I've just answered a simlair question on this forum (and I've posted my answer below)some people think that the wave length may be quabtized in Planck lengths and therefore the smallest wavelength possible for a photon is one that corrsponds to the Planck length, however due to differing refernce frames I find this unlikely:
It is highly debatebale whether or not the Planck length is the smallest possible divison esp. when referring to wavelenghths. I mentioned this above but I'll now illustrate this exactly:
The relativistic Doppler shift is given by the following:
z = Δλ/λ = [(1 + v/c)/(1 - v/c)]1/2 - 1
Where λ is the original wavelength, Δλ is the change in wavelength due to the Doppler effect, v is the relative velocity of the source and the observer and c is the speed of light in a vacuum.
This can be rearranged into the following:
λ' = (z + 1)λ
Where λ' is the observed wavelength (λ + Δλ ) and (z + 1) = [(1 + v/c)/(1 - v/c)]1/2
Now consider two beams of light with wavelengths (for an observer sationery to the source) Λ1 and Λ2 and two observers one sationery to the source and one moving with velocity, v, relative to the source. These two equations can then be derived from the equation above:
λ1' = (z + 1)λ1
λ2' = (z + 1)λ2
For the observer sationery to the source the difference between the wavelengths of the beams will be:
dλ = λ1 - λ2
For the observer moving with velocity, v, relative to the source the difference between the two wavelengths will be:
dλ' = λ1' - λ2'
We can then relate these two differences:
dλ' = (z+1)dλ
This tells us that the difference between the wavelengths of two beams of lights will be different for different reference frames, therefore in one refernce frame a difference between two wavelengths may be less than or equal to the Planck length yet in another it may be greater.