Let $G\leq \operatorname{SO}_{6}(\mathbb Z)$ be a finite-index normal subgroup, so it's a Zariski dense subgroup of $\operatorname{SO}_{6}(\mathbb C)$; and let $H$ be the subset of $\operatorname{SO}_{6}(\mathbb C)$ given by $$ H=\left\{e_{x,y}= \begin{pmatrix} 0 & 0 & 1 & 0 & 0 & 0 \\ 0 & 0 & 0 & 0 & x & 0 \\ 0 & 0 & 0 & 0 & 0 & y \\ \frac{1}{y} & 0 & 0 & 0 & 0 & 0 \\ 0 & \frac{1}{x} & 0 & 0 & 0 & 0 \\ 0 & 0 & 0 & 1 & 0 & 0 \\ \end{pmatrix} \middle| x,y\in\mathbb C^* \right\}. $$
Let $B=B(\mathbb{C})$ be the Borel subgroup which consists of upper triangular matrices.
Is it true that $G\cap Be_{1,1}B$ is necessarily non-empty?
As noticed in the comments, $G\cap H$ is empty for most congruence subgroups $G$.
Here $\operatorname{SO}_{n}(k)$ is defined as $\{A\in\operatorname{GL}_{n}(k)\mid A^TJ_nA=J_n\}$, where $J_n$ is the identity matrix flipped 90 degrees.
My progress: I've been trying to study this subset of $Be_{1,1}B\cap\operatorname{SO}_{6}(\mathbb Z)$: $$ L=\left\{\begin{pmatrix} b_{1} & b_{3} & 1 & a_{2} b_{3}+a_{1} b_{1} & a_{3} b_{1}-a_{2} & -a_{3} b_{3}-a_{1} \\ b_{2} & 0 & 0 & a_{1} b_{2}-b_{3} & a_{3} b_{2}+1 & 0 \\ 0 & -b_{2} & 0 & -a_{2} b_{2}-b_{1} & 0 & a_{3} b_{2}+1 \\ 1 & 0 & 0 & a_{1} & a_{3} & 0 \\ 0 & 1 & 0 & a_{2} & 0 & -a_{3} \\ 0 & 0 & 0 & 1 & 0 & 0 \\ \end{pmatrix} \middle| a_i,b_i\in\mathbb Z\right\}. $$
I noticed that the characteristic polynomials of the matrices in $L$ are of the form $p_{a,b,c}(x)=1+ax+bx^2+cx^3+bx^4+ax^5+x^6$, maybe
$ \{A\in \operatorname{SO}_{6}(\mathbb C)|\exists a,b,c\in\mathbb{C}\;\text{s.t.}\;p_{a,b,c}(x)=p_{\text{char}}(A)(x)\}\subset Be_{1,1}B$
Here:
- $p_{\text{char}}(A)(x)$ is the characteristic polynomial of $A$.
There is something similar in $\operatorname{SL}_{n}(\mathbb C)$ where instead of $e_{1,1}$ we take $e$ to be the permutation matrix of $(123..n)$ and $G$ a finite index normal subgroup of $\operatorname{SL}_{n}(\mathbb Z)$ and $B$ the upper triangular matrices. Then since $J=\{A\in \operatorname{SL}_{n}(\mathbb C)|disc(A)\neq 0\}$ is Zariski open there is a $g\in J\cap G$ and since every element of $J$ is similar to some $b_1eb_2$ ,$b_i\in B$ , we get $g\in G\cap BeB$.
Here $disc(A)$ is the discriminant of the characteristic polynomial of A.