Unicode data uses NCHAR / NVARCHAR / NTEXT datatypes rather than CHAR /
VARCHAR / TEXT. This is defined at the column level rather than the database
level.

To change your database you could script all the tables, replace the unicode
datatypes with their non-unicode equivalents and then run the creation
script to create the new database. This might affect the behaviour of stored
procedures and other code but if the app is designed to cope with this then
presumably it shouldn't be a problem. Correctly converting your unicode data
to a non-unicode collation might be more of a challenge though.

--
David Portas
SQL Server MVP
--