We also construct RiemannHilbert homomorphisms from the cyclotomic Hecke algebras of certain 2dimensional complex reflection groups to the “spherical” subalgebras of the deformed central extensions of preprojective algebras. This allows us to show that if all parameters of the cyclotomic Hecke algebra are equal to 1 (i.e. the generators are unipotent) then the block of the trivial representation is equivalent to the category of representations of the spherical subalgebra of the centrally extended preprojective algebra.
As a byproduct, we show that the dimension of the cyclotomic Hecke algebra of a 2dimensional complex reflection group for generic parameters is equal to the order of the group, as conjectured by Broue, Malle, and Rouquier. The organization of the paper is as follows. In Section 2 we recall basic facts about deformed preprojective algebras, Frobenius algebras, and CohenMacaulay and Gorenstein algebras. In Section 3 we state the main results regarding central extensions of preprojective algebras and their deformations, as well as the corresponding spherical subalgebras. In Section 4 we set up the machinery of quantum
$S{L}_{2}$
and quantum Heisenberg algebra, which we use in Section 5 to prove the results of Section 2. Finally, in Section 6 we introduce and study the RiemannHilbert homomorphism.
Acknowledgments. P.E. is grateful to W. CrawleyBoevey for the references [
Ru]
,[
CB2]
, and to M. Artin, J. de Jong, V. Ostrik, D. Rogalski, and R. Rouquier for useful discussions. The work of P.E. was partially supported by the NSF grant DMS9988796 and the CRDF grant RM12545MO03. E.R. was supported in part by NSF Grant No. DMS0401387.
2 Preliminaries
2.1 Quivers and deformed preprojective algebras
Let
$Q$
be a quiver of finite Dynkin (that is, ADE) type with Cartan matrix
$A$
, and Coxeter number
$h$
. Let
$I=\{1,...,r\}$
be the vertex set of
$Q$
, and
$R={\oplus}_{i\in I}\mathbb{C}{e}_{i}$
be the commutative algebra generated by idempotents
${e}_{i}$
with
${e}_{i}{e}_{j}=0$
,
$i\ne j$
. Thus any element
$\mu \in R$
can be written as
$\mu ={\sum}_{i\in I}{\mu}_{i}{e}_{i}$
,
${\mu}_{i}\in \mathbb{C}$
.
Let
${x}_{i}\in {R}^{*}$
be the dual basis to
${e}_{i}$
. It is well known that
${x}_{i}$
span a root system
$\Delta $
inside
${R}^{*}$
with Cartan matrix
$A$
. We will identify
$R$
and
${R}^{*}$
using the Weyl group invariant inner product such that
$({x}_{i},{x}_{i})=2$
. Then
${e}_{i}$
are the fundamental weights of
$\Delta $
.
Let
$\rho \in R$
be the sum of fundamental weights. Of course,
$\rho =1$
, but we'll use the notation
$\rho $
to emphasize the connection with Lie theory.
Let
$\overline{Q}=Q\cup {Q}^{*}$
be the double of
$Q$
, – the quiver with the same vertex set and an additional opposite edge
${a}^{*}\in {Q}^{*}$
for every edge
$a\in Q$
.
The generic deformed preprojective algebra
$\Pi $
of
$Q$
([
CBH]
,[
Ru]
) is the quotient of the path algebra of
$\overline{Q}$
over the ring
$\mathbb{C}[{x}_{1},...,{x}_{r}]=S{R}^{*}$
by the relation
$${\sum}_{a\in Q}[a,{a}^{*}]={\sum}_{i\in I}{x}_{i}{e}_{i}$$
(here
$\sum {x}_{i}{e}_{i}$
is the canonical element of
${R}^{*}\otimes R$
). This algebra is graded (
$\text{deg}\left({e}_{i}\right)=0$
,
$\text{deg}\left({x}_{i}\right)=2$
,
$\text{deg}\left(a\right)=\text{deg}\left({a}^{*}\right)=1$
). Also, it is clear that
$\Pi $
is independent on the orientation of the graph Q, up to an isomorphism: the reversal of orientation of an edge
$a$
may be accomplished by replacing
$a$
by
$a$
.
Let
${\Pi}_{0}$
be the zero fiber of
$\Pi $
, i.e. the quotient of the path algebra of
$\mathbb{C}\overline{Q}$
of
$\overline{Q}$
by the relation
${\sum}_{a\in Q}[a,{a}^{*}]=0$
. It is called the (GelfandPonomarev) preprojective algebra of
$Q$
. The fiber
${\Pi}_{\lambda}$
of
$\Pi $
at
$\lambda \in R$
is called the deformed preprojective algebra.
It is known that
${\Pi}_{0}$
is finite dimensional. Namely, as pointed out already by Gelfand and Ponomarev,
${\Pi}_{0}$
is a model for indecomposable representations of
$Q$
(i.e., the direct sum of all of them taken once). Thus the dimension of
${\Pi}_{0}$
is the sum of heights of all positive roots:
$$dim\left({\Pi}_{0}\right)={\sum}_{\alpha >0}(\alpha ,\rho )=2(\rho ,\rho )=\frac{h(h+1)r}{6}.$$
(see [
MOV]
). The last equality follows from Freudental's magic formula (
$(\rho ,\rho )=hdim\mathfrak{g}/12$
, where
$\mathfrak{g}$
is the simple Lie algebra attached to
$A$
) and Kostant's formula (
$(h+1)r=dim\mathfrak{g}$
). Moreover, it is known (see [
MOV]
) that the Hilbert polynomial of
${\Pi}_{0}$
with respect to its grading is
$$\begin{array}{c}{H}_{0}\left(t\right)=\frac{1+P{t}^{h}}{1Ct+{t}^{2}},\end{array}$$ 
(1)

where
$P$
is the permutation of fundamental weights corresponding to taking the dual representation of
$\mathfrak{g}$
.
This implies that
${\Pi}_{\lambda}$
is finite dimensional for each
$\lambda $
, and
$\Pi $
is a finitely generated
$\mathbb{C}[{x}_{1},...,{x}_{r}]$
module.
It is known [
CB1]
that unlike the nonDynkin case, the algebra
$\Pi $
is not a free
$\mathbb{C}[{x}_{1},...,{x}_{r}]$
module. More specifically, it is known (see e.g. [
CB1]
, Theorem 1.2) that the algebra
${\Pi}_{\lambda}$
is zero unless
$\lambda $
belongs to a reflection hyperplane.
2.2 Frobenius algebras
We recall the basic facts about Frobenius algebras.
Let
$A$
be a finite dimensional (unital) algebra over
$\mathbb{C}$
. Recall that it is called a Frobenius algebra if
$A={A}^{*}$
as a left
$A$
module. This is equivalent to saying that there exists a linear function
$f:A\to \mathbb{C}$
such that the bilinear for
$(a,b)=f\left(ab\right)$
is nondegenerate. Indeed, given
$A$
with an isomorphism of left modules
$\phi :A\to {A}^{*}$
, we set
$f=\phi \left(1\right)$
, and conversely, given
$f$
, we define
$\phi $
by
$\phi \left(a\right)\left(b\right)=f\left(ba\right)$
.
Lemma 2.1.
Let
$A$
be a
${\mathbb{Z}}_{+}$
graded finite dimensional algebra,
$A={\oplus}_{j\ge 0}A\left[j\right]$
such that
$R:=A\left[0\right]$
is a commutative semisimple algebra, and the Hilbert polynomial
${P}_{A}\left(t\right)$
satisfies the condition
${P}_{A}\left(t\right)={t}^{d}{P}_{A}\left({t}^{1}\right)$
. Then the following conditions are equivalent:
(i)
$A$
is Frobenius; (ii)
$A\left[d\right]$
is an invertible
$A\left[0\right]$
bimodule, and the multiplication map
$A\left[i\right]{\otimes}_{R}A[di]\to A\left[d\right]$
defines an isomorphism
$A[di]\to A\left[i{]}^{*}{\otimes}_{R}A\right[d]$
.

Proof.
Let
$R={\oplus}_{j=1}^{r}{R}_{j}$
, where
${R}_{j}$
are copies of
$\mathbb{C}$
. Then any
$R$
bimodule
$M$
can be written as
$M=\oplus {M}_{ij}\otimes {R}_{ij}$
, where
${R}_{ij}=\mathbb{C}$
is the unique irreducible
$({R}_{i},{R}_{j})$
bimodule.
Suppose (ii) holds. Since
$A\left[d\right]$
is invertible, we have
$A\left[d\right]={\oplus}_{q}{R}_{\sigma \left(q\right)q}$
for some permutation
$\sigma $
. Then the multiplication map defines an isomorphism
$$A[di{]}_{pq}\to (A[i{]}^{*}{)}_{p\sigma \left(q\right)}.$$
This implies that if
$f\in {A}^{*}$
vanishes on degrees
$<d$
, and
$f:A\left[d\right]\to \mathbb{C}$
is given by
$f({y}_{1},...,{y}_{r})={\sum}_{j=1}^{r}{y}_{j}$
(
${y}_{p}\in {R}_{p\sigma \left(p\right)}$
), then
$f\left(ab\right)$
is a nondegenerate form on
$A$
. Thus (i) holds.
Now suppose (i) holds. Then the map
$(a,b)\mapsto f\left(ab\right)$
defines pairings
$A\left[0\right]\otimes A\left[d\right]\to \mathbb{C}$
and
$A\left[d\right]\otimes A\left[0\right]\to \mathbb{C}$
which have trivial kernel in
$A\left[d\right]$
. Since the dimensions of
$A\left[0\right]$
and
$A\left[d\right]$
are the same, these pairings are nondegenerate.
This implies that
$A\left[d\right]$
is faithful as a left and right
$R$
module, which by dimension counting implies that
$A\left[d\right]$
is invertible, i.e.
$A\left[d\right]=\oplus {R}_{\sigma \left(q\right)q}$
. Now the multiplication map defines a linear map
${\psi}_{ipq}:A[di{]}_{pq}\to (A[i{]}^{*}{)}_{p\sigma \left(q\right)}$
.
Assume that (ii) fails. Then by dimension count for some
$i,p,q$
the map
${\psi}_{ipq}$
has a nontrivial kernel. Let us choose the largest such
$i$
(clearly
$i<d$
), and let
$a$
be an element in the kernel of
${\psi}_{ipq}$
. It is clear that
$A\left[m\right]a=0$
for all
$m>0$
, because for any
$b\in A[m{]}_{sp}$
,
$b\in Ker{\psi}_{i+m,sq}$
. Let us pick
${a}^{\prime}\in A\left[d\right]$
such that
$f\left(ca\right)=f\left(c{a}^{\prime}\right)$
for any
$c\in A\left[0\right]$
. Then
$f\left(c\right(a{a}^{\prime}\left)\right)=0$
for all
$c\in A$
. Contradiction. Thus (ii) must hold, as desired. □
2.3 CohenMacaulay and Gorenstein algebras
Let us now recall the basic properties of noncommutative CohenMacaulay and Gorenstein algebras. (see also [
YZ]
or the discussion in [
EG]
, beginning of Section 3).
Let
$A$
be a
${\mathbb{Z}}_{+}$
graded algebra over
$\mathbb{C}$
, such
$R=A\left[0\right]$
is finite dimensional and semisimple. We will assume that
$A$
is a finitely generated module over a finitely generated graded central subalgebra
$B$
, such that
$B\left[0\right]=\mathbb{C}$
. In this case, by Noether's normalization lemma, we may assume that
$B$
is a polynomial algebra; we will do so from now on.
In this situation, the dualizing complex of
$A$
may be defined by the formula
${\mathbb{D}}_{A}=RHo{m}_{B}(A,B)$
. It can be shown that it does not depend on the choice of
$B$
.
Definition 2.2.
The algebra
$A$
is said to be CohenMacaulay if the cohomology of its dualizing complex is concentrated in degree zero, and Gorenstein if this cohomology is furthermore isomorphic to
$A$
as a left module. ^{2 }
Thus we have the following proposition.
Proposition 2.3.
(i)
$A$
is CohenMacaulay if and only if
$A$
is a free
$B$
module.
(ii)
$A$
is Gorenstein if and only if furthermore there exists a
$B$
linear map
$f:A\to B$
such that the map
$(a,b)\mapsto f\left(ab\right)$
is a nondegenerate pairing
$A\times A\to B$
, i.e. defines an isomorphism
$A\to Ho{m}_{B}(A,B)$
.
This immediately implies the following.
Proposition 2.4.
(i) Let
$A$
be a finite dimensional algebra. Then
$A$
is Gorenstein if and only if it is Frobenius. (ii)
$A$
is Gorenstein if and only if its zerofiber
${A}_{0}$
(as a
$B$
module) is Gorenstein.
3 Central extensions of preprojective algebras
3.1 Gorenstein properties and Hilbert series of generic deformed preprojective algebras
One of the main results of this paper is the following theorem.
Theorem 3.1.
(i)
$\Pi $
is a Gorenstein algebra (in particular, CohenMacaulay). (ii) The matrix Hilbert series of
$\Pi $
(i.e. the matrix consisting of the Hilbert series of
${e}_{i}\Pi {e}_{j}$
) is equal to
$$H\left(t\right)=\frac{1{t}^{2h}}{(1{t}^{2}{)}^{r}(1Ct+{t}^{2})},$$
where
$C=2A$
is the adjacency matrix of
$\overline{Q}$
.
Theorem 3.1 is proved in Section 5 .
3.2 Central extensions of preprojective algebras
Another result of this paper is a construction of a central extension of the preprojective algebra, whose deformed version is flat, unlike that of the usual preprojective algebra. Namely, let
$\Pi \left[z\right]$
be the algebra of polynomials of a central variable
$z$
with coefficients in
$\Pi $
; it is graded with
$\text{deg}\left(z\right)=2$
. Let
$\mu \in R$
be a regular weight (i.e., does not belong to any reflection hyperplane
$(\alpha ,\mu )=0$
). For any weight
$\lambda \in R$
, let
${\Pi}_{\lambda}^{\mu}$
be the quotient of
$\Pi \left[z\right]$
by the relations
${x}_{i}={\mu}_{i}z+{\lambda}_{i}$
. Thus
${\Pi}_{\lambda}^{\mu}$
is the quotient of the path algebra
$\mathbb{C}\left[z\right]\overline{Q}$
by the relation
$${\sum}_{a}[a,{a}^{*}]={\sum}_{i}({\mu}_{i}z+{\lambda}_{i}){e}_{i}.$$
This algebra carries a natural filtration induced by the grading in the path algebra.
Theorem 3.2.
(i)
$\Pi \left[z\right]$
is a finitely generated free module over the algebra
${K}_{\mu}:=\mathbb{C}[{x}_{1}{\mu}_{1}z,...,{x}_{r}{\mu}_{r}z]$
.
(ii) The associated graded algebra of
${\Pi}_{\lambda}^{\mu}$
under the natural filtration is
${\Pi}_{0}^{\mu}$
.
(iii) The algebra
${\Pi}_{0}^{\mu}$
is Gorenstein (=Frobenius). One has
${z}^{h1}=0$
, and for generic
$\mu $
the socle of
${\Pi}_{0}^{\mu}$
is
${z}^{h2}R$
.
(iv) The Hilbert polynomial of
${\Pi}_{0}^{\mu}$
is
$$\stackrel{~}{H}\left(t\right)=H\left(t\right)(1{t}^{2}{)}^{r1}=\frac{1+{t}^{2}+...+{t}^{2(h1)}}{1Ct+{t}^{2}}.$$
(v) The dimension of
${\Pi}_{\lambda}^{\mu}$
is
${h}^{2}(h+1)r/12$
, and for generic
$\lambda $
it is semisimple.
Theorem 3.2 is proved in Section 5 .
Part (v) of Theorem
3.2 implies that for generic
$\lambda $
the element
$z$
is semisimple, and hence
$${\Pi}_{\lambda}^{\mu}={\oplus}_{\alpha >0}{\Pi}_{\lambda \frac{(\lambda ,\alpha )}{(\mu ,\alpha )}\mu}.$$
Thus
${\Pi}_{\lambda}$
is semisimple for a generic
$\lambda $
on a reflection hyperplane
$(\alpha ,\lambda )=0$
(where
$\alpha $
is a positive root). Moreover, by [
CB1]
, Theorem 1.2, it is actually simple:
${\Pi}_{\lambda}=Ma{t}_{(\alpha ,\rho )}(\mathbb{C})$
. This means that the Weyl denominator
$\delta \left(x\right)\in \mathbb{C}[{x}_{1},...,{x}_{r}]$
is zero in
$\Pi $
, so
$\Pi $
is schemetheoretically supported on the reflection hyperplanes. This was conjectured by Rump [
Ru]
and proved by CrawleyBoevey in [
CB2]
using a different method.
3.3 Subalgebras corresponding to nodal vertices
It is interesting to consider “spherical subalgebras” of the above algebras corresponding to nodal vertices.
Namely, let
$Q$
be of type
${A}_{2n1}$
,
$D$
or
$E$
, and
$p\in I$
be the nodal vertex (= the branching vertex for
$D$
and
$E$
, and the middle vertex for
${A}_{2n1}$
). After removal of
$p$
, the quiver becomes a union of “legs”, i.e. Dynkin diagrams of type
${A}_{{d}_{k}1}$
,
$k=1,...,m$
(
$m=2$
or
$3$
). Recall that to such
$Q$
one can attach a finite subgroup
$G$
of
$SO\left(3\right)$
generated by elements
${T}_{k}$
,
$k=1,...,m$
, with defining relations
${T}_{k}^{{d}_{k}}=1$
for
$k=1,...,m$
, and
${\prod}_{k=1}^{m}{T}_{k}=1$
(for type
$A$
we get the cyclic group, for type
$D$
the dihedral group, for type
${E}_{6}$
the tetrahedral group, for type
${E}_{7}$
the cube group, and for type
${E}_{8}$
the icosahedral group); all finite subgroups of
$SO\left(3\right)$
are obtained in this way.
The order
$\leftG\right$
of the group
$G$
can be represented as a product of two integers
${q}_{1}\le {q}_{2}$
such that
${q}_{1}+{q}_{2}1=h/2$
. Namely (see e.g. [
Ko]
),
${q}_{1}$
is the pth coordinate of the maximal root in the basis of simple roots. ^{3 }
Let
$B={e}_{p}\Pi {e}_{p}$
. According to [
MOV]
, the algebra
$B$
is generated over
$\mathbb{C}[{x}_{1},...,{x}_{r}]$
by
${U}_{k}$
,
$k=1,...,m$
, with defining relations
$${U}_{k}({U}_{k}{x}_{{i}_{1}\left(k\right)})({U}_{k}{x}_{{i}_{1}\left(k\right)}...{x}_{{i}_{{d}_{k}1}\left(k\right)})=0,$$
where
${i}_{1}\left(k\right),...,{i}_{{d}_{k}1}\left(k\right)$
are the vertices of the
$k$
th leg of
$Q$
enumerated from the nodal vertex, and
$${\sum}_{k=1}^{m}{U}_{k}={x}_{p}.$$
Namely, the elements
${U}_{k}$
are just the elements
${a}_{k}^{*}{a}_{k}$
, where
${a}_{k}$
are the edges of
$\overline{Q}$
starting at
$p$
and going along the
$k$
th leg.
This implies that the algebra
${B}_{\lambda}^{\mu}:={e}_{p}{\Pi}_{\lambda}^{\mu}{e}_{p}$
is generated over
$\mathbb{C}\left[z\right]$
by
${U}_{k}$
with the same defining relations, in which
${x}_{i}={\lambda}_{i}+{\mu}_{i}z$
. ^{4 }
Let
$\mu $
be a weight such that
$(\alpha ,\mu )\ne 0$
for any positive root
$\alpha $
that involves
${x}_{p}$
with a strictly positive coefficient. Also, for a number
$n$
, let
$[n{]}_{q}:=\frac{1{q}^{n}}{1q}$
.
For brevity, if
$q={t}^{2}$
, we will simply write
$\left[n\right]$
instead of
$[n{]}_{q}$
.
Theorem 3.3.
(i) The algebra
$B\left[z\right]$
is a finitely generated free module over the algebra
${K}_{\mu}$
.
(ii) The associated graded algebra of
${B}_{\lambda}^{\mu}$
under the natural filtration is
${B}_{0}^{\mu}$
.
(iii) The algebra
${B}_{0}^{\mu}$
is Gorenstein. We have
${z}^{h1}=0$
, and for generic
$\mu $
the socle of
${B}_{0}^{\mu}$
is spanned by
${z}^{h2}$
. (iv) The Hilbert polynomial of
${B}_{0}^{\mu}$
is
$E\left(t\right)=[h/2]\left[{q}_{1}\right]\left[{q}_{2}\right]$
.
(v) The dimension of
${B}_{\lambda}^{\mu}$
is
$h{q}_{1}{q}_{2}/2$
, and for generic
$\lambda $
it is semisimple.
Theorem 3.3 is proved in Section 5 .
Remark. If
$\mu $
is regular, then this theorem easily follows from Theorem 3.2 , but our statement is more general, so it requires a separate proof.
Considering the case
$\mu ={e}_{p}$
, we immediately get the following corollary.
For any numbers
${\lambda}_{ik}$
,
$i=1,...,{d}_{k}$
,
$k=1,...,m$
, define the algebra
$B\left(\lambda \right)$
generated by
${U}_{k}$
and a central element
$z$
with defining relations
$${}^{{d}_{k}}{\prod}_{i=1}({U}_{k}{\lambda}_{i})=0,$$
$${\sum}_{k=1}^{m}{U}_{k}=z.$$
Corollary 3.4.
(i)
$grB\left(\lambda \right)=B\left(0\right)$
.
(ii)
$B\left(0\right)$
is a Gorenstein algebra, and
${z}^{h1}=0$
in
$B\left(0\right)$
.
(iii) The Hilbert polynomial of
$B\left(0\right)$
is
$[h/2]\left[{q}_{1}\right]\left[{q}_{2}\right]$
.
For comparison note that the Hilbert polynomial of the algebra
${B}_{0}:=B\left(0\right)/\left(z\right)$
, according to [MOV] , equals
$\left[{q}_{1}\right]\left[{q}_{2}\right]$
.
Remark. We have checked using the Magma computer algebra system that the socle of
$B\left(0\right)$
is spanned by
${z}^{h2}$
.
4 The quantum Heisenberg algebra
4.1 Definition on properties of the quantum Heisenberg algebra
The proofs of the main results of this paper are based on the idea of [
MOV]
:
algebras related to quivers may be obtained from algebras in the category of representations of quantum
$SL\left(2\right)$
by application of tensor functors into
$R$
bimodules.
We are going to use the notions and notation from [
BaKi,
EO]
. Namely, for every
$q\in {\mathbb{C}}^{*}$
, denote by
${\stackrel{~}{\mathcal{C}}}_{q}$
the tensor category of finite dimensional comodules over the quantum function algebra
${F}_{q}\left(SL\right(2\left)\right)$
. If
$q$
is not a root of unity of order
$>2$
, then this category is semisimple and has simple objects
${V}_{i}$
,
$i\in {\mathbb{Z}}_{+}$
(representations with highest weight
$i$
) such that
${V}_{0}$
is the neutral object, with the ClebschGordan tensor product rule
$${V}_{i}\otimes {V}_{j}{=}^{min(i,j)}{\oplus}_{n=0}{V}_{2n+ij}.$$
On the other hand, if
$q$
is a root of unity of order
$n>2$
, then the category
${\stackrel{~}{\mathcal{C}}}_{q}$
is not semisimple. However, the objects
${V}_{i}$
are still well defined (the so called Weyl modules), and they are simple if
$i\le \overline{n}1$
, where
$\overline{n}=n$
if
$n$
is odd and
$\overline{n}=n/2$
if
$n$
is even.
Consider the tensor algebra
$T({V}_{1}\oplus {V}_{0})$
in
${\stackrel{~}{\mathcal{C}}}_{q}$
(note that it is an indobject).
We will regard it as a graded algebra in which
${V}_{1}$
has degree
$1$
and
${V}_{0}$
has degree 2. Then we have two morphisms from
${V}_{0}$
to
$T({V}_{1}\oplus {V}_{0})\left[2\right]$
: the map
$f$
which is a composition
${V}_{0}\to {V}_{1}\otimes {V}_{1}\to T({V}_{1}\oplus {V}_{0})\left[2\right]$
, and the map
$g:{V}_{0}\to T({V}_{1}\otimes {V}_{0})\left[2\right]$
coming from the embedding
${V}_{0}\to {V}_{1}\oplus {V}_{0}$
.
There are also two obvious embeddings
${f}_{1},{f}_{2}:{V}_{0}\otimes {V}_{1}\to T({V}_{0}\oplus {V}_{1})\left[3\right]$
corresponding to multiplication in two different orders (more precisely,
${f}_{1}$
is the multiplication map, and
${f}_{2}$
is the multiplication map composed with the canonical isomorphism
${V}_{0}\otimes {V}_{1}\to {V}_{1}\to {V}_{1}\otimes {V}_{0}$
). Let
$\stackrel{~}{\mathcal{J}}$
be the ideal generated by the images of
$fg$
and
${f}_{1}{f}_{2}$
.
Definition 4.1.
The quantum Heisenberg algebra in
${\stackrel{~}{\mathcal{C}}}_{q}$
is the algebra
$\stackrel{~}{A}=T({V}_{0}\oplus {V}_{1})/\stackrel{~}{\mathcal{J}}$
.
More explicitly, the algebra
$\stackrel{~}{A}$
is the usual quantum Heisenberg algebra generated by
$x,y,z$
with defining relations saying that the element
$z=xyqyx$
is central, i.e.
$$xyqyx=z,xzzx=0,yzzy=0,$$
and the usual coaction of the Hopf algebra
${F}_{q}\left(SL\right(2\left)\right)$
; it is convenient to express this coaction as an action of
${U}_{q}\left(sl\right(2\left)\right)$
given by
$$ex=y,ey=0,fy=x,fx=0,{q}^{h}y=qy,{q}^{h}x={q}^{1}x,$$
$$ez=fz=0,{q}^{h}z=z.$$
Here the coproduct of
${U}_{q}\left(sl\right(2\left)\right)$
is given by
$\Delta \left(e\right)=e\otimes {q}^{h}+1\otimes e$
,
$\Delta \left(f\right)=f\otimes 1+{q}^{h}\otimes f$
,
$\Delta \left({q}^{h}\right)={q}^{h}\otimes {q}^{h}$
.
Lemma 4.2.
Assume that
$q$
has order infinity or an even number
$2l$
. Let
$j<l$
, and
${E}_{j,s}$
be the copy of
${V}_{j2s}$
in
$\stackrel{~}{A}\left[j\right]$
. Then the multiplication map
${E}_{j,0}\otimes {E}_{1,0}\to {E}_{j+1,1}$
is nonzero.

Proof.
It is easy to see that^{5 }
$$\begin{array}{c}x{y}^{j}={q}^{j}{y}^{j}x+[j{]}_{q}{y}^{j1}z.\end{array}$$ 
(2)

Now, the highest weight vector of
${E}_{j,0}$
is
${y}^{j}$
. Thus it suffices to show that the vector
${y}^{j}x$
, which lies in
${E}_{j+1,0}\oplus {E}_{j+1,1}$
, has a nonzero component in
${E}_{j+1,1}$
, i.e. that it is not an eigenvector of the operator
$fe$
. But we have
$$fe\left({y}^{j}x\right)=f\left({y}^{j+1}\right)={\sum}_{s=0}^{j}{q}^{s}{y}^{s}x{y}^{js}.$$
Using equation ( 2 ), we get
$$fe\left({y}^{j}x\right)={\sum}_{s=0}^{j}{q}^{s}{y}^{s}({q}^{js}{y}^{js}x+[js{]}_{q}{y}^{js1}z)=$$
$$\frac{{q}^{j+1}{q}^{j1}}{q{q}^{1}}{y}^{j}x+\frac{\left[j{]}_{q}\right[j+1{]}_{{q}^{1}}}{[2{]}_{{q}^{1}}}{y}^{j1}z.$$
Since the second coefficient is never zero, we get the required statement. □
Now let
$q={e}^{\pi i/h}$
, where
$h\ge 2$
is an integer. Then the category
${\stackrel{~}{\mathcal{C}}}_{q}$
contains a nonabelian subcategory
${\mathcal{T}}_{q}$
of tilting modules, which is closed under taking direct summands. Furthermore, the category
${\mathcal{T}}_{q}$
contains a tensor ideal
$\mathcal{\mathcal{I}}$
, such that
${\mathcal{T}}_{q}/\mathcal{\mathcal{I}}={\mathcal{C}}_{q}$
is a semisimple tensor category (the fusion category). The simple objects in
${\mathcal{C}}_{q}$
are
${V}_{0},...,{V}_{h2}$
, and the tensor product is given by the Verlinde rule:
$${V}_{i}\otimes {V}_{j}{=}^{min(i,j,h2max(i,j\left)\right)}{\oplus}_{n=0}{V}_{2n+ij}.$$
Consider now the tensor algebra
$T({V}_{0}\oplus {V}_{1})$
in the fusion category
${\mathcal{C}}_{q}$
.
Then we can define the ideal
$\mathcal{J}$
in this algebra generated by the images of the maps
$fg$
and
${f}_{1}{f}_{2}$
, where
${f}_{1},{f}_{2},f,g$
are defined in the same way as above.
Definition 4.3.
The quantum Heisenberg algebra in
${\mathcal{C}}_{q}$
is the algebra
$A=T({V}_{0}\oplus {V}_{1})/\mathcal{J}$
.
Remark. The quotient
${A}_{0}$
of
$A$
by the ideal generated by the unique copy of
${V}_{0}$
in degree
$2$
is the quantum symmetric algebra considered in [MOV] ; as an object of
${\mathcal{C}}_{q}$
it is
${V}_{0}\oplus {V}_{1}\oplus ...\oplus {V}_{h2}$
, and the degree of
${V}_{i}$
is
$i$
.
Proposition 4.4.
(i) The degree
$n$
component
$A\left[n\right]$
of
$A$
is the object
${\oplus}_{j\le s/2}{V}_{s2j}$
, where
$s=min(n,2h4n)$
.
(ii) The algebra
$A$
is Gorenstein in the following sense: the highest nontrivial degree in
$A$
is
$A[2h4]={V}_{0}$
, and the multiplication map
$A\left[i\right]\otimes A[2h4i]\to A[2h4]$
defines an isomorphism
$A\left[i\right]\to A[2h4i{]}^{*}$
.

Proof.
Let
$L$
be the ideal in
$\stackrel{~}{A}$
generated by the copy of
${V}_{h1}\subset {V}_{1}^{\otimes h1}\subset T({V}_{0}\oplus {V}_{1})[h1]$
. It follows from Lemma 4.2 that the graded algebra
$\stackrel{~}{A}/L$
, as an object of the category
${\stackrel{~}{\mathcal{C}}}_{q}$
, has the structure specified in part (i) of the proposition. In particular,
$\stackrel{~}{A}/L$
belongs to the category of tilting modules
${\mathcal{T}}_{q}$
. In the category
${\mathcal{T}}_{q}$
, the algebra
$\stackrel{~}{A}/L$
can be written as
$T({V}_{0}\oplus {V}_{1})/\stackrel{~}{L}$
, where
$\stackrel{~}{L}$
is the preimage of
$L$
in
$T({V}_{0}\oplus {V}_{1})$
. Note that both the algebra
$T({V}_{0}\otimes {V}_{1})$
and the ideal
$\stackrel{~}{L}$
belong to
${\mathcal{T}}_{q}$
.
Now we can consider the image
${A}^{\prime}$
of
$\stackrel{~}{A}/L$
in the fusion category
${\mathcal{C}}_{q}$
(which is by definition a quotient category of
${\mathcal{T}}_{q}$
). We have
${A}^{\prime}=T({V}_{0}\oplus {V}_{1})/{J}^{\prime}$
, where
${J}^{\prime}$
is the image of
$\stackrel{~}{L}$
. Since the image of
${V}_{h1}$
in
${\mathcal{C}}_{q}$
is zero, we have
${J}^{\prime}=J$
and
${A}^{\prime}=A$
, hence (i).
To prove part (ii), it suffices to note that by Lemma
4.2 , if
${V}_{j}\subset A\left[i\right]$
,
$i\ge j>0$
then the map
${V}_{j}\otimes {V}_{1}\to {V}_{j1}$
defined by multiplication by the generating copy of
${V}_{1}$
and then projecting to
${V}_{j1}\subset A[i+1]$
is nonzero.
More precisely, this statement follows from the lemma by multiplying by
${z}^{(ij)/2}$
. □
4.2 The quantum Heisenberg algebra and the central extension of the preprojective algebra
Now assume that
$q={e}^{\pi i/h}$
, where
$h$
is the Coxeter number of
$Q$
. Recall (see [
EO,
MOV]
) that there exists a unique tensor functor
$\mathcal{\mathcal{F}}:{\mathcal{C}}_{q}\to Rbimod$
from
${\mathcal{C}}_{q}$
to the category of
$R$
bimodules, such that the bimodule
$\mathcal{\mathcal{F}}\left({V}_{1}\right)$
is the edge space
$E$
of the doubled quiver
$\overline{Q}$
. It is checked in [
MOV]
that
$\mathcal{\mathcal{F}}\left({A}_{0}\right)$
is the preprojective algebra
${\Pi}_{0}$
.
Proposition 4.5.
$\mathcal{\mathcal{F}}\left(A\right)={\Pi}_{0}^{\rho}$
.

Proof.
The algebra
${\Pi}_{0}^{\rho}$
is the quotient of the path algebra
$\mathbb{C}\left[z\right]\overline{Q}$
by the relations saying that
${\sum}_{a\in Q}[a,{a}^{*}]=z$
, and
$z$
is central. It is easy to show, similarly to [MOV] , that these relations are images under
$\mathcal{\mathcal{F}}$
of the relations
$xyqyx=z$
,
$xz=zx$
,
$yz=zy$
. The proposition is proved. □
Corollary 4.6.
The Hilbert series of
${\Pi}_{0}^{\rho}$
is given by the formula in Theorem 3.2 , (iv). In particular, the dimension of
${\Pi}_{0}^{\rho}$
is
${h}^{2}(h+1)r/12$
.

Proof.
In the Grothendieck ring of
${\mathcal{C}}_{q}$
, we have
${V}_{j}={P}_{j}\left({V}_{1}\right)$
, where
${P}_{j}$
is the Tchebysheff polynomial of the second kind:
${P}_{j}(2cosx)=\frac{sin(j+1)x}{sinx}$
. This means that by Proposition 4.4 , (i), the Grothendieckgroupvalued Hilbert polynomial of
$A$
is
$${\sum}_{j=0}^{h2}{\sum}_{i=0}^{h2j}{t}^{2i+j}{P}_{j}\left({V}_{1}\right),$$
and hence by Proposition 4.5 , the Hilbert polynomial of
${\Pi}_{0}^{\rho}$
is
$$\stackrel{~}{H}\left(t\right)={\sum}_{j=0}^{h2}{\sum}_{i=0}^{h2j}{t}^{2i+j}{P}_{j}\left(C\right)={\sum}_{j=0}^{h2}\frac{{t}^{j}{t}^{2(h1)j}}{1{t}^{2}}{P}_{j}\left(C\right)=\frac{{H}_{0}\left(t\right){t}^{2h2}{H}_{0}\left({t}^{1}\right)}{1{t}^{2}}.$$
Thus by formula ( 1 ),
$$\stackrel{~}{H}\left(t\right)=\frac{1{t}^{2h}}{(1{t}^{2})(1Ct+{t}^{2})},$$
as desired. Thus we have
$\stackrel{~}{H}\left(t\right)=\frac{1P{t}^{h}}{1{t}^{2}}{H}_{0}\left(t\right)$
, and hence
$dim{\Pi}_{0}^{\rho}=hdim{\Pi}_{0}/2={h}^{2}(h+1)r/12$
. The corollary is proved. □
Remark. Since
$F\left({A}_{0}\right)={\Pi}_{0}$
, we conclude from Proposition 4.4 that the algebra
${\Pi}_{0}$
, and hence
${B}_{0}$
, are Frobenius algebras.
5 Proofs of Theorems 3.1 , 3.2 , 3.3
5.1 Proofs of Theorems 3.1 , 3.2
We now prove Theorems 3.1 , 3.2 .
Lemma 5.1.
$\Pi \left[z\right]$
is a free module over
$\mathbb{C}[{x}_{1}z,...,{x}_{r}z]$
.

Proof.
The fiber of this module at
${x}_{i}={\lambda}_{i}$
is
${\Pi}_{\lambda}^{\rho}$
. Thus it suffices to show that the dimension of
${\Pi}_{\lambda}^{\rho}$
is at least as big as that of
${\Pi}_{0}^{\rho}$
, for generic
$\lambda $
. To do so, note that by Theorem 1.2 in [CB1] , for generic
$\lambda $
the algebra
${\Pi}_{\lambda}^{\rho}$
has an irreducible representation with dimension vector being each positive root
$\alpha $
. Therefore,
$$dim{\Pi}_{\lambda}^{\rho}\ge {\sum}_{\alpha >0}(\alpha ,\rho {)}^{2}=h(\rho ,\rho )={h}^{2}\frac{dim\mathfrak{g}}{12}=\frac{{h}^{2}(h+1)r}{12},$$
which together with Corollary 4.6 implies the desired inequality. □
Lemma 5.1 together with Corollary 4.6 imply part (ii) of Theorem 3.1 .
Now, by Proposition
4.4 , Proposition 4.5 , and Proposition 2.1 ,
${\Pi}_{0}^{\rho}$
is a Frobenius algebra. Hence by Proposition 2.4 (i) it is a Gorenstein algebra.
Therefore, since
$\Pi \left[z\right]$
is a free module over
$\mathbb{C}[{x}_{1}z,...,{x}_{r}z]$
, by Proposition 2.4 (ii), the algebra
$\Pi \left[z\right]$
is Gorenstein, and hence again by Proposition 2.4 (ii),
$\Pi $
is Gorenstein. This implies part (i) of Theorem 3.1 . Thus Theorem 3.1 is proved.
Now let
$\mu $
be a regular weight. Since
${\Pi}_{0}$
is finite dimensional, the algebra
${\Pi}_{0}^{\mu}$
is finitely generated as a module over
$\mathbb{C}\left[z\right]$
. Moreover, if
${z}_{0}\in \mathbb{C}$
, we have
${\Pi}_{0}^{\mu}/(z{z}_{0})={\Pi}_{{z}_{0}\mu}$
, which is zero by Theorem 1.2 of [
CB1]
for any
${z}_{0}\ne 0$
. Hence
${\Pi}_{0}^{\mu}$
is finite dimensional. This implies that
$\Pi \left[z\right]$
is a finitely generated
${K}_{\mu}$
module (since this module is
${\mathbb{Z}}_{+}$
graded and its zerofiber is
${\Pi}_{0}^{\mu}$
). Since
$\Pi \left[z\right]$
is a Gorenstein algebra, it is in particular CohenMacaulay, and by Proposition 2.3 (i),
$\Pi \left[z\right]$
is a free
${K}_{\mu}$
module, hence parts (i),(ii) of Theorem 3.2 .
We also see that the Hilbert polynomial of
${\Pi}_{0}^{\mu}$
is the same as that for
${\Pi}_{0}^{\rho}$
, so Corollary 4.6 implies Theorem 3.2 , (iv), and the dimension formula of (v). The semisimplicity of
${\Pi}_{\lambda}^{\mu}$
follows from the fact that for generic
$\lambda $
, by Theorem 1.2 of [
CB1]
, it has an irreducible representation with dimension vector being every positive root
$\alpha $
; as we've shown, the sum of dimensions of these representations is
${h}^{2}(h+1)r/12=dim{\Pi}_{0}^{\mu}$
.
We also see, by Proposition
2.4 (ii), that
${\Pi}_{0}^{\mu}$
is Gorenstein (=Frobenius).
This proves the first statement of Theorem
3.2 , (iii). The second statement follows from part (iv), and the third statement (the socle is
${z}^{h2}R$
) follows from the fact that this is so for
$\mu =\rho $
(by Proposition 4.4 and Proposition 4.5 ). Theorem 3.2 is proved.
Remark. Let
$N$
be the ideal in
${\Pi}_{0}^{\mu}$
generated by
$z$
, and
${H}_{k}\left(t\right)$
be the Hilbert polynomial of the quotient
${N}^{k}/{N}^{k+1}$
. The above arguments show that for generic
$\mu $
one has
$${\sum}_{k\ge 0}{H}_{k}\left(t\right){u}^{k}=\frac{{H}_{0}\left(t\right)u{t}^{h}P{H}_{0}\left(ut\right)}{1u{t}^{2}}.$$
Indeed, if
$\mu =\rho $
then using Propositions 4.4 and 4.5 as in the proof of Corollary 4.6 , we get
$${\sum}_{k\ge 0}{H}_{k}\left(t\right){u}^{k}={\sum}_{j=0}^{h2}{\sum}_{i=0}^{h2j}{t}^{2i+j}{u}^{j}{P}_{j}\left(C\right)=$$
$${\sum}_{j=0}^{h2}\frac{{t}^{j}{t}^{2(h1)j}{u}^{h1j}}{1u{t}^{2}}{P}_{j}\left(C\right)=\frac{{H}_{0}\left(t\right){t}^{2h2}{u}^{h1}{H}_{0}\left({t}^{1}{u}^{1}\right)}{1u{t}^{2}},$$
which gives the desired formula. Also, it is clear from this argument that for
$\mu =\rho $
the operator
$z:{\Pi}_{0}^{\mu}\left[j\right]\to {\Pi}_{0}^{\mu}[j+2]$
has maximal rank for all
$j$
.
Hence, the same formula applies to generic
$\mu $
.
5.2 Proof of Theorem 3.3
It follows from Theorems 3.1 , 3.2 that the algebras
$B$
and
$B\left[z\right]$
are Gorenstein. Since
${B}_{0}$
is finite dimensional, the algebra
${B}_{0}^{\mu}$
is finitely generated as a module over
$\mathbb{C}\left[z\right]$
. Moreover, by Theorem 1.2 of [
CB1]
,
$z$
must act by zero in every irreducible representation of
${B}_{0}^{\mu}$
. Thus,
${B}_{0}^{\mu}$
is finite dimensional.
This implies that
$B\left[z\right]$
is a finitely generated
${K}_{\mu}$
module. Since
$B\left[z\right]$
is a Gorenstein algebra, it is in particular CohenMacaulay, and by Proposition 2.3 (i),
$B\left[z\right]$
is a free
${K}_{\mu}$
module, hence parts (i),(ii) of Theorem 3.3 .
We also see that the Hilbert polynomial of
${B}_{0}^{\mu}$
is the same as that for
${B}_{0}^{\rho}$
, so Proposition 4.6 implies Theorem 3.3 , (iv), and the dimension formula of (v). The semisimplicity of
${B}_{\lambda}^{\mu}$
follows from the fact that for generic
$\lambda $
, by Theorem 1.2 of [
CB1]
, it has an irreducible representation corresponding to every positive root
$\alpha $
that involves
${x}_{p}$
; the sum of dimensions of these representations is
$h{q}_{1}{q}_{2}/2=dim{B}_{0}^{\mu}$
.
We also see, by Proposition
2.4 (ii), that
${B}_{0}^{\mu}$
is Gorenstein (=Frobenius).
This proves the first statement of Theorem
3.3 , (iii). The second statement follows from part (iv), and the third statement (the socle is
${z}^{h2}$
) follows from the fact that this is so for
$\mu =\rho $
(by Proposition 4.4 and Proposition 4.5 ). Theorem 3.3 is proved.
6 Relation to Hecke algebras of 2dimensional complex reflection groups
Let us discuss the connection of the above results with the theory of Hecke algebras for complex reflection groups, due to Broué, Malle, and Rouquier ([
BMR]
).
6.1 Finiteness of cyclotomic Hecke algebras in 2 dimensions
Let
$\Gamma $
be a 2dimensional irreducible complex reflection group, and
$H$
be its cyclotomic Hecke algebra (see [
BMR]
); it is defined by generators and relations which are deformations of the relations of
$\Gamma $
. The algebra
$H$
is a module over the algebra
$\mathbb{C}[\mathbf{T}]$
of functions on the torus
$\mathbf{T}$
of parameters.
Denote by
$H(\Lambda )$
the specialization of
$H$
at a point
$\Lambda \in \mathbf{T}$
.
It is known ([
BMR]
) that the dimension of
$H(\Lambda )$
is generically
$\ge \Gamma $
.
Moreover, it is conjectured (see [
BMR]
), that the algebra
$H$
is a free
$\mathbb{C}[\mathbf{T}]$
module, and hence
$H(\Lambda )$
has dimension exactly
$\Gamma $
for all values of
$\Lambda $
, i.e. it is a flat deformation of the group algebra
$\mathbb{C}[\Gamma ]$
. As far as we know, this conjecture has been checked by J. Müller using a computer for all cases except
${G}_{17},{G}_{18},{G}_{19}$
(see [
GGOR]
, Remark 5.12).
Here we give a proof of a weak version of this conjecture.
Theorem 6.1.
The algebra
$H$
is a finitely generated module over
$\mathbb{C}[\mathbf{T}]$
. Therefore, the algebra
$H(\Lambda )$
is finite dimensional for all
$\Lambda $
, and has dimension
$\Gamma $
for generic
$\Lambda $
, and
$\ge \Gamma $
for all
$\Lambda $
.
The rest of the subsection is devoted to the proof of this theorem^{6 }
.
The theorem is known for the infinite series of complex reflection groups ([
BMR]
), so it is sufficient to concentrate on the exceptional ones, i.e. the groups
${G}_{n}$
in the ShephardTodd classification [
ST]
,
$4\le n\le 22$
.
For the proof we need explicit presentations of the groups
$\Gamma $
. Recall that in all cases the image
$G$
of
$\Gamma $
in
$PG{L}_{2}(\mathbb{C})$
is the tetrahedral group, octahedral group, or icosahedral group. Thus the groups
${G}_{4},...,{G}_{22}$
fall into three families: tetrahedral, octahedral, and icosahedral. In each family, there is a maximal group
$\widehat{G}$
of order
$G{}^{2}$
, which is a central extension of
$G$
by a cyclic group of order
$\leftG\right$
(generated by an element
$Z$
), and all others are its subgroups.
The realizations of the particular groups
${G}_{n}$
,
$4\le n\le 22$
, are as follows.
The tetrahedral family.
The tetrahedral group
$G$
:
${a}^{2}={b}^{3}={c}^{3}=1$
,
$abc=1$
(order 12).
The maximal group
$\widehat{G}={G}_{7}$
(order 144):
${a}_{*}^{2}={b}_{*}^{3}={c}_{*}^{3}=1$
,
${a}_{*}{b}_{*}{c}_{*}=Z$
,
$Z$
is central; conjugacy classes of reflections are powers of
${a}_{*},{b}_{*},{c}_{*}$
.
Group
${G}_{4}$
(order 24) is generated by
$a={a}_{*}{Z}^{3},b={b}_{*}{Z}^{2},c={c}_{*}$
. The relations are
${a}^{2}={\zeta}^{1}$
,
${b}^{3}=\zeta $
,
${c}^{3}=1$
,
$abc=1$
,
$\zeta $
is central (here
$\zeta ={Z}^{6}$
).
Conjugacy classes of reflections are powers of
$c$
.
Group
${G}_{5}$
(order 72) is generated by
$a={a}_{*}{Z}^{1},b={b}_{*},c={c}_{*}$
. The relations are
${a}^{2}={\zeta}^{1}$
,
${b}^{3}=1$
,
${c}^{3}=1$
,
$abc=1$
,
$\zeta $
is central (here
$\zeta ={Z}^{2}$
).
Conjugacy classes of reflections are powers of
$b,c$
.
Group
${G}_{6}$
(order 48) is generated by
$a={a}_{*},b={b}_{*}{Z}^{1},c={c}_{*}$
. The relations are
${a}^{2}=1$
,
${b}^{3}={\zeta}^{1}$
,
${c}^{3}=1$
,
$abc=1$
,
$\zeta $
is central (here
$\zeta ={Z}^{3}$
).
Conjugacy classes of reflections are powers of
$a,c$
.
Octahedral family.
The octahedral group
$G$
:
${a}^{2}={b}^{3}={c}^{4}=1$
,
$abc=1$
(order 24). The maximal group
$\widehat{G}={G}_{11}$
(order 576):
${a}_{*}^{2}={b}_{*}^{3}={c}_{*}^{4}=1$
,
${a}_{*}{b}_{*}{c}_{*}=Z$
,
$Z$
is central; conjugacy classes of reflections are powers of
${a}_{*},{b}_{*},{c}_{*}$
.
Group
${G}_{8}$
(order 96) is generated by
$a={a}_{*}{Z}^{3},b={b}_{*}{Z}^{2},c={c}_{*}$
. The relations are
${a}^{2}={\zeta}^{1}$
,
${b}^{3}=\zeta $
,
${c}^{4}=1$
,
$abc=1$
,
$\zeta $
is central (here
$\zeta ={Z}^{6}$
).
Conjugacy classes of reflections are powers of
$c$
.
Group
${G}_{9}$
(order 192) is generated by
$a={a}_{*},b={b}_{*}{Z}^{1},c={c}_{*}$
. The relations are
${a}^{2}=1$
,
${b}^{3}={\zeta}^{1}$
,
${c}^{4}=1$
,
$abc=1$
,
$\zeta $
is central (here
$\zeta ={Z}^{3}$
).
Conjugacy classes of reflections are powers of
$a,c$
.
Group
${G}_{10}$
(order 288) is generated by
$a={a}_{*}{Z}^{1},b={b}_{*},c={c}_{*}$
. The relations are
${a}^{2}={\zeta}^{1}$
,
${b}^{3}=1$
,
${c}^{4}=1$
,
$abc=1$
,
$\zeta $
is central (here
$\zeta ={Z}^{2}$
).
Conjugacy classes of reflections are powers of
$b,c$
.
Group
${G}_{12}$
(order 48) is generated by
$a={a}_{*},b={b}_{*}{Z}^{4},c={c}_{*}{Z}^{3}$
. The relations are
${a}^{2}=1$
,
${b}^{3}={\zeta}^{1}$
,
${c}^{4}=\zeta $
,
$abc=1$
,
$\zeta $
is central (here
$\zeta ={Z}^{12}$
).
Conjugacy classes of reflections are powers of
$a$
.
Group
${G}_{13}$
(order 96) is generated by
$a={a}_{*},b={b}_{*}{Z}^{2},c={c}_{*}{Z}^{3}$
, and
$f={c}_{*}^{2}$
. The relations are
${a}^{2}=1$
,
${b}^{3}=\zeta $
,
${f}^{2}=1$
,
${c}^{2}\zeta =f$
,
$abc=1$
,
$\zeta $
is central (here
$\zeta ={Z}^{6}$
). Conjugacy classes of reflections are powers of
$a$
and
$f$
.
Group
${G}_{14}$
(order 144) is generated by
$a={a}_{*},b={b}_{*},c={c}_{*}{Z}^{1}$
. The relations are
${a}^{2}=1$
,
${b}^{3}=1$
,
${c}^{4}={\zeta}^{1}$
,
$abc=1$
,
$\zeta $
is central (here
$\zeta ={Z}^{4}$
).
Conjugacy classes of reflections are powers of
$a,b$
.
Group
${G}_{15}$
(order 288) is generated by
$a={a}_{*},b={b}_{*},c={c}_{*}{Z}^{1},f={c}_{*}^{2}$
.
The relations are
${a}^{2}=1$
,
${b}^{3}=1$
,
${f}^{2}=1$
,
${c}^{2}\zeta =f$
,
$abc=1$
,
$\zeta $
is central (here
$\zeta ={Z}^{2}$
). Conjugacy classes of reflections are powers of
$a,b$
, and
$f$
.
Icosahedral family.
The icosahedral group
$G$
:
${a}^{2}={b}^{3}={c}^{5}=1$
,
$abc=1$
(order 60).
The maximal group
$\widehat{G}={G}_{19}$
(order 3600):
${a}_{*}^{2}={b}_{*}^{3}={c}_{*}^{5}=1$
,
${a}_{*}{b}_{*}{c}_{*}=Z$
,
$Z$
is central; conjugacy classes of reflections are powers of
${a}_{*},{b}_{*},{c}_{*}$
.
Group
${G}_{16}$
(order 600) is generated by
$a={a}_{*}{Z}^{3},b={b}_{*}{Z}^{2},c={c}_{*}$
. The relations are
${a}^{2}={\zeta}^{1}$
,
${b}^{3}=\zeta $
,
${c}^{5}=1$
,
$abc=1$
,
$\zeta $
is central (here
$\zeta ={Z}^{6}$
).
Conjugacy classes of reflections are powers of
$c$
.
Group
${G}_{17}$
(order 1200) is generated by
$a={a}_{*},b={b}_{*}{Z}^{1},c={c}_{*}$
. The relations are
${a}^{2}=1$
,
${b}^{3}={\zeta}^{1}$
,
${c}^{5}=1$
,
$abc=1$
,
$\zeta $
is central (here
$\zeta ={Z}^{3}$
).
Conjugacy classes of reflections are powers of
$a,c$
.
Group
${G}_{18}$
(order 1800) is generated by
$a={a}_{*}{Z}^{1},b={b}_{*},c={c}_{*}$
. The relations are
${a}^{2}={\zeta}^{1}$
,
${b}^{3}=1$
,
${c}^{5}=1$
,
$abc=1$
,
$\zeta $
is central (here
$\zeta ={Z}^{2}$
).
Conjugacy classes of reflections are powers of
$b,c$
.
Group