Proving that the zeros of the Riemann zeta function with real part $1/2$ is reflected upon the imaginary axis

81 Views Asked by At

Consider the Functional formula of the Riemann Zeta function

$$\zeta(s)=2^{s}\pi^{s-1}\sin\left(\frac{\pi s}{2}\right)\Gamma(1-s)\zeta(1-s)$$ Let $s=\frac{1}{2}+it$ for $t\in\mathbb{R}$ $$\zeta\left(\frac{1}{2}+it\right)=2^{\frac{1}{2}+it}\pi^{-\frac{1}{2}+it}\sin\left(\frac{\pi}{2}\Big(\frac{1}{2}+it\Big)\right)\Gamma\Big(\frac{1}{2}-it\Big)\zeta\Big(\frac{1}{2}-it\Big)$$ Lets say $\zeta\left(\frac{1}{2}+it\right)=0$ for some $t$.

Then $2^{\frac{1}{2}+it}\pi^{-\frac{1}{2}+it}\sin\left(\frac{\pi}{2}\Big(\frac{1}{2}+it\Big)\right)\Gamma\Big(\frac{1}{2}-it\Big)\zeta\Big(\frac{1}{2}-it\Big)$ at least one of the products must be $0$. In order to prove that $\zeta\left(\frac{1}{2}+it\right)=0$ if and only if $\zeta\left(\frac{1}{2}-it\right)=0$. We will show that $2^{\frac{1}{2}+it}\pi^{-\frac{1}{2}+it}\sin\left(\frac{\pi}{2}\Big(\frac{1}{2}+it\Big)\right)\Gamma\Big(\frac{1}{2}-it\Big)$ is never zero.

$\mathbf{Lemma\, 1.1}$ $$2^{\frac{1}{2}+it}\pi^{-\frac{1}{2}+it}\sin\left(\frac{\pi}{2}\Big(\frac{1}{2}+it\Big)\right)\Gamma\Big(\frac{1}{2}-it\Big)\neq 0\,\,,\,\forall t\in \mathbb{R}$$ $\mathbf{Proof:}$ We will show that they are non-zero term by term.

Obviously, $2^{\frac{1}{2}+it}\pi^{-\frac{1}{2}+it}$ cannot be zero, that clears out. $$\sin\left(\frac{\pi}{2}\Big(\frac{1}{2}+it\Big)\right)=\sin\Big(\frac{\pi}{4}+it\frac{\pi}{2}\Big)=\sin\Big(\frac{\pi}{4}\Big)\cos\Big(it\frac{\pi}{2}\Big)+\cos\Big(\frac{\pi}{4}\Big)\sin\Big(it\frac{\pi}{2}\Big)$$ $$=\frac{1}{\sqrt{2}}\cosh\Big(\frac{\pi}{2}t\Big)+\frac{i}{\sqrt{2}}\sinh\Big(\frac{\pi}{2}t\Big)$$ But we know $\cosh(t)$ is non-zero for all real $t$. This shows that $\sin\Big(\frac{\pi}{4}+it\frac{\pi}{2}\Big)\neq 0$

Now we only have to prove that $\Gamma\Big(\frac{1}{2}-it\Big)$ is non-zero and our proof will be done.

Recall the Weierstrass definition of the Gamma function $$\Gamma(s)=\frac{e^{-\gamma s}}{s}\prod_{k=1}^{\infty}\frac{e^{s/k}}{1+\frac{s}{k}}$$ Note that $x=0$ if and only if $|x|=0$. Plugging in $s=\frac{1}{2}+it$ and take absolute value on both sides gives $$\Big|\Gamma\Big(\frac{1}{2}+it\Big)\Big|=\left|\frac{e^{\frac{-\gamma}{2}-it\gamma }}{\frac{1}{2}+it}\right|\left|\prod_{k=1}^{\infty}\frac{e^{\frac{1}{2k}+\frac{it}{k}}}{1+\frac{1}{2k}+\frac{it}{k}}\right|=\frac{e^{-\frac{\gamma}{2}}}{|\frac{1}{2}+it|}\prod_{k=1}^{\infty}\frac{e^{\frac{1}{2k}}}{|1+\frac{1}{2k}+\frac{it}{k}|}$$ $$=\frac{e^{-\frac{\gamma}{2}}}{\sqrt{\frac{1}{4}+t^2}}\prod_{k=1}^{\infty}\frac{e^{\frac{1}{2k}}}{\sqrt{(1+\frac{1}{2k})^2+(\frac{t}{k})^2}}$$ $$\Rightarrow \Big|\Gamma\Big(\frac{1}{2}+it\Big)\Big|^2=\frac{e^{-\gamma}}{t^2+\frac{1}{4}}\prod_{k=1}^{\infty}\frac{e^{\frac{1}{k}}}{(1+\frac{1}{2k})^2+(\frac{t}{k})^2}$$ Remember the goal is to show that the entire thing is non-zero. We can show that $\prod_{k=1}^{\infty}\frac{e^{\frac{1}{k}}}{(1+\frac{1}{2k})^2+(\frac{t}{k})^2}$ is bounded above some $a>0$.

consider: $$a_k=\frac{(1+\frac{1}{2k})^2+(\frac{t}{k})^2}{e^{\frac{1}{k}}}$$ We have: $$\Big|\Gamma\Big(\frac{1}{2}+it\Big)\Big|^2=\frac{e^{-\gamma}}{t^2+\frac{1}{4}}\prod_{k=1}^{\infty}\frac{1}{a_k}$$ Consider the product $$\prod_{k=1}^{m}a_k=\prod_{k=1}^{m}\frac{(1+\frac{1}{2k})^2+(\frac{t}{k})^2}{e^{\frac{1}{k}}}=e^{-H_m}\prod_{k=1}^{m}\Big(1+\frac{1}{k}+\frac{1}{4k^2}+\frac{t^2}{k^2}\Big)$$ Using the taylor expansion of $e^x$ $$1+x\leq e^x$$ $$\Rightarrow 1+\frac{1}{k}+\frac{1}{4k^2}+\frac{t^2}{k^2}\leq e^{\frac{1}{k}+\frac{1}{4k^2}+\frac{t^2}{k^2}}$$ $$\Rightarrow \prod_{k=1}^m \Big(1+\frac{1}{k}+\frac{1}{4k^2}+\frac{t^2}{k^2}\Big)\leq \prod_{k=1}^m e^{\frac{1}{k}+\frac{1}{4k^2}+\frac{t^2}{k^2}}=e^{H _m+(t^2+\frac{1}{4})\sum_{k=1}^m\frac{1}{k^2}}$$ $$\leq e^{H_m +(t^2+\frac{1}{4})\zeta(2)}$$ Therefore $$\prod_{k=1}^{m}a_k \leq e^{-H_m}\cdot e^{H_m +(t^2+\frac{1}{4})\zeta(2)}=e^{(t^2+\frac{1}{4})\zeta(2)}$$ $$\Rightarrow \prod_{k=1}^{m}\frac{1}{a_k}\geq e^{-(t^2+\frac{1}{4})\frac{\pi^2}{6}}\,\,,\, \forall m\in\mathbb{N}$$ At the end, we have $$\Big|\Gamma\Big(\frac{1}{2}+it\Big)\Big|^2=\frac{e^{-\gamma}}{t^2+\frac{1}{4}}\prod_{k=1}^{\infty}\frac{1}{a_k}\geq \frac{e^{-\gamma-(t^2+\frac{1}{4})\frac{\pi^2}{6}}}{t^2+\frac{1}{4}}$$ Which is certainly greater than zero. $$ \Big|\Gamma\Big(\frac{1}{2}+it\Big)\Big|^2 \neq 0 \Rightarrow \Gamma\Big(\frac{1}{2}+it\Big) \neq 0$$

Hence the lemma is proved

$$\zeta\left(\frac{1}{2}+it\right)=2^{\frac{1}{2}+it}\pi^{-\frac{1}{2}+it}\sin\left(\frac{\pi}{2}\Big(\frac{1}{2}+it\Big)\right)\Gamma\Big(\frac{1}{2}-it\Big)\zeta\Big(\frac{1}{2}-it\Big)$$ $$\zeta\left(\frac{1}{2}+it\right)=\mathrm{(Not zero)}\zeta\Big(\frac{1}{2}-it\Big)$$ Finally, this shows that if $\zeta\left(\frac{1}{2}+it\right)=0$, then $\zeta\left(\frac{1}{2}+it\right)=0$ as well for all $t\in\mathbb{R}$.

Graphically, this tells us that the zeros of the Riemann zeta function on the critical line is reflective

$2$ Questions:

$(1)$ Is my proof correct?

$(2)$ I somehow felt like this is an excessive solution. Is there an alternate proof?

1

There are 1 best solutions below

1
On BEST ANSWER

Your proof is correct, but the statement you are trying to prove is valid for any nontrivial zeros of $\zeta(s)$.

Note that when $s\in\mathbb R$ we have $\zeta(s)\in\mathbb R$, so by the principle of reflection we see that $\zeta(\overline s)=\overline{\zeta(s)}$. This indicates that $\zeta(s)=0$ iff $\zeta(\overline s)=0$. That is, all nonreal zeros of $\zeta(s)$ are symmetric about the real axis.