A complex number is called a root of unity if , for some positive integer .
Given a positive integer , let be the set of all roots of unity of the form
where and and are positive integers.
(i) Prove that has less than distinct elements.
(ii) Prove that is empty.
Solution
(i) For a root of unity , with and real and , we have , i.e. . So, if , for positive integers and , we have .
Since and are positive integers, and, once is known, is also known, so has fewer than distinct elements, as asserted.
(ii) Suppose to the contrary that is a root of unity, where and are positive integers. Then and we have two possibilities: either and or and .
First note that and that is a root of unity iff is a root of unity. Hence it is sufficient to consider the case
Suppose is a root of unity, say , where is a positive integer. We can write , and obtain , , so , for some integer , and if is any integer, then , for some non-negative integer .
In particular, there are only finitely many possible values for as runs through the set of integers.
We claim:
where is an integer not divisible by .
We prove the claim by induction on . When , , and the claim holds.
Assume that is an integer and that the claim holds for . This means that , with not divisible by , and thus
and is an integer not divisible by . Thus the claim holds for . Hence the claim is proved for all integers .
It follows that all the values of , are distinct, since, when written as a fraction in reduced form, all the denominators are distinct. But this contradicts what we have shown above, namely that there are only finitely many possible values for as runs through the set of integers. Hence is empty, as required.