The simplest way is conventional SEM. Just glue a piece of your paper to a stub with a conductive adhesive tape and coat with thin layer of gold or carbon with a sputter. BSE mode will be useful for detection of the particles - they will be looking as bright dots on the dark background.
Sample preparation for TEM is more tricky - resin embedding and slicing with an ultratome.
Do as Denis Korneev said, but be aware that carbon is better than gold for EDS and BSE, which you will need for nanoparticles identification. If your particles are really "nano" and their concentration is minor, you may need TEM, but anyway it's better to start with SEM.
You may need a cross section of your paper; for the beginning even a simple cut with a razor blade could bee good enough, but it is the second step of investigation.