A semitone in musical terms is roughly equivalent to a frequency ratio of 1.059463 - in other words, play a C at 1.059463x its normal speed and you get a C#. You can use that ratio to scale one note up or down a bit so you can use fewer samples to cover the range that you want.

Or you could just generate the tone directly, by taking a musical note frequency table and using that as the parameters to a sine wave, but making that sound good can be tricky.

Yes, in fact that is how cheap sample libraries will work - they give you something like 4 samples per octave and stretch them to fit the other semitones in between.

Note frequency table. Again, this is only of use if you want to generate the note yourself. If you just want to stretch wav files (or if you can play them back at a different rate, achieving the same effect), you just need to know the ratio I posted above. Note that it's an exponential scale, so to go up 3 semitones you multiply the original frequency by that ratio three times, for example.

I'm not sure how C# play sounds, short of using the WinMM PlaySound function, but if you can find a library that allows you to alter the frequency of playback, that will be the easiest route to take. Failing that, you may have to write a little routine to 'stretch' the wave data in memory.