0

Assume that $p>q$, p and q are prime number, that $p^2$ is not congruent to $1\mod q$ and that $q^2$ is not congruent to $1\mod p$. Classify all groups of order $p^2q^2$.

Can somebody give me some advice on where to start here, what I know that I can use Sylows theorem in some way here, but that is where I get stuck.

R.vW
  • 359

3 Answers3

2

By Sylow's Third Theorem we have that there's only one Sylow $p-$ subgroup, as neither $q^2$ nor $q$ is congruent to $p$. Therefore as the only subgroup of the given order that subgroup is normal. Similarly there's only one Sylow $q-$ subgroup and it's normal too. Let them be $P$ and $Q$ respectively.

Now note that $P \cap Q = \{e\}$, as we have that $P \cap Q \le P, Q$, but the only number that divides both the orders of $P$ and $Q$ is $1$. Also note that $PQ=G$, as $PQ \le G$ and $|PQ| = \frac{|P||Q|}{|P \cap Q|} = p^2q^2$. But as $P,Q$ are normal subgroup the commutator $pqp^{-1}q^{-1} = (pqp^{-1})q^{-1} = p(qp^{-1}q^{-1}) \in P,Q$, so $pqp^{-1}q^{-1} = e \implies pq = qp; \quad\forall q \in Q, p \in P$

Now consider the map $\phi: P \times Q \to G$, defined $\phi(p,q)=pq$. Then we have that:

$$\phi((p,q)(p',q')) = \phi(pp',qq') = pp'qq' = (pq)(p'q') = \phi((p,q))\phi((p',q'))$$

So $\phi$ is a group homomorphism and if $\phi((p,q)) = e \implies pq = e$. But this means that $p,q \in P,Q$, so we must have that $p=q=e$. Hence $\ker(\phi) = \{e\}$. So $\phi$ is injective and we've proven that it's surjective already, so: $G \equiv P \times Q$.

Now as every group of order square of prime is abelian we have that $G$ is a the direct product of abelian groups, hence it's abelian. Now using the fundamental theorem of finitely generated abelian groups we have that:

$$G \equiv \mathbb{Z}_{p^2} \times \mathbb{Z}_{q^2}$$ $$G \equiv \mathbb{Z}_p \times \mathbb{Z}_{pq^2}$$ $$G \equiv \mathbb{Z}_{p^2q}\times \mathbb{Z}_{q}$$ $$G \equiv \mathbb{Z}_{pq} \times \mathbb{Z}_{pq}$$

Stefan4024
  • 35,843
  • It looks like you are missing the cyclic group of order $p^2q^2$. This is because every non-cyclic group of order of a square of a prime is abelian, as the duplicate of the linked question correctly claims. – wythagoras Oct 22 '16 at 11:39
  • @wythagoras But $\mathbb{Z}{p^2q^2} \equiv \mathbb{Z}{p^2} \times \mathbb{Z}_{q^2}$ – Stefan4024 Oct 22 '16 at 11:40
  • Yeah, nevermind. I should refrain from commenting on things I barely know something about. – wythagoras Oct 22 '16 at 11:49
  • 2
    @wythagoras Quite opposite. If you're not interested in things you don't know and discuss them no matter how wrong you're you will never learn about them. At the end there's nothing to be ashamed if you don't know something, but you have a will to learn. Greetings! – Stefan4024 Oct 22 '16 at 11:51
  • 1
    Thank you for your support. Let me rephrase my comment: I should refrain form trying to "correct" people on things I barely know something about, and phrase it as a question instead. – wythagoras Oct 22 '16 at 12:08
1

While I can't answer the entire question, I imagine this will help. There is only 1 Sylow p-subgroup, and only 1 Sylow q-subgroup.

Using Sylow's third theorem, the number of Sylow p-subgroups of $G$, $n_p$, must divide the index of the Sylow p-subgroup in $G$ (index $=q^2$). Similarly, $n_q$ must divide $p^2$. Since $p$ and $q$ are prime, only $1$,$p$,and $p^2$ divide $p^2$, while only $1$, $q$ and $q^2$ divide $q^2$.

Using Sylow's third theorem again, $n_p$ is congruent to $1\mod(p)$, and $n_q$ is congruent to $1\mod(q)$.

If $q$ were congruent to $1\mod(p)$, then $q^2$ would be congruent to $1\mod(p)$ (which the question tells us is false). Thus we know that $q$ can't be congruent to $1 \mod(p)$. We then know that $n_p = 1$, and $n_q = 1$ also (by exactly analogous reasoning).

If the above reasoning was too terse, you need to try $1$, $p$ and $p^2$ as possible values for $n_q$, and then you will find that only $n_q = 1$ works.

0

Try looking at the easier case where the order of $G$ is $pq$ (with similar conditions on $p$ and $q$ as your case for $p^{2}$ and $q^{2}$), which gives unique Sylow $p$-subgroup and unique Sylow $q$-subgroup. This then can tell you what type of group $G$ must be. Now try and generalize the ideas to your case.