I don't know the XDocument API as I do not practice C#, but from my own experience with others languages you are probably right when you are suspecting the '\t' and '\n'.
So there is two solution that come to my mind. You could try the following options:
- Use the CDATA element as the text inside a CDATA section will be ignored by the parser.
- Use a trim function to remove the white spaces and line breaks.