Crystal bases for the quantum superalgebra $U_q(\mathfrak {gl}(m,n))$ (Q4942940)

This paper adapts the celebrated crystal bases introduced by the third named author to the case of the Lie superalgebra \({\mathfrak {gl}}(m,n)\) and its quantized enveloping algebra. If \({\mathfrak g}\) is a symmetrizable Kac-Moody Lie algebra with quantized enveloping algebra \(U_q({\mathfrak g})\), each integrable \(U_q({\mathfrak g})\)-module \(M\) has a base at \(q=0\) called the crystal base. More precisely, if \(A\) denotes the ring of rational functions with no pole at 0, then \(M\) contains an \(A\) lattice \(L\) called the crystal lattice, and the crystal base \(B\) is a certain basis for the \(\mathbb{Q}\)-vector space \(L/qL\). The basis is acted on by the so-called Kashiwara operators, and it has many important applications in combinatorial representation theory and other areas. The general Lie superalgebra \({\mathfrak {gl}} (m,n)\) arises from a \(\mathbb{Z}_2\)-graded vector space \(V= V_0\oplus V_1\) for which \(V_0\) and \(V_1\) have dimensions \(m\) and \(n\) respectively. Let \(\text{End} (V)_a\) be the set of endomorphisms \(x\) of \(V\) such that \(xV_b \subseteq V_{a+b}\) for \(a,b\in \{0,1\}\), where addition is taken modulo 2. Then \({\mathfrak {gl}} (m,n)\) is \(\text{End}(V)= \text{End}(V_0) \oplus \text{End}(V_1)\) regarded as a Lie algebra under the supercommentator product \([x,y]:= xy- (-1)^{ab} yx\), where \(x\in \text{End} (V)_a\) and \(y\in \text{End} (V)_b\). The representation \(V\) plays a key role in the theory, and the corresponding quantized enveloping algebra \(U_q ({\mathfrak {gl}} (m,n))\) (a Hopf superalgebra) has a corresponding representation \({\mathcal V}\). Tensor powers of \({\mathcal V}\) are completely reducible and the irreducible summands are indexed by partitions having an \((m,n)\)-block shape (meaning that there is no box at the intersection of row \((m+1)\) and column \((n+1)\)). The semistandard tableaux associated to such shapes are used to describe combinatorially the crystal structure of these irreducible modules. The generalization of the crystal basis machinery to the superalgebra case is far from routine, and the approach here differs from that of Y. M. Zou. One of the main difficulties is that representations for Lie superalgebras need not be completely reducible, so the authors restrict their study to certain representations that behave well under tensor products. Highest and lowest weight vectors play a key part in the theory of crystal bases, and another problem to be overcome in this particular case arises from the presence of what the authors call ``fake'' highest and lowest weight vectors.