I may not have an exact same picture with you, but here's the diagram I drew from my understanding:
You can depict that the asynchronous call is made to the existing execution specification on diagram, but is really a separate one (colored). But as a model, asynchronous messages should not be made between the synchronous message and its reply message.
What we'd suggest is to create another Server lifeline (e.g. Sever #2) and create Asynchronous message to the #2.
My answer may not be sufficient, so if you could paste a visual of how you'd like to model with a mark where you have a doubt for, it'd be very helpful.