Optimizing Focal Plane Alignment for Optical Tweezers: A Guide for Researchers

[kubikat]

I have a question concerning optical tweezers.
How do you make sure that the focal plane of the beam and the laser are the same..?
In other words, I can see the laser on my CCD image when it is reflected off the cover slide, but if I see the reflection of the laser then the beads that are stuck to the cover slide are not in focus... If the beads are in focus then I can not see the laser...
I have read papers on it and they all mention that issue and just say that it should be resolved... but they don't say how. I do realize that it is probably something trivial, but I just can not figure it out. Any hints will be greatly appreciated.

 

[Andy Resnick]

It's not clear what your setup is- what's the wavelength of the laser, for example?

Are you asking how to ensure the trapped object is in focus? It's also important to remember that the trapped object is not centered on the focal spot, but slightly "downstream" from it.

 

[kubikat]

Thank you so much for your response (:

Here is the sketch of the setup (the imaging part)... apart from it I have 2 lenses that expand the beam and another mirror to move the beam.

The wavelength of the laser I use is 750 nm, the dielectric mirror reflects most of it, but let's some through so I can still observe it on the CCD screen, when it is reflected off the slide.
I can get clear pictures of the spheres and the beam separately, but not both... And i want to know how to make the planes of the laser focus and the plane of the microscope focus coincide?

 

[Andy Resnick]

I'm not trying to be difficult, I'm trying to get a clear picture:

From what I understand, you are trapping spheres (plastic?) in a fluid solution, and you want to simultaneously image the trapped sphere and the trapping beam. You are currently able to image one or the other, but not both.

On the systems I've built, if I bring the focus of the trap into coincidence with the focal plane of the image (by, for example, imaging both the coverslip and the backscattered laser light), then the trapped particle is not in focus.

But the only reason I need to image the laser is for alignment- once the tweezers are aligned to the microscope, I don't image the beam anymore- so I'm a little confused by your question. I'm not interested in the laser focus being in the object plane, I'm interested in the trapped object being held at the object plane.

In any case, if you want to adjust the axial position of the beam focus, you can 'detune' the beam expander to add a bit of divergence (or convergence) to the trapping beam prior to it entering the objective- if you are using an infinity corrected objective lens, this won't add too much aberration to the trapping beam. But, if your optics are misaligned, you can introduce a lot of off-axis aberrations which will rapidly deteriorate the quality of the trap.

Does this help?

 

[kubikat]

Yes I am trying to trap 3micrometer polystyrene spheres in the fluid. I think I understand that I will not be able to see both beam and the sphere ( since I can only see the laser beam when it is reflected off the cover slide, or slide cover or anywhere where the index of refraction is changing). I should be able to see something when I trap the sphere because of refractions... if I am thinking about it correctly...

I think I understand what you mean... since the trapped particles are not exactly at the same plane as the focal point, you won't be able to see them, so they do not need to coincide... I think my main question was how does one change the focal plane of the laser beam and your last paragraph about "detuning" the beam expander answered this question (:

My microscope objective is not infinity corrected though... I think it is the one where the intermediate image is formed 160mm behind the objective, so I will have to align everything very carefully.

Thank you so much for answering my questions. We don't have grad students in our school and none of my professors done any optics since their undergrad, so you are helping me so much. Could I ask you a few more questions?

 

[Andy Resnick]

Could I ask you a few more questions?

Of course! I may be able to help with alignment techniques/tricks as well.

 

[kubikat]

Oh thank you so much...

The first question I have is the following...

On the picture below I drew two ways I had my telescope set up. In the first case the laser light was first reflected off a mirror ( which seems to be first surface and has been cleaned and I also tried a few more first surface mirrors) and then going through 2 lenses. After the focal point of the first lens, the pattern was produced ( Looks like Airy disc pattern, which is due to Fraunhofer diffraction)

Then at some point I needed a post, so I figured since the mirror wasn't really "doing" anything ( I use 2 other mirrors for alignment of the beam)... so I took the mirror out. The airy pattern was gone then and replaced by a pattern of concentric rings, which all seem to be of equal intensity.

Have you ever come across something like this?

 

[Andy Resnick]

What you describe is very similar to spherical aberration, but it's not clear what changed from (1) to (2). I suppose if the distance from the laser to the first lens decreased, you would tend to notice the effect more since the beam uses less and less of the available lens power. Is the source a diode laser? Has the output beam been circularized or anything? What does the raw beam look like?

Something to try- if the lenses are plano-convex, put the curved face of lens 1 facing toward the laser (and the curved face of lens 2 should face the objective), see if that helps.

 

[kubikat]

The beam is an 80 mw diode laser... pretty cheap one. I think the diode only cost $30... It is supposed to give out an elliptical beam.
I have it collimated and the beam looks pretty circular (to within you can't tell that it's not circular) and about 1.5-2 mm in diameter. I was going to use 2 prisms to make the beam circular in case the trap was not working.

It seems to me that the mirror is to blame for producing an Airy disc, since it kind of acts like a slit (maybe)... but mirrors technically shouldn't do anything to the beam except changing its direction.

I do not completely understand why does a beam use less lens power if it is closer to it...

Also the lenses I use are biconvex. I read recently that the plano convex are better with aberrations but it was much after we have purchased the biconvex ones.

 

[Redbelly98]

Hmm, since the beam diameter (1.5-2 mm) is a lot smaller than f=50 mm, I wouldn't expect much spherical aberration. But maybe the fact that they are biconvex lenses makes for significant aberrations after all.

Some other possibilities:
1. Are you sure you are using the front surface of the mirror?
2. What is the total beam path length from the laser to the first lens, in setups 1 and 2? I.e. are they the same?
3. Are you able to adjust the lens position so that the beam goes right through the lens's center?
4. Are the lenses AR-coated for your wavelength or uncoated?

Another thing to try -- can you look at the beam's reflection from a convex surface? (No lenses, just a convex mirror, located the same distance from the laser as the f=50mm lens.) That will expand the beam, and you can see if the beam pattern more closely resembles setup 1 or setup 2.

If you don't have a convex mirror, you can try looking at the beam reflected from a curved convex lens surface, preferably one that is uncoated or at least not coated for 750 nm. Lacking that, even an AR-coated mirror usually reflects 1/4 to 1/2 %, if you have the room lights off you might be able to see the reflected beam even then.

 

[Andy Resnick]

The beam is an 80 mw diode laser... pretty cheap one. I think the diode only cost $30... It is supposed to give out an elliptical beam.
I have it collimated and the beam looks pretty circular (to within you can't tell that it's not circular) and about 1.5-2 mm in diameter. I was going to use 2 prisms to make the beam circular in case the trap was not working.

Is your collimator part of the diagram you showed in post #7, or is this an additional component? But otherwise, this all seems reasonable.

It seems to me that the mirror is to blame for producing an Airy disc, since it kind of acts like a slit (maybe)... but mirrors technically shouldn't do anything to the beam except changing its direction.

I do not completely understand why does a beam use less lens power if it is closer to it...

You re correct- the mirror should not do anything other than redirect the beam. What I meant as far as your question refers to the fact that the refractive power (and amount of aberrations) of a lens depends in part on how much of the lens is illuminated, since the effective f-number changes. Laser beams have a divergence, so by moving the laser closer to the lens, you are operating at a higher f/-number: normally it's not a problem, but illuminating a 50 mm focal length lens with a 2mm diameter beam means the lens is operating at f/25, which is pretty high, but not awful.

To compare, my beam expander uses a 10 mm / 230 mm telescope (IIRC), expanding a Nd:YAG raw beam (0.6 mm diameter).

Also the lenses I use are biconvex. I read recently that the plano convex are better with aberrations but it was much after we have purchased the biconvex ones.

Plano-convex (and plano-concave) are better than biconvex/bicocave if they are oriented properly, but you should be able to get your system to work ok.

Another thing to try -- can you look at the beam's reflection from a convex surface?

This is an excellent idea- see if you can locate the back-reflection off the lens surface. You can get a lot of clues for troubleshooting with this technique.

Back-reflection is the #1 way to perform the initial alignment of optical elements. If you intentionally decenter the first lens, you can (usually) more easily find the back-reflected beam. Then, center the lens by colocating the reflected and incident beams, and perform any needed tip-tilt adjustments as well.

 

[Redbelly98]

Back-reflection is the #1 way to perform the initial alignment of optical elements. If you intentionally decenter the first lens, you can (usually) more easily find the back-reflected beam. Then, center the lens by colocating the reflected and incident beams, and perform any needed tip-tilt adjustments as well.

Note that there will be two back-reflected beams, one from each lens surface. Both need to be colocated with the incident beam.

The two reflections will both diverge, and with different divergence angles. So they will be easier to find the closer you are to the lens, where the beams are smallest and brightest.

kubikat, just to make sure of something, are you using an IR viewer *** to see the laser beam? To help with seeing faint IR beams, it generally helps to turn off the room lights, and also use an IR-pass filter to eliminate stray visible room light from the viewed image.

***Note: I mean a viewer along the lines of the http://www.findrscope.com/SPD/infrared-viewer-laser-kit-85268a-ir-laser-package-1350nm--12F0000-1149959918.jsp" might also work.

 

[kubikat]

Thank you both again very very much. I didn't expect to receive that much useful information and advice!

I think my mirror is the front surface. I have seen both front surface ones and non front surface ones and and the ones I tried look like front surface. Is there a way to check it?

The total beam path length in the setups 1 and 2 are not the same but don't differ by much (maybe a 30 cm)... I have tried putting a mirror right next to the laser and I still get the pattern.

I did adjust the lens position so the beam goes straight through the center, but I am going to try looking at reflections to ensure that.

My lenses are AR coated I think. I was thinking about buying some more lenses (maybe getting the plano-convex ones) and uncoated ones are a lot cheaper... would they still work for what I am doing? I mean the laser I am using is 80 mW and I read that 10 mW is usually enough for trapping, so I can afford losing a lot of power.

And I am not using the IR-viewer (we don't have one)... I use a card. Or just turn the power of the beam down and look at it with naked eye (since the beam is only 780 nm I can see it and looking at it with naked eye makes alignment a lot easier)

 

[Redbelly98]

Important question: (!)

Or just turn the power of the beam down and look at it with naked eye (since the beam is only 780 nm I can see it and looking at it with naked eye makes alignment a lot easier)

Do you mean that the beam hits a sheet of white paper or a card, and you are looking at the spot on the card where the beam hits it?

 

[kubikat]

Yea, I look at the beam when it hits a white piece of paper. It is red. Since I turn the power down to 2-3 mW it shouldn't be harmful right. Usually when I am working at the high powers I use glasses.

 

[Redbelly98]

Yea, I look at the beam when it hits a white piece of paper. It is red. Since I turn the power down to 2-3 mW it shouldn't be harmful right. Usually when I am working at the high powers I use glasses.

Okay, good. I was fearful you might be doing something really really wrong when you said "look at it with naked eye", I am glad you did not mean what I was thinking.

I think my mirror is the front surface. I have seen both front surface ones and non front surface ones and and the ones I tried look like front surface. Is there a way to check it?

Yes. Look at the reflected beam at 45° angle of incidence. If the reflector is on the back side then you'll see two extra reflected beams from the front surface, weaker than the main beam but noticeable, one on either side of the main reflected beam.

If the reflector is on the front side, you shouldn't see the two extra reflected beams ... but if you do see extra beams, they'll all be on the same side of the main reflected beam.

The total beam path length in the setups 1 and 2 are not the same but don't differ by much (maybe a 30 cm)... I have tried putting a mirror right next to the laser and I still get the pattern.

Well, if the beam paths are 5 cm and 35 cm, that's a big difference. But if they're 200 cm and 235 cm, then you're right it's not a big difference.

I did adjust the lens position so the beam goes straight through the center, but I am going to try looking at reflections to ensure that.

It's good that you have extra power to spare in terms of being able to see the reflected beams.

My lenses are AR coated I think. I was thinking about buying some more lenses (maybe getting the plano-convex ones) and uncoated ones are a lot cheaper... would they still work for what I am doing? I mean the laser I am using is 80 mW and I read that 10 mW is usually enough for trapping, so I can afford losing a lot of power.

Hmmm, I'll defer that question to Andy... but it sounds like you could be okay with uncoated ones, if your project budget is tight.

 

[Andy Resnick]

My lenses are AR coated I think. I was thinking about buying some more lenses (maybe getting the plano-convex ones) and uncoated ones are a lot cheaper... would they still work for what I am doing? I mean the laser I am using is 80 mW and I read that 10 mW is usually enough for trapping, so I can afford losing a lot of power.

And I am not using the IR-viewer (we don't have one)... I use a card. Or just turn the power of the beam down and look at it with naked eye (since the beam is only 780 nm I can see it and looking at it with naked eye makes alignment a lot easier)

It's good that you have extra power to spare in terms of being able to see the reflected beams.

Hmmm, I'll defer that question to Andy... but it sounds like you could be okay with uncoated ones, if your project budget is tight.

I would use AR coated lenses, for a couple of reasons:

1) Since you can adjust the power output of the diode directly, you can always dial down the trap power that way. Using AR lenses will increase the maximum trap force, which can translate into the ability to trap smaller objects, or objects with a smaller index mismatch.

2) The microscope objective was (most likely) not designed to image 780 nm light, and so the lens aberrations and the transmission will be worse than the visible.

Personally, I can't see 780 nm light directly; I'd have to use an IR viewer (mine's a View-It, genetically identical to the Find-R-Scope). So I'd always run the laser at full power and wear glasses. Plus, I'm not very familiar with your diode source, but I wonder if the beam characteristics change when you dial down so much.

Then, of course, there's the issue of what happens to the scattered light (off a non AR lens)- the light has to go somewhere. My systems have been built on upright microscopes, which places the beam line directly at eye level. This does not make the safety people happy. Again, 780 nm is on the edge of visibility, so you (or someone else using your system) may not have a blink reflex, which could translate into eye damage.

One bit of advice on alignment technique- *always* align to the laser beam. Assuming you are not using a microscope, the alignment steps proceed: shine beam directly onto the sample, then place lens #1 in the path, center and adjust tip/tilt. Then add lens #2, repeat the alignment steps, the add the microscope objective. Once that has all be done, perform final alignment by imaging the focused laser (I replace the samples with a mirror) and adjust first the objective, then lens #2, then lens #1 to minimize the various aberrations.

If you are using a microscope, then first align the beam to the microscope optical axis, then proceed as above. If there are mirrors, align them all prior to adding any lenses. I perform my final alignment by adjusting the mirror tip/tilt, rather than the lenses.

I guess, after all this discussion, I am wondering why the Airy pattern was present at the focus of lens #1 in the first place: can you reproduce the original setup and verify that?

 

[Redbelly98]

I guess, after all this discussion, I am wondering why the Airy pattern was present at the focus of lens #1 in the first place: can you reproduce the original setup and verify that?

I think the pattern was viewed after the focus, not at it? From Post #7:

After the focal point of the first lens, the pattern was produced

 

[kubikat]

Andy, that's strange that we can see the beam then. Maybe our diode is not 780 nm after all :/ That's what the specs say though... (By the way I did not find out that I can see the beam until one day I forgot to put the safety glasses on...) I always wore the glasses and used to use cheap cell phone camera to view the beam until we got an IR card.

Thank you for the advice about the alignment! I used to align step by step first lens, then the second lens etc...

Beam paths do not differ as drastically as 5 and 35... but still pretty close to that.
I will reproduce both setups tomorrow with the same beam path, when I am at school and take the pictures of the beams. (my paint job if drawing them is not really good)... Maybe then it will look like something you've seen before. And I will also check my mirror to see if it's really first surface.

Also I have one other question: Do you think it matters much that my 45 degree dielectric mirror is not at 45 degrees? This way I get more of the image of the sample of the screen and also am able to see the ray reflected off the slide cover on my CCD camera (since some of it is transmitted)

Here is an image to show what I mean.

Oh an also, does the beam have to be single mode, because mine isn't? And I know that converting it to a single mode will result in quite lot of power lost. Some papers I read talk about using a fiber, to convert it to single mode and I built a single mode laser before by using a diffraction grating. Some papers say that multimode laser beams result is bad aberrations, however I was told by some people who have built a working tweezer appartus that it is not necessary and it works just fine with multimode.

Thank you so much for all the help. I am printing out the whole thread and taking it to school with me tomorrow to make sure I don't forget anything :)

 

[kubikat]

Oh yes and the pattern was viewed after the focus (it might have been there before, just was too tiny to see)

 

[kubikat]

And I did figure out the mirror had nothing to do with the pattern :D it was all in the path length difference... Thank you ! I still don't know why the path length difference matters so much but it's nice to narrow it down :)

 

[Andy Resnick]

I think the pattern was viewed after the focus, not at it? From Post #7:

Good question- I was assuming both patterns were obtained at the plane of 'best focus'.

 

[Andy Resnick]

Oh lordy... you just opened up a whole new can o' worms...

Andy, that's strange that we can see the beam then. Maybe our diode is not 780 nm after all :/ That's what the specs say though... (By the way I did not find out that I can see the beam until one day I forgot to put the safety glasses on...) I always wore the glasses and used to use cheap cell phone camera to view the beam until we got an IR card.

<snip>

Also I have one other question: Do you think it matters much that my 45 degree dielectric mirror is not at 45 degrees?

<snip>

Oh an also, does the beam have to be single mode, because mine isn't? And I know that converting it to a single mode will result in quite lot of power lost. Some papers I read talk about using a fiber, to convert it to single mode and I built a single mode laser before by using a diffraction grating. Some papers say that multimode laser beams result is bad aberrations, however I was told by some people who have built a working tweezer appartus that it is not necessary and it works just fine with multimode.

Oh yes and the pattern was viewed after the focus (it might have been there before, just was too tiny to see)

And I did figure out the mirror had nothing to do with the pattern :D it was all in the path length difference... Thank you ! I still don't know why the path length difference matters so much but it's nice to narrow it down :)

I'm not even sure where to start... Dielectric mirrors are almost always designed to reflect a given wavelength having a particular polarization at a particular angle of incidence. Going off-nominal will always decrease the reflected power.

If your laser is not single mode, you will *never* see an Airy pattern. Honestly, I am convinced that the only way to get a stable 3D trap is to use a single mode beam. If all you are doing is pushing particles against a glass surface, then yes- a crappy beam will still give you some decent 2-D confinement. But for a full 3D trap, you simply must have steep intensity gradients, and this will only be possible with a single-mode beam.

At this point, I'll await some images of your setup, beam spot, etc. Maybe it's worth sharing some images of my current setup?

 

[kubikat]

My setup is very simple ( it is inexpensive for one and then my professor wants me to set it up and write a manual so students in the advanced physics course would be able to set it up themselves --- so it has to be simple). It is all in one plane - horizontal. I personally think that the vertical would work better, but horizontal is simpler and more safe (the beam never leaves the plane of the table). In any case if you could post the images of your setup that would be awesome. I will post mine in a few posts after this one.

 

[kubikat]

Here is the image of my whole setup. I am using a diode laser. 100x oil immersion objective and the telescope lenses I currently use are 50mm and 250mm focal length.

And here is where I am trying to trap them. I am planning on making sort of a flow cell once I can actually trap something to measure how strong the trap is.

 

[kubikat]

And here is the beam pattern that I see when beam comes out of the laser and hits the first lens after the focal point (this one is when the path length of the beam is rather short). I am not sure why I get the rings.

Here is how the beam profile changes as it travels longer distances. At certain distance I can make the "hole" in the center disappear and then the center of the beam seems to be gaussian.

Thank you a lot for your help. I know it is a lot of work looking over this kind of stuff. And Andy, the images of your setup would certainly help. Also if you could please send me your paper on the optical tweezers from space station. ( I am not sure if that is an appropriate request, since I tried downloading it online but every website I visited required me to pay for it.) My school library might have it, so if you are not able to send it to me, I should be able to get it there. Thank you again!

 

[kubikat]

Also I am sorry the pictures of the beam are so huge. I tried making them smaller on the host website and then add them in here again. but they wouldn't change

 

[Redbelly98]

And here is the beam pattern that I see when beam comes out of the laser and hits the first lens after the focal point (this one is when the path length of the beam is rather short). I am not sure why I get the rings.

Here is how the beam profile changes as it travels longer distances. At certain distance I can make the "hole" in the center disappear and then the center of the beam seems to be gaussian.

This is making more sense now, I'm pretty sure this is a basic diffraction effect.

It looks like part of the circularization process built into the laser is to send the beam through a circular aperture. At shorter distances, the beam looks pretty much like the aperture, though rings are present to some extent. At longer distances, the diffraction effects become more pronounced, with the center going through bright and (nearly) completely dark regions.

 

[kubikat]

Oh okay, so that's just diffraction? I do send a beam through a small circular lens that I put into the diode module for collimation. Where shall I keep the telescope then? At the position that gives me a more pronounced diffraction pattern (far away from the laser) or near the laser where I get the ring pattern? I think further away would work better since the central maximum is (should be) Gaussian. But I am not sure.

 

[Andy Resnick]

Thank you a lot for your help. I know it is a lot of work looking over this kind of stuff. And Andy, the images of your setup would certainly help. Also if you could please send me your paper on the optical tweezers from space station. ( I am not sure if that is an appropriate request, since I tried downloading it online but every website I visited required me to pay for it.) My school library might have it, so if you are not able to send it to me, I should be able to get it there. Thank you again!

These pictures were very helpful. I'm taking a mental health day today, I'll be able to take some images and send you the PDF file tomorrow.

 

[Andy Resnick]

And here is the beam pattern that I see when beam comes out of the laser and hits the first lens after the focal point (this one is when the path length of the beam is rather short). I am not sure why I get the rings.

Here is how the beam profile changes as it travels longer distances. At certain distance I can make the "hole" in the center disappear and then the center of the beam seems to be gaussian.

Before I get to these images, here's two photos of my setup:

[PLAIN]http://img585.imageshack.us/img585/3264/dsc01324u.jpg

[PLAIN]http://img691.imageshack.us/img691/5079/dsc3306.jpg

The tweezers are coupled into an upright microscope that uses infinity-corrected objectives, so the beam coming out of the tweezer module is (nearly) collimated. The top view of the module has the following components (in order): Nd:YAG laser, mirror, 40 mm plano-convex lens, mirror, 250 mm plano-convex lens. Inside the microscope there's another mirror, directing the beam down into the objective lens. the tweezer module sits on top of a 4-axis positioner (x, z, tip, tilt) that allows me to bring the tweezer beam into alignment with the optic axis of the microscope.

Here's an image of the spot at the sample plane:

[PLAIN]http://img707.imageshack.us/img707/1190/flycap0.jpg

and slightly defocused:

[PLAIN]http://img574.imageshack.us/img574/7902/flycap1.jpg

For these images, I used a 100x NA 0.9 objective- I generally trap with a 63X 0.9 dipping objective, tho. The two small dots/interference fringes are caused by the mirror in the microscope, and are not actually present at the sample, AFAIK.

Your beam images look ok, there's perhaps a hint of astigmatism (the spots are not perfectly symmetric, there appears to be a feature at about 45 degrees on one image). This can be minimized by tip/tilt adjustments of your lens. At the trap, astigmatism will manifest by a jet of particles shooting off to one side.

I'm not seeing any evidence that your laser is multimode- can you take some images of the focused spot? Put a mirror in place of the sample (lay a coverslip on top of the mirror if your objective requires one), and take a picture of the reflected spot.

Can you make a trap?

 

[kubikat]

Hi Andy! Thank you a lot for your pictures. They were really helpful, since now I had a better idea of what I was looking for, looking at the beam. I finally got trapping! It was really really exciting. The trap was very strong and kept literally sucking in the particles from within of like 30-50 micrometers. My friend and I decided it looked like a gravitational potential well except in optics. I was able to move the particles around as well. The weird thing is though, I got trapping when I only had one of the telescope lenses in (the one with the large wavelength) so the beam wasn't even close to filling up the back aperture. I put the other lens in afterwards and it seemed to have made the trap less strong, but I still need to play around with it some more.

Here is the image of the trapped spheres:

I am so happy it finally worked. Do you think it looks okay?

 

[kubikat]

Oh and the cluster of the particles on the bottom is where a beam used to be before I moved it. I can also use the beam to grab particles one by one and arrange them in shapes.

 

[Andy Resnick]

Hi Andy! Thank you a lot for your pictures. They were really helpful, since now I had a better idea of what I was looking for, looking at the beam. I finally got trapping! It was really really exciting. The trap was very strong and kept literally sucking in the particles from within of like 30-50 micrometers. My friend and I decided it looked like a gravitational potential well except in optics. I was able to move the particles around as well. The weird thing is though, I got trapping when I only had one of the telescope lenses in (the one with the large wavelength) so the beam wasn't even close to filling up the back aperture. I put the other lens in afterwards and it seemed to have made the trap less strong, but I still need to play around with it some more.

Here is the image of the trapped spheres:


I am so happy it finally worked. Do you think it looks okay?

Oh and the cluster of the particles on the bottom is where a beam used to be before I moved it. I can also use the beam to grab particles one by one and arrange them in shapes.

Excellent- glad to hear it!

In response to your questions, here's my best guess:

1) You do not have a 3-D trap, you are pushing particles against the glass, where they stick due to electrostatic interactions. This may or may not be a problem, depending on your goal.

2) The optical trap does indeed share common characteristics with a gravitational well- the trapping energy is approximately a quadratic potential well, and dielectric particles 'fall' into it.

3) Your objective lens is not infinity-corrected, so the trap beam should not be collimated when it enters the objective. My suspicion is that you are focusing the trap beam with the singlet when the trap is optimized. If you want to fill the back aperture (which results in a 'stronger' trap) while still trapping objects in the focal plane, you need to adjust a few aspects of the telescope. It can get involved since the objective is not infinity corrected, and I'm not sure I can easily walk you through it- and I've never done it.

My design was based on this:

http://www.opticsinfobase.org/abstract.cfm?URI=ao-36-10-2107

and you may be able to get some ideas as well, especially from page 2109.

 

[hopeful]

This is a very helpful thread with some great information.

I'm currently doing a research project regarding laser tweezers and hope to build one after I learn enough about it. I've printed out the thread, and just wanted to let you three know that your work is greatly appreciated and probably helps more people than you realize. Thank You!