why escape sequence can't be represented as unicodeEscape in java?

The Unicode escape sequences are applied earlier in the source transformation than the character literal escape sequences. Unicode escape sequences are transformed very early in the process - before any other lexing happens, including before line breaks are detected. See JLS 3.2 for details.

So when you put \u000a into a Java source file, it will behave exactly as if you'd put an actual line feed in there - causing a line break as far as the rest of the compiler is concerned.

(Personally I think this was a design mistake; I prefer the C# approach of only allowing Unicode escape sequences at very specific points in the code, but that's a different matter.)