lifetime of fs via the q ~2.4 `2.9 Doppler shift attenuation method also allows for an independent determination of (by using q +/!) and therefore provides a test of internal consistency. The present experimental results establish the rate for the 24Mg(p, c)25Al reaction for temperatures of T9 0.02E2. The reaction rate deviates from the previous estimates (Caughlan & Fowler 1988) by 18%E45% with statistical uncertainties of 5%E21%, and therefore the suggestion of Zaidins & Langer (1997) that the total width of the E R 223 keV resonance may have a signiÐcant inNuence on the total 24Mg]p reaction rate for can be ruled out. T9B 0.04 P. A. Denissenkov, G. S. Da Costa, J. E. Norris, & A. Weiss (A&A, 333, 926 [1998]) suggested that the existence of an undetected low-energy resonance could also enhance the 24Mg(p, c)25Al reaction rate. Based on our current experimental and theoretical understanding of the 25Al level structure, this possibility is extremely unlikely. Our results suggest that hydrogen burning of 24Mg at temperatures below cannot account for the anticorrelation of T9 0.055
Read full abstract