0

Let $\mathbb{Q}(\sqrt{d})$ be finite extension with square free $d$, with $R_d$ being integral closure of subring $\mathbb{Z}$.

It is known that $R_d=\mathbb{Z}[\sqrt{d}]$ if $d\not\equiv 1\pmod{4}$ and $R_d=\mathbb{Z}[\frac{1+\sqrt{d}}{2}]$ if $d\equiv 1\pmod{4}$ are Dedekind domains.

With tools solely from number theory, is it possible to calculate ideal class group for $R_d$ when $d$ is small?

For instance, show ideal class group for $R_{-5}=\mathbb{Z}[\sqrt{-5}]$ is $\mathbb{Z}/2\mathbb{Z}$; for $R_{-23}=\mathbb{Z}[\frac{1+\sqrt{-23}}{2}]$ is $\mathbb{Z}/3\mathbb{Z}$.


Attempt: For prime ideal $P$ in $\mathbb{Z}[\theta_d]$, $P\cap\mathbb{Z}=(p)$ must be prime. Here $\theta_d$ is the adjoined element in closure.

If $(p)\subseteq\mathbb{Z}[\theta_d]$ is prime, then $P=(p)$, otherwise $(p)=(p,\theta_d+k)(p,\theta_d-k)$ and $P$ is one of them.

$R_{-5}=\mathbb{Z}[\sqrt{-5}]$ has nonprincipal prime ideal $P_2=(2,1+\sqrt{-5})$, so the ideal class group is not trivial.

To show ideal class group is $\mathbb{Z}/2\mathbb{Z}$ it suffices to show for every nonprincipal prime $Q=(p,\sqrt{-5}+k)$, $P_2Q$ principal.

This is where I stuck, is there elementary way to find single generator for $(2,1+\sqrt{-5})(p,\sqrt{-5}+k)$?

$R_{-23}=\mathbb{Z}[\frac{1+\sqrt{-23}}{2}]$ has nonprincipal prime ideal $P_2=(2,\frac{1+\sqrt{-23}}{2})$ and its square $P_2^2=(2,\frac{1-\sqrt{-23}}{2})$.

Similarly, it suffices to show for every nonprincipal prime $Q=(p,\frac{1+\sqrt{-23}}{2}+k)$, either $P_2Q$ or $P_2^2Q$ is principal.

I assume similar argument applies here, but there might be more case analysis.

Many answers I encountered uses (mentions) Minkowski bound, which I would like to avoid.

  • When you say With tools solely from number theory, I would think, Minkowski's bound is solely algebraic number theory (geometry of numbers). It is part of every ANT lecture, at least. So for the class number of $\Bbb Q(\sqrt{-23})$ I would use it, see here, as you mentioned. But there is, say, also this post, by the way, on $I=(3,1+\sqrt{-23})$. – Dietrich Burde Nov 24 '24 at 09:45
  • for imaginary quadratic fields, the group agrees with the group of (positive, primitive) binary quadratic forms. There is an explicit map http://zakuski.math.utsa.edu/~jagy/BLOG_2014_July_15.pdf Next, the forms classes will simply be the Gauss reduced forms. For $-5$ we have discriminant $-20$ and $x^2 + 5 y^2, ; ; ; 2x^2 + 2xy + 3 y^2$ where we say that the two forms lie in different genera. For $-23$ one genus, $x^2 + xy + 6 y^2,$ then $2x^2 + xy + 3 y^2$ finally $2x^2 - xy + 3 y^2$ – Will Jagy Nov 24 '24 at 17:38

0 Answers0