Rock blocks, wreath products and KLR algebras (Q1674592): Difference between revisions

For any integer \(d\geq 0\), \textit{J. Chuang} and \textit{R. Rouquier} showed in [Ann. Math. (2) 167, No. 1, 245--298 (2008; Zbl 1144.20001)] that any two blocks of symmetric groups \(S_n\) and \(S_m\) (or more generally, of their Iwahori-Hecke algebras \(H_n\) and \(H_m\)) of weight \(d\) are derived equivalent. For each weight, there are certain special blocks known as RoCK blocks, also sometimes called Rouquier blocks, that are more tractable than general blocks. For the symmetric groups, \textit{J. Chuang} and \textit{R. Kessar} showed in [Bull. Lond. Math. Soc. 34, No. 2, 174--184 (2002; Zbl 1033.20009)] that, for symmetric groups, RoCK blocks of weight less than the field characteristic \(p\) are Morita equivalent to the wreath product \(S_{\varnothing,1} \wr S_d\), where \(S_{\varnothing,1}\) denotes the principal block of \(\mathbb{F}S_p\) symmetric group. This was known to be false for blocks of weight \(d \geq p\), as the RoCK block has more simple modules than \(S_{\varnothing,1} \wr S_d\). However, in [Mem. Am. Math. Soc. 947, vii, 102 p. (2009; Zbl 1269.20010)], \textit{W. Turner} conjectured that \(S_{\varnothing,1} \wr S_d\) is Morita equivalent to a certain idempotent truncation of a RoCK block. The proof of this conjecture is the purpose of the article under review. In fact, the author proves a more general statement that applies to the Hecke algebra of the symmetric group. Namely, the author shows that a certain idempotent truncation of a RoCK block of the Hecke algebra is isomorphic to \(H_{\rho,0} \otimes_\mathbb{F} (H_{\varnothing,1} \wr S_d)\), where \(H_{\rho,0}\) is the corresponding block of weight 0, and is thus a split semisimple algebra, whence the Morita equivalence between a RoCK block and \((H_{\varnothing,1} \wr S_d)\) follows. The proof uses (a special case of) the Brundan-Kleshchev isomorphism between Hecke algebras and cyclotomic KLR algebras of type \(A\) to transport to problem to the setting of the latter, where a non-trivial \(\mathbb{Z}\)-grading is put to good use. Finally, it should be mentioned that Turner also made conjectures about the whole RoCK block when \(d \geq p\). He constructed an algebra that he expected to be Morita equivalent to the whole RoCK block, rather than just to the idempotent truncation handled in this paper. Building on the present work, the author and \textit{A. Kleshchev} have proved this conjecture in [Ann. Math. (2) 188, No. 2, 453--512 (2018; Zbl 1435.20010)].