PSF Measurements In Fluorescence Imaging

Question

Quite a technical question!

I have measured the Point Spread Function of 100nm fluorescent breads with my Olympus scanning head. I'm two-photon exciting the beads with a wavelength of 800nm and focused in the sample with a 100x with N.A.: 1.4

The theory suggests me the resolution of such a system to be: $$d=\frac{0.7\cdot\lambda}{N.A.}\approx 400nm$$

Now that value should be equal (with at most a 18% correction) to the FWHM of the 2D gaussian I obtain on the image.

But from my analysis of the images I obtain a FWHM close to 2um. Now surely the formula is for an ideal optical setup but a factor five seems to me quite strange! Is that possible to obtain such a result, in the case of very not ideal optical elements, or should I look for some sort of problem with the acquisition sw that tells me the pixel dimension of the images?

Answer 1

I've found the solution to my problem. First of all, I had a factor 2 discrepancy due to the pixel dimension and, more important from an optical point of view, the laser beam was underfilling the entrance pupil of the objective. This means the focusing was worse!

Answer 2

You have many things which can go wrong when you try to go down to the wavelength limit, without resorting to Very not ideal optical element : non-Gaussian beam, wrong wavelength, spherical aberration, etc. I fear there is no easy answer to your question.

Answer 3

The resolution you obtain depends on the detection numerical aperture and your camera pixel size (Nyquist criteria). Here in your case, the illumination numerical aperture can be 1.4; the same as what your objective company says, but the detection numerical aperture is different and the immersion media of your sample and lens determine that. Also, your total detection magnification can determine your effective pixel size and image formation that can degrade your resolution.