1

"The number $\sqrt{2}$ is not a rational number."

Proof. Suppose that $\sqrt{2}$ is a rational number.

Let $\mathbb{P}$ denotes the set of all prime numbers.

It follows that,

\begin{equation} \sqrt{2} = \frac{p}{q} \text{ where p,q $\in > \mathbb{Z}$ and $q \ne 0$} \end{equation}

By squaring both sides of the equation, one obtains:

\begin{equation} 2 = \frac{p^2}{q^2} \end{equation}

From

\begin{equation} 2q^2 = p^2, \end{equation}

it follows that,

\begin{equation} p^2 = 2r \text{ where r $\in \mathbb{Z^+}$}. > \end{equation}

From

\begin{equation} q^2 = \frac{p^2}{2}, \end{equation}

it follows that,

\begin{equation} q^2 = 2s \text{ where s $\in \mathbb{Z^+}$}. > \end{equation}

Because, for every positive even integers 2$\phi$ where $\phi \in > \mathbb{Z^+}$,

\begin{equation} 2\phi = uv \text{ where $u \in \mathbb{Z^+}$ and $v > \in \mathbb{P}$ ,} \end{equation}

it follows that,

\begin{equation} 2 = \frac{p^2}{q^2} = \frac{2r}{2s} = \frac{ax}{by} > \text{ where a,b $\in \mathbb{Z^+}$, $b \ne 0$, and $x,y \in > \mathbb{P}$}. \end{equation}

Consider the equation:

\begin{equation} x = 2\left(\frac{b}{a}\right)y \end{equation}

Because the $LHS$ of the above equation is a prime number and the $RHS$ is not a prime number, one obtains a contradiction.

Hence, the number $\sqrt{2}$ is not a rational number.


I have realized that the proof I had demonstrated before is incorrect, so I have put the incorrect proof in the above blockquote. The new proof is in the texts below.


I am trying to prove that $\sqrt{2}$ is not a rational number on my own using proof by contradiction. However, I am not sure whether my proof is correct or not.

The following is my proof.


"The number $\sqrt{2}$ is not a rational number."

Proof. Suppose that $\sqrt{2}$ is a rational number.

It follows that there must be an integer $p$ and a nonzero integer $q$ such that:

\begin{align} \sqrt{2} &= \frac{p}{q} \end{align}

By squaring both sides of the above equation, one obtains:

\begin{align} 2 &= \frac{p^2}{q^2} \end{align}

Rearranging the above equation so that $p^2$ is the subject, one obtains:

\begin{align} p^2 &= 2q^2 \end{align}

Thus,

\begin{align} p^2 = 2r \text{ where r $\in$ $\mathbb{Z^+}$} \end{align}

Also, from the equation $2 = \frac{p^2}{q^2}$, by rearranging the equation so that $q^2$ is the subject, one obtains:

\begin{align} q^2 = \frac{p^2}{2} \end{align}

Thus,

\begin{align} q^2 = 2s \text{ where s $\in$ $\mathbb{Z^+}$} \end{align}

Substituting $2r$ and $2s$ into the equation $2 = \frac{p^2}{q^2}$, one obtains:

\begin{align} 2 = \frac{p^2}{q^2} = \frac{2r}{2s} \end{align}

Since $p^2 = 2q^2$, it follows that $p^2 > q^2$.

Because both $p^2$ and $q^2$ are positive even numbers and $q^2 \geq 4$, it follows that $p^2 \geq 16$.

Thus, the equation $2 = \frac{p^2}{q^2} = \frac{2r}{2s}$ is expressed as:

\begin{align} 2 = \frac{p^2}{q^2} = \frac{(4+2m)^2}{(2+2n)^2} \text{ where m,n $\in$ $\mathbb{Z^*}$ } \end{align}

Because the term $\frac{(4+2m)^2}{(2+2n)^2} = \frac{(m+2)^2}{(n+1)^2}$ and $n \leq m+1$ for $\frac{(m+2)^2}{(n+1)^2}$ to be an integer, it follows that:

\begin{align} 2 = \frac{p^2}{q^2} = \frac{(4+2m)^2}{(2+2n)^2} = \frac{(m+2)^2}{(n+1)^2} \text{ where m,n $\in$ $\mathbb{Z^*}$ and $n \leq m+1$} \end{align}

Expanding $(m+2)^2$ as $m^2 +4m + 4$ and factor $2$ from the term, one obtians:

\begin{align} \frac{p^2}{q^2} = \frac{(m+2)^2}{(n+1)^2} = \frac{(m^2 +4m +4)}{(n^2 +2n +1)} = \frac{2(\frac{m^2}{2}+ 2m + 2)}{(n^2 + 2n +1)} = 2 \text{ where $m,n \in \mathbb{Z^*}$ and $n \leq m+1$} \end{align}

For the term $\frac{2(\frac{m^2}{2}+ 2m + 2)}{(n^2 + 2n +1)}$ to be equal to $2$, $\frac{(\frac{m^2}{2}+ 2m + 2)}{(n^2 + 2n +1)} =1$.

For $\frac{m^2}{2}+ 2m + 2$ to be an integer, $m=2k$ where $k \in \mathbb{Z^*}$.

Thus,

\begin{align} 1 = \frac{(\frac{m^2}{2}+ 2m + 2)}{(n^2 + 2n +1)} = \frac{(2k^2+ 4k + 2)}{(n^2 + 2n +1)} = \frac{2(k^2+ 2k + 1)}{(n^2 + 2n +1)} \text{ where $k,n \in \mathbb{Z^*}$ and $n \leq 2k +1$} \end{align}

Assuming there exists $k,n \in \mathbb{Z^*}$ such that $k^2 + 2k +1$ equals to the number $1$ and $n^2 + 2n +1$ equals to the prime number $2$.

Such $k,n$ make the term $\frac{2(k^2+ 2k + 1)}{(n^2 + 2n +1)} = 1$.

Because $n^2 +2n +1 \ne 2$ for all $n \in \mathbb{Z^*}$, one obtains a contradiction.

Hence, the number $\sqrt{2}$ is not rational.


Approxiz
  • 431
  • 3
  • 1
    Based on what you wrote, $\frac{b}{a} = \frac{x}{2y}$, so I don't believe there is a contradiction to derive. (I.e. the RHS is a prime number, it is $x$). I think the issue here is that $\frac{b}{a}$ is not necessarily an integer, hence it is not possible to deduce that the RHs is composite. – masiewpao Mar 31 '21 at 18:54
  • 1
    You must have a misprint: you appear to use $q^2=2s$, but introduced $s$ in the claim $p^2=2s$. – J.G. Mar 31 '21 at 19:14
  • 1
    How do you know $2\left(\frac{b}{a}\right)y$ isn’t a prime number? Maybe it’s $2\left(\frac{3}{6}\right)11$, say. – Steve Kass Mar 31 '21 at 19:21
  • @SteveKass I know where I have made a mistake now. I have not consider the term $\frac{b}{a}$ as $\frac{1}{2}$ when I tying to prove this making my proof incorrect. – Approxiz Mar 31 '21 at 19:23
  • @J.G. I have edited my original post and have posted a new proof. Is the new proof correct? Thank you in advanced. – Approxiz Apr 02 '21 at 16:26
  • @Approxiz How do you know $(m+2)^2=2(n+1)^2$ has no solutions? – J.G. Apr 02 '21 at 17:25
  • @J.G. If I were to factor out $2$ from $(m+2)^2$, I obtained $2(\frac{m}{\sqrt{2}} +\sqrt{2})^2$. Then the term $\frac{(m+2)^2}{(n+1)^2}$ becomes $\frac{2(\frac{m}{\sqrt{2}} +\sqrt{2})^2}{(n+1)^2}$. Because for $\frac{2(\frac{m}{\sqrt{2}} +\sqrt{2})^2}{(n+1)^2}$ to become $2$, the term $\frac{(\frac{m}{\sqrt{2}} +\sqrt{2})^2}{(n+1)^2}$ must be $1$. However, no integer $m$ could make the term $(\frac{m}{\sqrt{2}} +\sqrt{2})^2$ an integer, so there should not be a solution to $(m+2)^2 = 2(n+1)^2$. – Approxiz Apr 02 '21 at 18:46
  • @J.G. I do not know whether the above statement can be put in the prove or not because I am proving that $\sqrt{2}$ is not rational. – Approxiz Apr 02 '21 at 18:47
  • @Approxiz You don't know $m+2=(n+1)\sqrt{2}$ is impossible until you've proven $\sqrt{2}$ is irrational. – J.G. Apr 02 '21 at 18:48
  • @J.G. In this case, I need to explain further that why the last equation in my proof has no solution or that the equation is incorrect? – Approxiz Apr 02 '21 at 18:53
  • @Approxiz Once you've proven $p,,q$ are both even, you should finish the proof the usual way (by noting this implies $p/q$ has no in-lowest-terms expression, which is impossible for rational numbers.) What you've tried doesn't look like it gets you anywhere. – J.G. Apr 02 '21 at 18:55
  • @J.G. I have coded the equation in Python and it turns out that the equation seems to be true (at least for $m,n$ in range($100$)). Is there a way to prove or disprove the last equation? Sorry for asking many questions. I am trying to learn proof-writing on my own. – Approxiz Apr 02 '21 at 19:17
  • 1
    @Approxiz It's equivalent to the original problem, so you're really asking how prove $\sqrt{2}$ is irrational, given that your approach doesn't work. I've already advised you of one way to do this, which starts largely as per your strategy. A few proofs are given here. – J.G. Apr 02 '21 at 19:22
  • @masiewpao Sorry for asking another question after you have answered this question for some time ago. I have made a final attempt to prove that $\frac{(m+2)^2}{(n+1)^2} \ne 2$ in order to prove that the number $\sqrt{2}$ is not rational. Is my current proof a proof or a conjecture? Thank you in advanced! – Approxiz Apr 05 '21 at 14:49

1 Answers1

1

It is all fine up until you introduce $2\phi = uv$. Fundamental theorem of arithmetic states that $2\phi$ is already a product of powers of prime numbers. Then $v$ is just one specific prime that makes up $2\phi$ and does not tell us much about $2r/2s$.
To continue the proof, observe that $p^2 = 2r$ and $q^2 = 2s$ are both even numbers and $2|p^2,q^2$. But for $\sqrt{2}$ to be rational, we need that $\text{gcd}(p,q)=\text{gcd}(p^2,q^{2}) = 1$, which is not what we get. Hence, a contradiction and the proof is complete.