"quantum" symmetric plane partitions beget alternating sign matrices?

Question

The "quantum" version qTSPP of the number of totally symmetric plane partitions, contained in the cube $[0,n]^3$, is enumerated by $$f_n(q):=\prod_{j=1}^n\prod_{k=1}^j\prod_{\ell=1}^k\frac{1-q^{j+k+\ell-1}}{1-q^{j+k+\ell-2}}.$$ L'Hopital $f_n(1)=\lim_{q\rightarrow1}f_n(q)$ restores the classical version $\prod_{1\leq\ell\leq k\leq j\leq n}\frac{j+k+\ell-1}{j+k+\ell-2}$. Although $f_n(-1)=0$ trivially, when $n$ is odd, I observe the case $n$ even is decidedly striking; namely that, $$f_{2n}(-1)=\lim_{q\rightarrow -1}f_{2n}(q)=\prod_{k=0}^{n-1}\frac{(3k+1)!}{(n+k)!},$$ the number $A_n$ of $n\times n$ Alternating Sign Matrices or $ASMs$.

QUESTION still waiting for an answer.

Is there a non-analytic (more conceptual) reason for this connection between qTSPP and ASMs?

Answer 1

This is not really an answer to the question per se (in that it is not conceptual), but I wanted to point out that this $q=-1$ evaluation was essentially already observed in Stembridge's original paper about the $q=-1$ phenomenon (as hinted at by the comment of Richard Stanley). Indeed, the point of that paper is that if $G$ is a subgroup of $S_3$ acting on plane partitions, and $P'_G(q)$ is the "appropriate" $q$-analog generating function for the plane partitions invariant under $G$, then the number of plane partitions invariant under $G$ and also self-complementary is $P'_G(-1)$. A particular case is $G=S_3$, which then says that the evaluation of the $q$-analog of the number of totally symmetric plane partitions at $q=-1$ is the number of totally symmetric, self-complementary plane partitions. You ask about alternating sign matrices rather than TSSCPPs, but of course the number of ASMs and TSSCPPs is the same.