It depends on the size of your plasmids. If they're smallish, you might be able to align them using Muscle or Mafft. Otherwise, or in the case where synteny might not be conserved across plasmids, you ought to use a genome aligner like Mauve.

Unless your plasmids are very small, your best bet to get a phylogeny that will finish in a reasonable time is RAxML, assuming you want to do some sort of model-based phylogenetics. With a dataset of this size, most Bayesian approaches will not converge quickly.