We would like to show the converse, that the clopen sets are base sets in the zariski topology.
Let e1 through en generate base open sets. The complement of their union is vH, where H is the ideal generated by e1 through en. Every finitely generated ideal is principal, so let f generate H. The union of our base sets is of, another base set.
If y is open and closed, then it is the union of base sets, and being a closed subset of a compact space, it is the finite union of base sets. Therefore y is a base set, and the open and closed sets are exactly the base sets.
Convert the boolean lattice to a ring, and then derive spec R. The clopen sets in spec R correspond to the elements of R, which are the elements of the lattice. For x and y in R, the intersection of ox and oy is oxy, and their union is ox+y+xy, since x+y+xy is the element that generates the principal ideal spanned by x and y. These are the meet and join respectively. The lattice has been mapped onto clopen sets, with join as union and meet as intersection.
Is the map injective? Given distinct points in the lattice, assume x is not below y. If the ideal R*y contains x then x = zy, and x is below y in the lattice, which is a contradiction. The ideal generated by y misses x, and the powers of x. Drive this up to a prime ideal containing y, and missing x. This shows vy ≠ vx, and clopen sets faithfully represent the lattice.