Hello everyone,
Ive recently been tasked with building an inverted objective-tirf microscope. Im rather new to this so im learning by doing. I have built the system and aligned the laser, as far as I know. After aligning everything and connecting the cameras and hardware, I decided to do some test images. However, these images are a little strange. Perhaps you could help me out. The images were taken in the epifluorescence position (e. g the laser shining directly through the sample) with just a solution of fluorophores and with a solution of tetra speck beads. When imaged I just get a blob, which is expected for the fluorophore solution but not for the tetraspeck. theres also a strange pattern that I can’t quite identify even after messing with the optics more. Ive uploaded the images and a rough sketch of the setup. I appreciate any input.
Is this fully custom microscope? Or was it built off a standard microscope body, such as an Olympus or Nikon inverted microscope?
My guess is that what you are seeing is scattered light that is passing through your fluorescence filters, but it’s really hard to say from the image you show. What are the different filters and dichroics you are using? For example “Chroma AT435lp”.
hey thanks for the response. It’s full homebuilt. I have this one at the intersection of the emission and excitation path ( zt375/473/532/635rpc). I have these in the excitation path to bring the lasers in the path. And I took the rest along the excitation path away, to test just a single laser.
now, that I think about it, that could be the problem, That I didnt have any downstream of the intersection one. 
Yes, the emission filter is very important. A dichroic alone does not block a sufficient amount of the excitation light.
okay so after I added the necessary emission filter, some of the scattering goes down. Now the main problem is that I still see the laser beam in the image ( a bright dot). What do you think is the cause of this? Is this a problem with the alignment through the objective (BFP)?
What is the OD of your emission filter at the wavelength of the excitation laser?
It may be an alignment issue, but if the filters are good and the alignment isn’t too far off I think you should be fine. Does it look like the beam is more or less focused in the center of the back focal plane? Does the beam come through the objective reasonably collimated and on the optical axis?
The emission filter is the HC 685/40 (OD>5). Im using tetraspeck beads to test the system. I have a 642 nm Laser. The beam comes out as it should. I get the nice pattern on the ceiling indicating the alignment on the BFP.
How bright is the bright dot relative to the tetraspeck beads (what size beads?). What kind of objective are you using, for example 100x oil immersion?
Maybe an excitation filter in front of the 642nm laser will help? Some lasers emit at other colors, though this is usually more of a problem with gas lasers.
Are you sure there are no other sources confounding the measurement? All the other lasers are either off or manually shuttered? You don’t have a focus lock laser?
Looking again at your schematic I would suggest changing the AOTF so that you use the deflected beam and dump the non-deflected beam. This should give you a much larger dynamic range.
Hi
I think i can help you with this, you have an ensemble of things to check
1.What tetraspeck beads are you using ? Are they adapted for 640nm excitation ? How do you prepare sample ? With White light you should already be able to see them and check sample and focus.
2. How do you move objective for focus on this home made microscope ?
3. Also what is your objective NA ? You probably know that you cannot do TIRF if the NA is less than the refractive index of sample
4. What is laser power at sample plane ?
5. Is the laser focused at back focal plane of objective and collimated at sample plane (in EPI… in TIRF you cannot check collimation). How do you perform this ? How do you choose the position in back focal plane ? You mention a spot for laser at sample but this should not be a spot but a wide gaussian.
.
hey thanks for your resonse. I will address the comments in order.
- the tetra speck beads are the 1 and 0.1 um with emissions at 4 wavelengths. For quick checks I simply dilute some of the beads in pbs. Nothing fancy, just to check the work real quick. I have, meanwhile observed the beads. However, other problems have arisen.
- The objective itself does not move. It is an inverted tirf setup. I can move the sample via the piezo nano stages.
- im using a tirf objective from nikon with a NA of 1.49
- I have not measured the laser power at the sample plane.
The laser is focused at the back focal plane. I aligned this by shining the laser directly through the objective and onto the ceiling. - I fixed the spot on the laser. And now, at least in epi, I can see the beads, but only in a certain region of the image. I can upload an example. Theres also weird patterns occurring around dirt, but I think I know the cause. I’ll upload an example image shortly .
Thanks for the feedback, from the fact that you can see the beads fluorescence in EPI system components and alignment must be corrects.
- ok
- is cool
- ok with Nikon oil and such
- ok for power if you see in EPI. Alignment procedure is enough but you can think about adding a removable lens in front of the camera, separated from the sensor by its focal distance. This will image the Back Focal Plane of the objective and monitor TIRF alignment.
- is not surprising if you have a small gaussian beam. Usually people use a wide gaussian to approximate a wide flat top profile (but some have more sophisticated technologies like ASTER).
Weird pattern is difficult to diagnose without image. can be dirt somewhere.
How do you plan to switch from EPI to TIRF ? You may look at the ASTER publication for this
beads only ever show up on the right side of the imag. something im not really sure about. Ive tried messing with the orientation of the sample to see if that was the problem, but it wasnt. So it seems like it is some problem with one of the optical components. Im troubleshooting a bit today since I have some downtime from experiments. Id love to hear any ideas though.
An alignment protocol that I use:
-
Focus on a sparse bead sample.
-
Remove the emission filter and the last lens in the excitation path (the lens that is focusing your collimated excitation beam onto the back aperture of the microscope objective) and reduce the laser power. You should now see a single very bright spot at focus and rings as you focus up and down. Use the alignment mirrors in the excitation path to adjust the spot so that it is centered and the rings are spatially uniform when defocused (spatially uniform = means the brightness around the ring is roughly even, one side of the ring is not obviously brighter than the other side). If you have an adjustable aperture in the excitation path close it down some and repeat. Closing the aperture down should increase your sensitivity to how spatially uniform the rings are).
-
Remove the objective and the sample and mark where on the wall / ceiling the beam hits. Then replace the objective and the sample. You only need to do this once, it isn’t necessary to redo this for every alignment.
-
Replace the emission filter and the lens and reopen the excitation aperture. Adjust the lens position in Z to create the smallest spot that you can at some distance after the beam exits the microscope objective. Adjust the lens X/Y position so that the beam is centered on the mark you created at step 3.
Your microscope will now hopefully perform as expected.
Ha nice alignment protocol through the Back Focal Plane for 2.
For 4. the lens you are referring to is the last excitation lens before objective ? Indeed you need to set the objective at the general sample focus height and then move axially the last excitation lens so that you focalize the beam in BFP (small spot at ceiling)
I think @mediocremicroscoper maybe one thing lacking for me is the general understanding of your system, what part you can use to align and the purpose of lenses. Usually a tube lens (~200mm focal) is present and there is a relay to the camera but i just noticed on your scheme that there is no such thing. The objective and the final lens in front of each camera acts as relay ?
It could be helpful to double check that you provide us with distance and focal lengths of your elements. Maybe indicate where/when there is an intermediate image plane.
For your imaging that only works on one side… maybe you could use a more general sample such as calibration USAF targets and look at it in white light. You would be able to look at the light traveling through the detection path and notice potential cropping or unusual effects.