“…x u, ∇p ∈ L q (0, ∞; L p (R n + )), t β ∇u ∈ L q 2 (0, ∞; L p 2 (R n + )), t γ u ∈ L q 3 (0, ∞; L p 3 (R n + )) for some β, γ > 0, 1 < p 2 , p 3 , q 2 , q 3 < ∞ with α = β + γ, 1 p = 1 p 2 + 1 p 3 and 1 q = 1 q 2 + 1 q 3 . The limiting case q = ∞ has been studied by M. Cannone, F. Planchon, and M. Schonbek [5] for h ∈ L 3 (R 3 + )(⊂ B −1+ 3 p p∞ (R 3 + )), by H. Amann [3] for h ∈ b −1+ n p p,∞ (R n + ), p > n 3 , p = n, and by M. Ri, P. Zhang and Z. Zhang [26] for h ∈ b 0 n∞ (R n + ), where b s pq (Ω) denotes the completion of the generalized Sobolev space H s p (Ω) in B s pq (Ω). In particular, in [5], the solution exists globally in time when h…”