Unicode superscript (and index) code points are designed for backward compatibility with older character sets and are essentially not recommended. They are not intended to be filled in or may be used to format text. If you need to use top-level characters, the recommended (and most comprehensive / compatible) approach is to use a markup language.
If the markup language is not available, you're just out of luck. Hacking another character that looks similar may be a useful workaround, but it hurts any semantic meaning for the text - for example, U+22C5 can be mechanically analyzed as a multiplication of neighboring numbers.
source share