How can I prove this statement:
A formula is satisfiable iff its existential closure is satisfiable(the formula has $n$ free variables which are $x_1,\dots,x_n$).
A formula $\varphi$ is satisfiable if and only if there is a structure $\mathcal{M}$ such that $\mathcal{M} \vDash \varphi$ . I applied this definition to $\varphi\left(x_{1}, \ldots, x_{n}\right),$
and say there is a tuple $\vec{a}=(a_{1}, \ldots, a_{n})$ in M satisfying $\varphi$ in $\mathcal{M}$ s.t.
$ \mathcal{M} \vDash \varphi\left(a_{1}, \ldots, a_{n}\right)$
And applying this definition to $\exists x_{1}, \ldots x_{n} \varphi\left(x_{1}, \ldots, x_{n}\right),$
$ \mathcal{M} \vDash \exists x_{1}, \ldots, x_{n} \varphi\left(x_{1}, \ldots, x_{n}\right)$
and after I say these, which steps should I follow?
Welcome to MSE!
Hint:
Recall a formula $\psi$ is satisfiable if and only if in some model $\mathfrak{M} \models \psi$. This is the definition of satisfiability.
So applying this definition to $\varphi(x_1, \ldots, x_n)$, we see this is asking for
And applying this definition to $\exists x_1, \ldots x_n . \varphi(x_1, \ldots, x_n)$, we see we want
Do you see why these are equivalent conditions? It might be worth reviewing the definition of $\mathfrak{M} \models \exists z . \psi(z)$.
I hope this helps ^_^