Coupling a pulsed laser into an upright microscope for um scale structuring of 2D materials

I’m adapting a Nikon LV150 upright microscope to structure exfoliated 2D materials using a pulsed laser. Since I’m not so familiar with optics, I would like to ask for help on the hardware.

Does anyone know of a proper way to couple a fiber laser into Nikon LV150 so that I can get ~1um spotsize?

I have a Thorlabs ns pulsed laser (520nm, 50kHz repetition, ) coupled to a multimode optical fiber connected to a collimator which shines into a beamsplitter reflecting most of the light down into the optical path into the objective.

We lose a lot of intensity before and after the microscope and therefore, eventhough in principle, this laser should be able to ablate graphene, the laser does no damage to it.

The collimators we bought from Thorlabs still result in slightly diverging beams. I think as a result of this we get a badly shaped “spot” of ~10-20 um in diameter. Does anyone know how I can make the collimation better?

We are now considering to buy a more powerful laser. Does anyone know the damage threshold for the optical components in Nikon Microscopes? Will objectives be okay with 1uJ pulses? Any recommendations for a cheap high powered pulsed laser?

These are the parts I bought


Thank you

Can you provide a quick sketch / diagram of the optical layout? Also, what objective are you using on the microscope?

1 Like

I was going to suggest using a single mode fiber instead as I believe that one thing that is limiting your ability to focus the laser to the spot size you want is the multi-mode fiber. However the beam quality of this laser looks to be quite poor (NPL52C) so you will likely lose a lot of power when trying to couple it into such a fiber.

I agree that an optical layout would be great, from your description it seems you reach the graphene while collimated… why not focalise to concentrate power ?

For collimation it is ‘bad’ at 10-20µm because any beam of small size will quickly diverge. But I don’t think you want collimation ? You need a small size to reach high irradiance and ablate.

As a note, you are rightly worried about spot size and you should check what is achievable with your Numerical aperture, but also check for pulse broadening. In my experience dielectric mirrors tends to broaden the beam and it is useful to switch to silver coated ones.

Thank you Hazen_Babcock and Adrien_Mau for the comments.

I have attached a simple schematic of how I envision this setup would work. One problem I have is that we seem to lose quite a bit of power between the beamsplitter (we also tested a silver mirror) and the objective. The Nikon microscope is a black box for us and we cannot get any information from the Nikon representative. I’ve sent him 4-5 emails since last November after he sold us a wrong component - not to blame him for making so many mistakes with our orders in the past few years but just to ask about the internal optics of the microscope - and we haven’t received a single email back. We don’t know the wavelength dependence of the absorption/ transmission nor the damaging threshold of the internal components nor the objectives.

In the past few weeks, we started looking into different pulsed lasers on the market to replace the NPL52C. We have also placed an order for a single mode fiber. The multimode fiber was not suitable for 520 nm to begin with. Hopefully the new fiber can make a difference in the shape of the beam.

As for the spot size, the theoretical diffraction limit of our laser together with the 0.6NA objective should be <1 um but we are seeing ~20 um spots. I am thinking that the collimation is not perfectly done before reaching the objective, therefore the beam does not focus into the smallest spot. The collimators I bought from Thorlabs have not been collimating well. They seem to have a rather easily noticeable divergence. I am thinking that the collimation should be done perfectly so that the beam reaches the objective in a parallel path.

Am I on the right track here? Do I need to add a pair of lenses in between the collimator and the beamsplitter to correct for this divergence? or do I also need to as a pinhole?


You will only be able to achieve a diffraction limited spot with a coherent Gaussian beam. My understanding is that currently you are using a multimode fiber? In this case I would expect that the best you would be able to do would be roughly proportional to the average mode index. A document with examples of the different kinds of modes you might see from such a fiber are here: rp-photonics..

1 Like

Another issue you may be having is that the laser may be focused in a different plane than the camera. Ideally you would be able to adjust the collimator to change the beam divergence so you could search for the setting that gave you the smallest spot.

Thank you for the information! We didn’t have a clue about fiber optics, so we just went with the fiber with the highest power rating, but it seems that we made a mistake in choosing a multimode fiber over single mode. The single mode fiber will arrive this week, so we will give that a try.

However, we also did try shining the laser without the multimode fiber and the collimator. The laser from Thorlabs is quite compact, so it was possible to hold it in place right beside the beam splitter. The beam diameter coming out from the laser is a few mm but is better collimated than with the fiber+collimator. Unfortunately the spot on the sample wasn’t so well defined still…

Apologies for missing this. In my opinion yes a pair of lenses will help. You want the distance between them to be adjustable so that you can control the divergence and find the setting that gives you the smallest spot. A (correctly sized) pinhole will also clean up the laser beam and give you an even smaller spot, but it will reduce the total power. Using a single mode fiber will have the same affect. I think that with the laser you have you will have a hard time coupling a significant fraction of the laser output into a single mode fiber.

1 Like

Could you check the shape of the laser at its direct output, before the objective and after the objective ? You may control the size via a lens pair but any way the shape should always look gaussian-ish, if else there may be an issue with the objective or the optical alignment.

images of the laser beam and of the focused spot at sample plane could also guide us toward specific issues