A More or Less Edible Laser
I recently found Stephen Wilk’s article “Edible Lasers: What’s the Next Course?” (which appeared in Optics & Photonics Magazine) and got into email correspondence with him about it, during the course of which I realized that I couldn’t actually recall having lased a dye that was dissolved in glycerol. (I probably did lase Rhodamine B in glycerol a long time ago, in a galaxy far, far away; but that was then and this is now, and that effort, if it did indeed occur, has long since faded from my memory.)
Accordingly, on Tuesday evening I set up a homebrew cuvette, and lased Rhodamine B in glycerol with longitudinal pumping from a nitrogen laser, something I think I’ve only done once or twice before. Almost nobody ever bothers to pump a dye laser longitudinally with the output of a nitrogen laser. There are several reasons for this, one of which is that half of the output from the dye goes back into the nitrogen laser and is lost unless you take steps to prevent that from happening, which is itself a nontrivial exercise. I didn’t take photos on Tuesday, but what I saw was interesting enough that I returned on Wednesday afternoon and did it again; see below.
I used Rhodamine B because it has an intriguing character: the efficiency of its fluorescence emission is directly proportional to the viscosity of the solvent that it is dissolved in. If you put RhB in methanol, the quantum efficiency is only about 0.43; by the time you get to glycerol it is more like 0.96, which is excellent that’s comparable to the efficiency of Rhodamine 6G or Fluorescein, both of which are extremely good laser dyes. RhB absorbs very little at the wavelength that the nitrogen laser provides, so the increased efficiency is extremely helpful. I could have added a second dye to absorb the UV and convert it to a wavelength range that RhB absorbs better (I have done this in the past to assist several dyes that don’t absorb much at 337 nm, as have various other folks), but that would involve extra tweaking, and I wanted to do this expeditiously. In addition, most laser dyes are toxic, and that’s not what I’m trying for here.
After the longitudinal pumping test succeeded I added more RhB, and lased the solution with transverse pumping. Here are two photos. The first shows the setup, and the second shows the output on a piece of paper. The solution was probably inadequately mixed; in addition, the glycerol has been heated unevenly by waste energy from previous pumping pulses, and it has almost certainly developed Schlieren, which interfere with the optical path and cause the beam to spread out. (Schlieren are irregularities in refractive index; think about the “wiggles” you see on the pavement when the sun shines down through the exhaust from a bus or truck, or what happens when you mix two liquids that have significantly different refractive indices, for example water and isopropyl alcohol or even isopropyl alcohol and ethanol.)
[There is a PRA LN-1000 TEA Nitrogen laser just outside of the picture, at right. It puts out pulses of ultraviolet light with wavelength of about 337.1 nm; the pulses last a little less than 1 nanosecond. The beam from the nitrogen laser goes through a cylindrical lens (visible at the right edge of the “setup” photo), and then through an ordinary achromat (which fluoresces in the “lasing” photo, clearly indicating that it is absorbing some of the UV from the LN-1000, grump). These two lenses focus the pump beam to a narrow line across the front of the dye solution in the cuvette, which you can’t see because the fluorescence of the dye is so bright. This is the usual way of pumping a dye laser with a nitrogen laser, though people typically use only a single cylindrical lens; I added the second lens to get a more compact setup.]
I like Rhodamine B, and it is less toxic than Rhodamine 6G; but you still wouldn’t want to drink it, so I redid the demonstration with Fluorescein:
I’m sure this solution would taste nasty, but you could certainly drink a small amount of it without hurting yourself, so it qualifies as edible. I didn’t even have to add any dish detergent to it. (The original edible laser was a slight cheat they were obliged to add a few drops of detergent to their material. In fact, I have heard Ted Hänsch say, concerning that particular issue, “At that point, Art stopped insisting on eating the experimental subjects.” Ahem. ;o)
[T. A. Hänsch, M. Pernier, and A. L. Schawlow, “Laser Action of Dyes in Gelatin”, IEEE Journal of Quantum Electronics, Volume QE-7 (January 1971), page 47. Also see http://histsoc.stanford.edu/pdfmem/Schawlow_Arthur.pdf , though this and other texts I find on the Web contain an apparent inaccuracy: AFAIK, they did not put Fluorescein into Jell-O™. They put it into unflavored gelatin and, as I mention above, it did not immediately work; like RhB, Fluorescein does not absorb well at 337 nm, on top of which gelatin is not as transparent as glycerol. They had to put so much dye into the gelatin that they got concentration quenching, which interfered with the fluorescence enough to prevent lasing. The detergent helps prevent this problem.]
Serendipity Strikes Again
As I mention above, I returned this afternoon to see whether I could get a photo of the results with longitudinal pumping. I removed the cylindrical lens, leaving the ordinary lens, and this is what I saw on the paper viewing screen with the cuvette of dye moved to the side:
This is a fine image of the inside of the laser head, showing the electrodes and the discharge. When I put the cuvette back in position, I was surprised to discover that I was getting lasing on both axes at once:
(Lasing on more than one axis at a time is far from unheard-of; but it’s not common, and this is certainly the first time I’ve done it or seen it with nitrogen-pumped dye I just happen to have chosen a dye concentration that supports it under my conditions, and a lens with appropriate focal length.)
Here are separate photos of the two outputs, for a bit more detail:
What I find particularly spiffy and interesting is the fact that the longitudinal lasing output from the dye cuvette is still an image of the inside of the nitrogen laser head (!). Notice that the L output is yellowish, while the T output is green. My belief is that this is because the L output has to travel through a lot more unexcited dye solution before it gets out of the cuvette, and the dye absorbs the short-wavelength tail of the emission. (This is a well-known and well-understood issue, very common with fluorescent dyes.)
Note: I took these photos with my iPhone. If anyone reading this has a strong need for higher resolution images, please contact me via email.
Accordingly, on Tuesday evening I set up a homebrew cuvette, and lased Rhodamine B in glycerol with longitudinal pumping from a nitrogen laser, something I think I’ve only done once or twice before. Almost nobody ever bothers to pump a dye laser longitudinally with the output of a nitrogen laser. There are several reasons for this, one of which is that half of the output from the dye goes back into the nitrogen laser and is lost unless you take steps to prevent that from happening, which is itself a nontrivial exercise. I didn’t take photos on Tuesday, but what I saw was interesting enough that I returned on Wednesday afternoon and did it again; see below.
I used Rhodamine B because it has an intriguing character: the efficiency of its fluorescence emission is directly proportional to the viscosity of the solvent that it is dissolved in. If you put RhB in methanol, the quantum efficiency is only about 0.43; by the time you get to glycerol it is more like 0.96, which is excellent that’s comparable to the efficiency of Rhodamine 6G or Fluorescein, both of which are extremely good laser dyes. RhB absorbs very little at the wavelength that the nitrogen laser provides, so the increased efficiency is extremely helpful. I could have added a second dye to absorb the UV and convert it to a wavelength range that RhB absorbs better (I have done this in the past to assist several dyes that don’t absorb much at 337 nm, as have various other folks), but that would involve extra tweaking, and I wanted to do this expeditiously. In addition, most laser dyes are toxic, and that’s not what I’m trying for here.
After the longitudinal pumping test succeeded I added more RhB, and lased the solution with transverse pumping. Here are two photos. The first shows the setup, and the second shows the output on a piece of paper. The solution was probably inadequately mixed; in addition, the glycerol has been heated unevenly by waste energy from previous pumping pulses, and it has almost certainly developed Schlieren, which interfere with the optical path and cause the beam to spread out. (Schlieren are irregularities in refractive index; think about the “wiggles” you see on the pavement when the sun shines down through the exhaust from a bus or truck, or what happens when you mix two liquids that have significantly different refractive indices, for example water and isopropyl alcohol or even isopropyl alcohol and ethanol.)
[There is a PRA LN-1000 TEA Nitrogen laser just outside of the picture, at right. It puts out pulses of ultraviolet light with wavelength of about 337.1 nm; the pulses last a little less than 1 nanosecond. The beam from the nitrogen laser goes through a cylindrical lens (visible at the right edge of the “setup” photo), and then through an ordinary achromat (which fluoresces in the “lasing” photo, clearly indicating that it is absorbing some of the UV from the LN-1000, grump). These two lenses focus the pump beam to a narrow line across the front of the dye solution in the cuvette, which you can’t see because the fluorescence of the dye is so bright. This is the usual way of pumping a dye laser with a nitrogen laser, though people typically use only a single cylindrical lens; I added the second lens to get a more compact setup.]
I like Rhodamine B, and it is less toxic than Rhodamine 6G; but you still wouldn’t want to drink it, so I redid the demonstration with Fluorescein:
I’m sure this solution would taste nasty, but you could certainly drink a small amount of it without hurting yourself, so it qualifies as edible. I didn’t even have to add any dish detergent to it. (The original edible laser was a slight cheat they were obliged to add a few drops of detergent to their material. In fact, I have heard Ted Hänsch say, concerning that particular issue, “At that point, Art stopped insisting on eating the experimental subjects.” Ahem. ;o)
[T. A. Hänsch, M. Pernier, and A. L. Schawlow, “Laser Action of Dyes in Gelatin”, IEEE Journal of Quantum Electronics, Volume QE-7 (January 1971), page 47. Also see http://histsoc.stanford.edu/pdfmem/Schawlow_Arthur.pdf , though this and other texts I find on the Web contain an apparent inaccuracy: AFAIK, they did not put Fluorescein into Jell-O™. They put it into unflavored gelatin and, as I mention above, it did not immediately work; like RhB, Fluorescein does not absorb well at 337 nm, on top of which gelatin is not as transparent as glycerol. They had to put so much dye into the gelatin that they got concentration quenching, which interfered with the fluorescence enough to prevent lasing. The detergent helps prevent this problem.]
Serendipity Strikes Again
As I mention above, I returned this afternoon to see whether I could get a photo of the results with longitudinal pumping. I removed the cylindrical lens, leaving the ordinary lens, and this is what I saw on the paper viewing screen with the cuvette of dye moved to the side:
This is a fine image of the inside of the laser head, showing the electrodes and the discharge. When I put the cuvette back in position, I was surprised to discover that I was getting lasing on both axes at once:
(Lasing on more than one axis at a time is far from unheard-of; but it’s not common, and this is certainly the first time I’ve done it or seen it with nitrogen-pumped dye I just happen to have chosen a dye concentration that supports it under my conditions, and a lens with appropriate focal length.)
Here are separate photos of the two outputs, for a bit more detail:
What I find particularly spiffy and interesting is the fact that the longitudinal lasing output from the dye cuvette is still an image of the inside of the nitrogen laser head (!). Notice that the L output is yellowish, while the T output is green. My belief is that this is because the L output has to travel through a lot more unexcited dye solution before it gets out of the cuvette, and the dye absorbs the short-wavelength tail of the emission. (This is a well-known and well-understood issue, very common with fluorescent dyes.)
Note: I took these photos with my iPhone. If anyone reading this has a strong need for higher resolution images, please contact me via email.