We have studied how the third dredge-up and the carbon star formation in low-mass Asymptotic Giant Branch stars depends on certain key nuclear reaction rates .
We find from a set of complete stellar evolution calculations of a 2 { M } _ { \odot } model with Z = 0.01 including mass loss , that varying either the ^ { 14 } \mathrm { N } ( \mathrm { p } , \gamma ) ^ { 15 } \mathrm { O } or the triple- \alpha reaction rate within their uncertainties as given in the NACRE compilation results in dredge-up and yields that differ by a factor of 2 .
Model tracks with a higher rate for the triple- \alpha rate and a lower rate for the ^ { 14 } \mathrm { N } ( \mathrm { p } , \gamma ) ^ { 15 } \mathrm { O } reaction both show more efficient third dredge-up .
New experimental results for the ^ { 14 } \mathrm { N } ( \mathrm { p } , \gamma ) ^ { 15 } \mathrm { O } reaction rates are surveyed , yielding a rate which is about 40 \% lower than the tabulated NACRE rate , and smaller than NACRE ’ s lower limit .
We discuss the possible implications of the revised nuclear reaction stellar evolution calculations that aim to reproduce the observed carbon star formation at low mass , which requires efficient third dredge-up .