Comment on ‘Twisted bialgebroids versus bialgebroids from a Drinfeld twist’

A class of left bialgebroids whose underlying algebra <?CDATA A H?> A ♯ H is a smash product of a bialgebra H with a braided commutative Yetter–Drinfeld H -algebra A has recently been studied in relation to models of field theories on noncommutative spaces. In Borowiec and Pachoł (2017 J. Phys. A: Math. Theor. 50 055205) a proof has been presented that the bialgebroid <?CDATA AF HF?> A F ♯ H F where H F and A F are the twists of H and A by a Drinfeld 2-cocycle <?CDATA F = ∑ F¹⊗ F²?> F = ∑ F 1 ⊗ F 2 is isomorphic to the twist of bialgebroid <?CDATA A H?> A ♯ H by the bialgebroid 2-cocycle <?CDATA ∑ 1 F¹⊗ 1 F²?> ∑ 1 ♯ F 1 ⊗ 1 ♯ F 2 induced by F . They assume H is quasitriangular, which is reasonable for many physical applications. However the proof and the entire paper take for granted that the coaction and the prebraiding are both given by special formulas involving the R-matrix. There are counterexamples of Yetter–Drinfeld modules over quasitriangular Hopf algebras which are not of this special form. Nevertheless, the main result essentially survives. We present a proof with a general coaction and the correct prebraiding, and even without the assumption of quasitriangularity.