Er.... I think you fundamentally misunderstand the purposes of XML and
PDF. PDF is a document image format. XML is a document markup/structure
format. The twain are fundamentally different, although PDF could
conceivably be stored in an XML format rather than a binary format. If
you're even thinking along the lines of "my document format -> XML ->
PDF", you need to understand both of those technologies a whole lot
more before you start making software design decisions. It sounds to me
like you're designing by buzzword.
PDF files are essentially graphics images. There's a lot more to it
than that, but it boils down basically to this: you write to PDF files
just like you write to the screen or the printer or an image. If you
want to convert documents to .pdf format in your software, you'll need
to draw it manually, and you'll need a third-party library. Try
googling for stuff like "C# pdf converter" for third-party library
options. Here's the first item in the results:
http://www.colorpilot.com/pdfsample_csharp.html
Good luck.