This is my second question on Problem 8.21 on p.119 of The Theory of Algebraic Numbers by Harry Pollard and Harold G. Diamond (Dover edition); my first question has been answered.
Let $R=\mathcal O_K$ be the ring of integers of an algebraic field $K=\mathbb Q(\theta)$ for some $\theta$ algebraic over $\mathbb Q$. If $I\ne\{0\}$ is an ideal of $R$ and $\beta\in K$ is such that $\beta\alpha\in I$ for all $\alpha\in I$, prove that $\beta\in R$.
It is important that $K$ and $R$ be so defined, otherwise the result won't be true, as the counterexamples in my first question show. By studying the counterexamples, I presume that what may be important here are principally generated ideals. The ideal $I$ need not be principal but (by Theorem 8.13 on p.109 of the book) there exists an ideal $J\ne\{0\}$ such that the product $IJ$ is principally generated by a rational integer, i.e. $IJ=(c)$ where $c\in\mathbb Z$. I suspect this may be useful in solving the problem.
Another notion which may be useful is, given an ideal $I$, the set defined as $I^{-1}=\{\gamma\in K:\gamma\alpha\in R\ \text{for all}\ \alpha\in I\}$. Thus $\beta\in I^{-1}$ in the problem. If $J$ is an ideal, $JI^{-1}$ is defined as the set of all finite sums of products of elements of $J$ and of $I^{-1}$; if $J\subseteq I$ then $JI^{-1}$ is an ideal.
I have been puzzling over this problem for a while and would welcome any help with it. Thanks.
We are given that $\beta I\subseteq I$. If $I$ is principal, say $I=(\iota)$ with $\iota\ne 0$, then $\beta \iota= r\iota $ for some $r\in R$ and hence $\beta=r\in R$.
For any other non-zero ideal $J$, we can replace $I$ with $IJ$: If $\beta I \subseteq I$ then also $\beta IJ\subseteq IJ$. Using Theorem 8.19 we can pick $J$ so that $IJ$ is a non-zero principal ideal and are done.