I understand what you mean now. The encoding of unicode string in sql server storage is UCS2 or utf-16, not utf-8. I don’t think you can change it.
What I said in my previosu post is that SQL accepts unicode string in any encoding, including utf-8. SQL server just converts it into utf-16.
I don’t know why you care SQL server’s encoding for unicode. You can always convert it back to utf-8 after you read the text into your application, right? Since utf-16 takes two bytes for each char, the space required in database is doubled. Is this your concern?