Emission spectrum changes with microscope objective?

[Yinxiao Li]

I am using a Renishaw inVia confocal microscope to study emission spectrum of molecules. The excitation wavelength of laser that I use is 532nm. However, I met with a very basic problem. When I use the same material (say plain silicon), and use different objectives (20X, 50X, 100X), then I will get totally different emission spectra. The spectra from each objective are consistent. My understanding is that the emission spectrum should definitely not change when I use different objective. This is driving me crazy. Any possible explanations?

 

[mfb]

I don't know what is causing it, but here are some ideas to narrow it down:
What is the wavelength range of your spectra?
What do you get without any sample?
What do you get without laser?

 

[Yinxiao Li]

I don't know what is causing it, but here are some ideas to narrow it down:
What is the wavelength range of your spectra?
What do you get without any sample?
What do you get without laser?

The wavelength that I tested is 550nm-740nm. The excitation wavelength is 532nm.
Without any sample, I will only get some noise.
Currently I use 5% of laser power. If I change it to 0.00001% of laser power (close to without laser), I will only get noise.

 

[Drakkith]

What is different about the emission spectra? How big of a change are we talking about?

 

[Andy Resnick]

I am using a Renishaw inVia confocal microscope to study emission spectrum of molecules. The excitation wavelength of laser that I use is 532nm. However, I met with a very basic problem. When I use the same material (say plain silicon), and use different objectives (20X, 50X, 100X), then I will get totally different emission spectra. The spectra from each objective are consistent. My understanding is that the emission spectrum should definitely not change when I use different objective. This is driving me crazy. Any possible explanations?

The lenses may have slightly different (spectral) transmissions. Do you have any information about the objective lenses?
I'm slightly familiar with the inVia Raman microscope and I looked up the spectrum of Silicon, but I can't easily parse the change from wavenumber shift to wavelength. Can you tell us the actual wavelength range you are looking at?

Edit- I just saw Post #3. Hang on...

Edit #2: Ok. I'd still like more information about the objective lenses, since you are looking out into the far red, past the wavelength range most lenses are chromatically corrected for.

 

[mfb]

Different spectral transmissions, and probably different lensing behavior as well, if the lenses are not designed for this wavelength range. That could explain some trends. If completely new lines show up or vanish then things are more complicated. It would help to see example spectra.

 

[Yinxiao Li]

Different spectral transmissions, and probably different lensing behavior as well, if the lenses are not designed for this wavelength range. That could explain some trends. If completely new lines show up or vanish then things are more complicated. It would help to see example spectra.

I need to correct what I said before. When there is no sample, I can still get some signal with the same settings (5% power, 10s expose), although the signal is weaker.
I attach three example spectra here. Two of them correspond to the silicon substrate----one is absolute counts, and the other one is normalized. The third one correspond to no-sample based on absolute counts. Hopefully these will help!

For the objective lens, I am not sure how to check it. Maybe I should call the company for more information?

 

[Yinxiao Li]

What is different about the emission spectra? How big of a change are we talking about?

In my last post, I attach the example spectra. You can clearly see there is a big difference in peak. I would say the spectra are totally different.

 

[Drakkith]

I don't know anything about how the whole system is set up. Is it possible you need to change the distance between different elements in the optical train when you change lenses?

 

[mfb]

The two peaks around 560 nm are visible in all three spectra and they are at the same place, that is nice.
Also, 50 and 100 look very similar. 20 is odd, even without sample.

For the objective lens, I am not sure how to check it. Maybe I should call the company for more information?

They certainly have some data on transmission and refractive index of their lenses as function of the wavelength. It could be in the documentation as well.

 

[Yinxiao Li]

The lenses may have slightly different (spectral) transmissions. Do you have any information about the objective lenses?
I'm slightly familiar with the inVia Raman microscope and I looked up the spectrum of Silicon, but I can't easily parse the change from wavenumber shift to wavelength. Can you tell us the actual wavelength range you are looking at?

Edit- I just saw Post #3. Hang on...

Edit #2: Ok. I'd still like more information about the objective lenses, since you are looking out into the far red, past the wavelength range most lenses are chromatically corrected for.

The objective lens is PLAN EPI--so it is designed for epi illumination. The 100X has a numerical aperture of 0.85 while 50X has a numerical aperture of 0.75. These information can be found directly on the body of the lens. They are the information you are looking for or not?

 

[Andy Resnick]

The objective lens is PLAN EPI--so it is designed for epi illumination. The 100X has a numerical aperture of 0.85 while 50X has a numerical aperture of 0.75. These information can be found directly on the body of the lens. They are the information you are looking for or not?

Not exactly- I was hoping for the manufacturer name and model number of each- Renishaw doesn't fabricate objective lenses.

I agree, there are significant differences between the 20X and the other two: the 50X and 100X could potentially be brought into agreement fairly easily with 'reasonable' spectral weighting. OTOH, if the 20X lens was specifically designed for far-red transmittance, that could also explain the difference- again, if you can provide the manufacturer and model number of each lens, that's how to start chasing down objective lens-dependent changes.

Now, here's a question since I don't know that much about Raman imaging: what is the angle and/or polarization dependence of the Raman-shifted light? I ask because the numerical aperture of the lenses are different, so you could be detecting differences due to changes in either incident angle and polarization state (which will vary azimuthally for a high NA lens)

 

[Yinxiao Li]

Not exactly- I was hoping for the manufacturer name and model number of each- Renishaw doesn't fabricate objective lenses.

I agree, there are significant differences between the 20X and the other two: the 50X and 100X could potentially be brought into agreement fairly easily with 'reasonable' spectral weighting. OTOH, if the 20X lens was specifically designed for far-red transmittance, that could also explain the difference- again, if you can provide the manufacturer and model number of each lens, that's how to start chasing down objective lens-dependent changes.

Now, here's a question since I don't know that much about Raman imaging: what is the angle and/or polarization dependence of the Raman-shifted light? I ask because the numerical aperture of the lenses are different, so you could be detecting differences due to changes in either incident angle and polarization state (which will vary azimuthally for a high NA lens)

The objective lenses are LEICA 566066 N PLAN EPI 20X/0.4, LEICA 566072 N PLAN EPI 50X/0.75, and LEICA 566073 N PLAN EPI 100X/0.85. Honestly I cannot find any useful information from these information...Do you have any clues?

To the best of my knowledge, there should be no angle or polarization dependence of the Raman-shifted light...

 

[Andy Resnick]

The objective lenses are LEICA 566066 N PLAN EPI 20X/0.4, LEICA 566072 N PLAN EPI 50X/0.75, and LEICA 566073 N PLAN EPI 100X/0.85. Honestly I cannot find any useful information from these information...Do you have any clues?

To the best of my knowledge, there should be no angle or polarization dependence of the Raman-shifted light...

Thanks for the info... Leica's database is currently down, tho... bleah...

 

[Andy Resnick]

The objective lenses are LEICA 566066 N PLAN EPI 20X/0.4, LEICA 566072 N PLAN EPI 50X/0.75, and LEICA 566073 N PLAN EPI 100X/0.85. Honestly I cannot find any useful information from these information...Do you have any clues?

To the best of my knowledge, there should be no angle or polarization dependence of the Raman-shifted light...

The Leica database is partially up and running today- still no spectral transmission curves- so I was only able to get partial information. It appears that the objectives require use of a #0 coverslip (as opposed to no coverslip or a #1.5 coverslip)- are you using those? Use of the proper coverslip is important because it is an optical element and the objective lens designers take it into account for the design of the lens. Incorrect coverslip thickness leads to uncorrected spherical aberration, which could account for your results.

I hesitate to create additional work, but if you are able to use 1 objective and 2 or 3 different coverslip thicknesses (say #0, #1.5, and no coverslip) you may see a difference.