Converging Lasers of Different Wavelengths

[Yondaime5685]

I hope this is the right place to ask this. It is an experiment question.

I have a 785 nm laser and a 1550 nm laser and I need to converge them down to a point. Now I could just manually align the lasers on the same XY-plane and have them hit that spot, but for reasons I don't want to explain, I need to converge them down with a lens.

So far I have the two lasers parallel and really close together. (Beams separated ~2-3 cm) They go though a spherical plano-convex lens and then hit a detector card. Now I know the focal points won't match due to the difference in wavelength, but is there any combination with this setup that I can get them to converge to the same point.

One thing I was thinking was letting the 1550 laser go straight through the center and have the 785 laser refract and meet up at the center.

Any input on this will be very helpful?

(The above image shows that I can make them aligned horizontally, but getting them to align vertically is killing me. Also this diagram is very crude.)

 

[UltrafastPED]

If you have the space try using two long focal length lenses and a dichroic mirror (lookup "cold mirror"); don't bother aligning the two laser beams. Calculate the effective focal length for each of your lenses - be sure to check that the transmission is good enough for your wavelengths - and then "converge" the beams at the dichroic mirror.

The mirror will be at a 45 degree angle with both beams; the long wavelength light will simply pass thru the mirror, seeing it like a flat plate of transparent glass. The shorter wavelength will be turned 90 degrees. If you use some diaphragms to assist with your alignment, and perhaps a dogleg for your local up/down, left/right alignment controls ... then you should be able to get the two beams perfectly aligned, and focused to the same spot.

If you require a smaller spot size you will have to shorten the working distance.

Another technique I have seen is to split a lens - cut it with a diamond saw (e.g., at the geology department), and position the two halves at different distances so that the focal point is the same. Then one beam is coming in high, the other low ... or left and right. You may even find that giving them a slight tilt is useful for converging the focal points. Once again the doglegs and some iris diaphragms will make your adjustments easier.

Dogleg: two mirrors in the beam line, each mirror turns the beam 90 degrees. Use the first for horizontal adjustments, the second for vertical. Use the iris diaphragms (a) the first one to generate Poisson rings in the beam, the second one, as far away as possible, as your alignment target.

When I setup a new optical system I always start with a dogleg for control, then an iris diaphragm nearby, then a second iris diaphragm as far downstream as possible. Close them down as needed for alignment, then open them back up for your actual work. With multiple beams, more doglegs & iris diaphragms. Makes things go faster, and more repeatable.

 

[Cthugha]

Also, for good results insert the lens the other way round than drawn in the picture. You usually get the best results following the "parallel to curve" rule. Place the convexside of the lens next to the parallel beam and the "plano" side towards the focus.

Besides that, I fully agree that the dogleg and iris method already mentioned definitely is the gold standard.

 

[Yondaime5685]

-UltrafastPED
Thanks for that info. I'm starting to do Optics for my research and I'm slowly getting my feet wet. The "dogleg" setup will hugely help me. The idea is so simple, but effective. I don't know about cutting the lens in half, my adviser might kill me, but I'll give your other suggestion a try.

Some clarification questions:
A beamsplitter has a dichroic mirror right? So I can use that instead?
When you say converge the beams at the dichroic mirror, you mean just let the beams hit the mirror or they have to overlap at the interface of the mirror?
Also after the beams pass through the mirror, you say that it "should be able to get the two beams perfectly aligned and focused" do you mean with the plano-convex lens?

Sorry for the noobish questions, but I really appreciate it.

-Chthuga
I see. Thanks!

 

[UltrafastPED]

I don't know about cutting the lens in half,

Most lenses used with modern lasers cost $60-100; thus cutting them up can be a worthwhile expense. I've seen this done a number of times, along with drilling holes through a lens or mirror, or just shaving off a portion to make room for optical clearance.

A beamsplitter has a dichroic mirror right?

There are many types of beam splitters; you might spend some time looking at all of the offerings of one of the major vendors, trying to understand how each one works, and when to use it.

For example, some beam splitters divide the beam by polarization state; some divide the beam by intensity - and not necessarily 50-50!

For example, see: https://www.newport.com/Optical-Beamsplitter-Selection-Guide/994059/1033/content.aspx


If you need to divide by wavelength - then you need a dichroic mirror. This has a special multi-layer dielectric coating - similar in some ways to anti-reflection coatings - and you must choose the dichroic mirror that has the characteristics which you require.

See http://search.newport.com/?q=dichroic mirror
These are all "turning mirrors"; you want one which can turn the one wavelength while transmitting the other wavelength. Depending upon requirements these can be quite expensive, especially if they must be custom designed and manufactured.

When you say converge the beams at the dichroic mirror, you mean just let the beams hit the mirror or they have to overlap at the interface of the mirror?

You should make a drawing of your planned setup, scaled to the physically available space. Then you can hopefully see what is the best approach. Your beams don't need to actually "converge" until they reach the focal point ... but they will be "converging" on the way there. Since they must both strike the dichroic mirror, at 90 degree angles to each other, you have limited freedom in where the dichroic mirror can be placed.

Also after the beams pass through the mirror, you say that it "should be able to get the two beams perfectly aligned and focused" do you mean with the plano-convex lens?

You want to do the beam alignment with the mirrors in the doglegs; if that is not enough, you can build another dogleg towards the end of the system, perhaps including the dichroic mirror as one piece.

In one of my setups where the second and third harmonics were being generated used two dichroic mirrors arranged as a dogleg for the final alignment of the focal spot - the focusing lens had a 2 meter focal length, and the converging beam went through the doubling and tripling crystals, then the dichroic mirror dogleg, and finally passed through a sapphire window (so that the UV could pass), on into a vacuum chamber ... where the final, 250 nm laser pulses generated electron pulses via the photoelectric effect.

PS: I'm guessing that your laser beams are CW, and not pulsed. Most of my work is with ultrafast lasers, but the general ideas are all the same. The safety procedures vary with the laser, but must always be considered first, and not as an after-thought.