[Math] The converse of “nilpotent elements are zero-divisors”

abstract-algebranilpotencering-theory

For commutative rings $A$ with identity $1\ne0$, nilpotent elements are zero-divisors. The converse is false, i.e. there is a commutative ring $A$ with identity $1\ne0$ and a zero-divisor $x$ in $A$ which is not nilpotent. Where is such an example?

Is it meaningful to ask the "percentage" of commutative rings $A$ with identity $1\ne0$ for which the converse holds, i.e. zero-divisors are nilpotent?

Best Answer

An example is in the integers mod $6$, where $2\cdot 3=0$, but no power of either of these individually is zero.

If we had some sort of measure on a set of rings (which could not possibly be all rings, because there are too many) with finite total measure, we could ask about proportion. I know of no such commonly used measure, but it's possible such a thing has been considered.

In my intuition, the proportion would be small, but I can't think of a quick reason to see why, other than that being nilpotent seems like a special property while being a zero divisor seems very common. For example, the only time this is true for rings of integers modulo $n$ is when $n$ is a prime power, and there are vanishingly few of those compared to all integers.

If you look at the answer by rschwieb in the question Under what conditions does a ring R have the property that every zero divisor is a nilpotent element? linked above in a comment by Dietrich Burde, which is not the accepted answer, it characterizes these rings. A ring has this property if and only if it is the quotient of an arbitrary commutative ring by a primary ideal. An ideal $I$ is said to be primary if whenever we have that $a,b\in R$ satisfy $ab\in I$, then we have either $a\in I$ or $b^n\in I$ for some $n>0$. This answers part of your question. It is nicely illustrated in the $\Bbb Z_n$ case: an ideal $n\Bbb Z$ of $\Bbb Z$ is primary if and only if $n=p^k$ for some prime $p$ and integer $k>0$.

So rings like these are actually "close" to integral domains, which are quotients of arbitrary commutative rings by prime ideals. Again, no quantitative argument for why they should be rare, only an intuitive one.

Related Question