Theorem. (Generalized affinoid principal zero purity; F8ZP-G). Let be a nonzero generalized -affinoid algebra pure of analytic dimension . If avoids every minimal prime of , then is empty or pure of analytic dimension . In particular this holds for every non-zero-divisor . For , every element avoiding the minimal primes is a unit. Moreover, in the original algebra , for every and contained in , the ordinary height of in is one.
Proof
Proof. Put . This is an actual affinoid quotient by NAG-001561–001562. If , there is nothing to prove. Choose one finite -free polyradius , and put , such that and are strict; NAG-001557 permits this simultaneous choice. If is trivially valued, include any additional radius other than in the strictification request, so that is nontrivially valued. The quotient comparison and faithful flatness in NAG-001566 give
By NAG-002291, the minimal primes of are exactly for , with contraction and with the same componentwise analytic dimension. Thus is pure of dimension , and avoids each of its minimal primes. Analytic and Krull dimensions agree for these strict algebras by NAG-001664. Proposition C.1.21 makes pure of dimension (and rules out when ).
Apply NAG-002291 again, now to the full quotient . For every , the prime is minimal and This is the required descent of componentwise analytic purity. Nonzero affinoids have nonempty spectra by NAG-001572, so the empty-space and zero-algebra alternatives agree. The regular case follows from Lemma C.1.13; its regularity also persists upstairs by the flatness in NAG-001566. The ordinary height assertion in the original follows by the minimal-prime and Tag 00KV argument of Proposition C.1.21: it uses only that is a Noetherian domain and that is nonzero there. It is the strictification argument, not this ordinary height statement alone, that supplies the asserted analytic dimension.
Only this finite free-radius extension, with its exact minimal-prime bijection, was used. No assertion of purity ascent by arbitrary flat extension, of freeness over completed residue fields, or of equality between generalized analytic dimension and ordinary Krull dimension has entered the argument. ◻