My attempt: by contradiction suppose 'a' (nonzero) is a zero divisor so there exists some 'b' (nonzero) which a.b=0 since ax=b has an answer so a(ax)=0 implies a.a(x)=0 now i have no idea.
Suppose R is a ring, in which for any nonzero a and b in R the equation ax=b has an answer. prove that R doesnt have left or right zero divisor.
486 Views Asked by user625854 https://math.techqa.club/user/user625854/detail AtThere are 3 best solutions below
On
Assume for contradiction that $a\neq 0$ is a right zero divisor, with $c\neq 0$ such that $ca = 0$. By assumption, there is some $b\in R$ such that $cb\neq 0$ (such a $b$ must exist; take any $d\neq 0$ in the ring*, and let $b$ be the solution to $cx = d$). Now $ax = b$ has a solution, but $0 = cax = cb$ is a contradiction.
The same contradiction shows that there cannot be any left zero-divisor $c$ in our ring (let $c$ be a left zero-divisor with $ca = 0$ and again look at $cax = cb$).
*It may be that the ring just has the zero element. But then it clearly doesn't have zero divisors, so we're still good.
On
I'm adding this answer to augment Wuestenfux's solution by demonstrating the ring has an identity.
Let $a$ be nonzero. Then $ae=a$ for some $e$. Then also $ay=e$ for some $y$. Stringing these together, $aay=ae=a$, so $R$ is strongly regular, and therefore a reduced ring.
With this, you can compute $(ea-a)^2=eaea-eaa-aea+aa=eaa-eaa-aa+aa=0$ so $ea=a$ as well.
Now let $b$ be any other nonzero element. There must exist $z$ such that $az=b$. You immediately have $eb=eaz=az=b$.
We similarly compute that $(be-b)^2=bebe-beb-bbe+bb=bbe-bb-bbe+bb=0$ so $be=b$ as well.
So $e$ is the identity for the ring. At this point you can either proceed as Wuestenfux has outlined, or else observe that the condition makes the ring a simple right module over itself, so therefore it is a division ring, and has no nonzero zero divisors.
Note that if the equation $ax=b$, $a\ne 0$, can be solved, then in particular the equation $ax=1$, where $1$ is the unit element, can be solved. So each element $a\ne 0$ has a multiplicative inverse.
But units are no zero divisors. Indeed, if $a\in R$ is a unit and $ab=0$, then $0 = a^{-1}0 = a^{-1}(ab) = (a^{-1}a)b=1b=b$.