Dicyclic groups
On this page
Introduction
The dicyclic group is defined by $$\mathrm{Dic}_n = \left<a,\,x \mid a^{2n}=e,\; x^2=a^n,\; xax^{-1} = a^{-1}\right>.$$ It follows from this that \(x\) has order 4 and \(a\) order \(2n\), and \(\mathrm{Dic}_n\) has order \(4n\). For n=2, \(\mathrm{Dic}_2\) is the quaternion group Q8. Indeed, \(\mathrm{Dic}_n\) can be viewed as a subgroup of the unit quaternions, by putting \(x = j\) and \(a=\mathrm{e}^{i\pi/n}\).
The centre of \(\mathrm{Dic}_n\) is of order 2, generated by \(x^2 = a^n\), corresponding to -I in SU(2)), leading to $$0 \longrightarrow C_2 \longrightarrow \mathrm{Dic}_n \longrightarrow D_n \longrightarrow 0,$$ (where Dn is the dihedral group of order 2n, generated by a and x with x2= an=1). Presented in this way \(\mathrm{Dic}_n\) is seen to be the binary dihedral group.
The other normal subgroup (containing the centre) is the cyclic subgroup C2n generated by a, with $$0 \longrightarrow C_{2n} \longrightarrow \mathrm{Dic}_n \longrightarrow C_2 \longrightarrow 0.$$ The element x projects to the nontrivial element of C2. (Neither of these short exact sequences splits.)
Each of the two normal subgroups C2n and C2 lead to representations of Dicn from those of C2 and Dn respectively. The one arising from the non-trivial representation of C2 is denoted A1. (Since C2 is a subgroup of C2n this coincides with the A1 representation of Dn.)
These are all (isomorphic to) subgroups of the infinite group \(\mathrm{Dic}_\infty\), which is the subgroup of SU(2) generated by j and all \(\mathrm{e}^{i\theta}\). This has centre C2 (generated by -I) and satisfies $$0 \longrightarrow S^1 \longrightarrow \mathrm{Dic}_\infty \longrightarrow C_2 \longrightarrow 0.$$
Conjugacy Classes
In Dicn there are n+3 conjugacy classes, the first two forming the centre of the group. They are:
A simple argument shows that over C, Dicn has 4 1-dimensional representations. If n is even these are all real, while if n is odd two of these are complex conjugates, and their sum is then a real irreducible representation of dimension 2, of complex type (denoted B in the tables below).
Character tables
Note: ak and a2n-k are conjugate, as are xak and xa2n-k, and x and x-1.
Dic2
Dic2 is isomorphic to Q8. For example, put \(x\mapsto i\) and \(a\mapsto j\), so \(xa\mapsto k\).
| Dic2 | 1 | x2 | a | x | xa | notes |
| # | 1 | 1 | 2 | 2 | 2 | |Dic2|=8 |
| A0 | 1 | 1 | 1 | 1 | 1 | trivial rep |
| A1 | 1 | 1 | -1 | 1 | -1 | |
| A2 | 1 | 1 | 1 | -1 | -1 | |
| A3 | 1 | 1 | -1 | -1 | 1 | |
| H | 4 | -4 | 0 | 0 | 0 | quaternionic type |
- H is the representation arising from the identification of Dic2 with a subgroup of the unit quaternions (as described above, in fact with Q8). If we identify the quaternions with C2 one finds that H is the underlying real representation of the complex 2-dimensional irreducible representation, with character 2,-2,0,0,0.
Dic3
| Dic3 | 1 | x2 | a | a2 | x | xa | notes |
| # | 1 | 1 | 2 | 2 | 3 | 3 | |Dic3|=12 |
| A0 | 1 | 1 | 1 | 1 | 1 | 1 | trivial rep |
| A1 | 1 | 1 | 1 | 1 | -1 | -1 | |
| B | 2 | -2 | -2 | 2 | 0 | 0 | complex type |
| E | 2 | 2 | -1 | -1 | 0 | 0 | from D3 |
| H | 4 | -4 | -2 | 2 | 0 | 0 | quaternionic |
Notes
- H is the underlying real rep of the complex rep with character 2,-2,-1,1,0,0. And H=B⊗ E.
- B is the sum of two complex irreducibles, with characters
- For odd values of n, the real representations do not distinguish between {x} and {xa}.
Dic4
| Dic4 | 1 | x2 | a | a2 | a3 | x | xa | notes |
| # | 1 | 1 | 2 | 2 | 2 | 4 | 4 | |Dic4|=16 |
| A0 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | trivial rep |
| A1 | 1 | 1 | 1 | 1 | 1 | -1 | -1 | |
| A2 | 1 | 1 | -1 | 1 | -1 | -1 | 1 | |
| A3 | 1 | 1 | -1 | 1 | -1 | 1 | -1 | = A1⊗A2 |
| E | 2 | 2 | 0 | -2 | 0 | 0 | 0 | from D4 |
| H1 | 4 | -4 | 2√2 | 0 | -2√2 | 0 | 0 | quaternionic |
| H2 | 4 | -4 | -2√2 | 0 | 2√2 | 0 | 0 | quaternionic |
Dic5
| Dic5 | 1 | x2 | a | a2 | a3 | a4 | x | xa | notes |
| # | 1 | 1 | 2 | 2 | 2 | 2 | 5 | 5 | |Dic5|=20 |
| A0 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | trivial rep |
| A1 | 1 | 1 | 1 | 1 | 1 | 1 | -1 | -1 | |
| B | 2 | -2 | -2 | 2 | -2 | 2 | 0 | 0 | complex type |
| E1 | 2 | 2 | \(\gamma\) | \(\bar\gamma\) | \(\bar\gamma\) | \(\gamma\) | 0 | 0 | from D5 |
| E2 | 2 | 2 | \(\bar\gamma\) | \(\gamma\) | \(\gamma\) | \(\bar\gamma\) | 0 | 0 | from D5 |
| H1 | 4 | -4 | -2\(\bar\gamma\) | 2\(\gamma\) | -2\(\gamma\) | 2\(\bar\gamma\) | 0 | 0 | quaternionic |
| H2 | 4 | -4 | -2\(\gamma\) | 2\(\bar\gamma\) | -2\(\gamma\) | 2\(\gamma\) | 0 | 0 | quaternionic |
Notes
- \(\gamma = 2\cos(2\pi/5) = \frac12(\sqrt5-1)\) (=golden ratio)
- \(\bar\gamma = 2\cos(4\pi/5) = -\frac12(\sqrt5+1)\)
Dic6
| Dic6 | 1 | x2 | a | a2 | a3 | a4 | a5 | x | xa | notes |
| # | 1 | 1 | 2 | 2 | 2 | 2 | 2 | 6 | 6 | |Dic6|=24 |
| A0 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | trivial rep |
| A1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | -1 | -1 | |
| A2 | 1 | 1 | -1 | 1 | -1 | 1 | -1 | 1 | -1 | |
| A3 | 1 | 1 | -1 | 1 | -1 | 1 | -1 | -1 | 1 | |
| E1 | 2 | 2 | 1 | -1 | -2 | -1 | 1 | 0 | 0 | from D6 |
| E2 | 2 | 2 | -1 | -1 | 2 | -1 | -1 | 0 | 0 | from D6 |
| H1 | 4 | -4 | 2√3 | 2 | 0 | -2 | -2√3 | 0 | 0 | quaternionic |
| H2 | 4 | -4 | -2√3 | 2 | 0 | -2 | 2√3 | 0 | 0 | quaternionic |
| H3 | 4 | -4 | 0 | -4 | 0 | 4 | 0 | 0 | 0 | quaternionic |
General n
The pattern continues: for all n there are four 1-dimensional irreducible complex representations:
- n even: these are all real, denoted A0 ... A3;
- n odd: two are real (A0 and A1), and two are complex, say B1 and B2=B1*, and so the real 2-dimensional rep B = B1 + B1* is of complex type.
There are then representations (denoted Ei) arising from the homomorphism Dicn → Dn, and there are quaternionic representations (here denoted Hi) from Dicn → SU(2).