As per K8S documentation ( Running in multiple zones ), there is no requirement for master and worker node to be in same zone. It can span across multiple zones.

It also states that if availability is an important concern, replicas should be created for control plane components and for pods across availability zones.

I would like to expand a bit the answer.

In the documentation one can find:

Kubernetes is designed so that a single Kubernetes cluster can run across multiple failure zones, typically where these zones fit within a logical grouping called a region.

There is no requirement that you have both master and worker node in the same zone.

If availability is an important concern, select at least three failure zones and replicate each individual control plane component (API server, scheduler, etcd, cluster controller manager) across at least three failure zones.

You should remember, that Kubeentes doesn't provide cross-zone resilience for the API server endpoints. To improve availability one can use such techniqes like DNS round-robin, SRV records.