Suppose we work in number field $Q(\omega_{pq})$ where $\omega_{pq}$ is the $p*q$th

root of unit and $p*q$ are the product of two distinctive prime $p,q$. Similarly $\omega_{p}$ and $\omega_{q}$ are the $p$th and $q$th root of unit respectively. Now we note the binomial expansion

$(1-(1-\omega_{p}))^{\frac{p}{q}}=1-\frac{p}{q}*(1-\omega_{p})+ C_{\frac{p}{q}}^{2}*(1-\omega_{p})^{2} + ...$

By simple analysis, we know it should converge by non-archimedes topology to one of the $q$th root of unit $\omega_{q}^{k}$ at certain place $\beta$ over $1-\omega_{p}$, or equivalently over $p$. And denote this expansion as $s$, again by simple analysis of the expansion $s-1$, we know $s\neq1$.

So, by the property of product of converge series in non-archimedes topology we know $1, s, s^{2}, ... s^{q-1}$ will go through all the $q$th root of unit while all of them belong to $1+\pi*Z_{\beta}$ where $\pi$ is a primal element of place $\beta$ . So $1+s+s^{2}+...+s^{q-1}=q+\pi*Z_{\beta}\neq0$ which is a contradiction with the property of $q$th root of unit. So what is wrong in the above argument?