The classical theory of grain nucleation suffers from both theoretical and predictive deficiencies .
We strive to alleviate these deficiencies in our understanding of dust formation and growth by utilizing an atomistic model of nucleation .
Carbon cluster geometries are determined with a set of global minimization algorithms .
Using density functional theory , the binding energies of carbon clusters from n = 2 to n = 99 are then calculated .
These energies are used to calculate the critical size and nucleation rate of carbon clusters .

We find that the critical cluster size is largely determined by the changes in geometry of the clusters .
Clusters with size n = 27 and n = 8 , roughly corresponding to the transition from ring-to-fullerene geometry and chain-to-ring geometry respectively , are the critical sizes across the range of temperature and saturation where nucleation is significant .
In contrast to the classical theory , nucleation is enhanced at low-temperatures , and suppressed at high temperatures .
These results will be applied to a modified chemical evolution code using results from supernova simulations .