by
Matthew Eardley
Department of Physics,
SUNY at Stony Brook
Advisor: Harold Metcalf
Introduction
Part I: In the Lab
A. Setup
In the Laser Teaching Center, this was the setup:
B. Equipment
I used the following equipment:
To measure the power of the laser, I used the photo-detector, which gives a current output proportional to the power coming in. I used a 10K resistor across the terminals of the digital multimeter to convert the signal to a voltage in the proper range. Thus on the multimeter I read a voltage that was proportional to the power of the beam. I first measured the power coming directly out of the laser as a reference. Then, as I was aligning the beam into the fiber, I aligned the output end of the fiber with the power meter to measure the amount of beam that was getting coupled.
The PAF fiber port, shown below, is what couples the laser beam into the fiber.
The beam comes in one end, and the fiber port lens focuses the
beam to a point smaller than the size of the optical fiber.
For good coupling, the beam needs to come in at normal incidence
to the fiber port lens, and it needs to hit the center
of the lens. The lens has five degrees of freedom: x, y, z,
and two angles, pitch and yaw. The x and y are axes perpendicular
to the plane of the lens, and z is along the beam line, as shown:
All of these can be controlled by slowly turning screws on the
back of the fiber port. However, it is more convenient to use the two
mirrors to adjust the x, y, pitch, and yaw of the beam. Since it is
further away from the coupler, mirror 1 only affects the x/y position
of the beam to good approximation, and mirror 2 affects mainly the pitch
and yaw of the beam since it is closest to the coupler.
Step 1: Align the beam in the coupler without the fiber plugged in. I adjusted the mirrors until the beam went through the coupler without reflecting off the walls. I also used a cross drawn on a piece of cardboard held at various distances from the coupler to ensure that the beam going through perpendicular to the plane of the lens.
Step 2: Plug the fiber in the coupler and see if anything comes out. With the lights off, I could see a very faint spot projected on a piece of paper (the signal was too faint for the power meter to read). Then I used the knobs on the mirrors to maximize the output of the beam. Adjustment of the mirrors must be done iteratively, because you are basically walking up a "hill" in a four dimensional phase space of two directions and two angles. At this point it is best to have the fiber backed off of the coupler, not screwed all the way in, as shown:
Further away from the coupler, the beam is broader so it is
easier to find a maximum.
Step 3: Once the fiber was in the center of the beam, I slowly inserted the fiber into the coupler until it was screwed all the way on, maximizing the output with the mirrors each time. The output of the fiber was then bright enough to be measured by the power meter, but the output was only about one or two percent of the input.
Step 4: The final parameter to adjust is the distance from the fiber to the lens. The fiber needs to sit exactly at the focus of the lens for maximum coupling. I slowly adjusted the z-position of the lens with the screw on the fiber port, each time maximizing the output with the mirrors.
D. Data
The following table shows my results:
The first alignment took several days of trying. It took me a while to figure out the proper method for aligning the mirrors and the coupler, based on the coupler's manual and previous web reports on fiber optics (see References below). The second alignment took about an hour, and based on others' experiences, this is about as fast as one could reasonably expect to do this. Aligning a laser into a fiber is not a trivial thing to do!
E. Application
The next step was to apply this hard-earned skill to a real experiment. One of Professor Metcalf's students, Oleg Kritsun, was doing an experiment where he needed to take light from a Titatium-Sapphire laser in one room and use it for an experiment in another. An optical fiber went between the two rooms. The process of coupling the laser to the fiber was the same, except for one thing: there was only one mirror, so I used the mirror to adjust the angles, and the x and y screws on the coupler for the position. This made things slightly more difficult because the screws on the coupler are much more sensitive than the mirros, but the end result was the same.
Things were more complicated because the two ends of the fiber were in different rooms. The output end was again pointed at a power meter, and a video camera was trained on the power meter. Next to the coupler was a moniter, so I could watch the power meter as I adjusted the beam alignment. Since the whole apparatus was on an optical bench, the alignment was stable as long as no one hit the coupler or the mirror.
Part II: Frequency Combs
A. Background
A frequency comb (a.k.a. supercontinuum, frequency chain) is white light that has "the brightness of a laser with the bandwidth of a lightbulb." This light is produced by putting femtosecond laser pulses through special sections of optical fiber that produce a broad range of frequencies, in many cases from ultraviolet to infrared.
B. Applications
There are a wide variety of possible applications of light with such characteristics. They include:
Frequency combs were first produced using microstructure photonic crystal fibers (PCFs) [paper 1]. This fiber is basically a silica core surrounded by large air holes that run the length of the fiber. Unfortunately, these are very expensive to make and are quite delicate. T. A. Birks et al [paper 2] discovered another way to generate a frequency comb, using a tapered fiber, shown below:
The tapered fibers were made from conventional multimode telecommunication fiber by heating and stretching in a flame. The transition from untapered to tapered is about 35 mm, and the taper waste has a diamater of about 2µm with a length of 90mm. They can be produced in the lab in a matter of minutes, and are not as delicate PCFs. In addition, the untapered fiber has a much larger core than a PCF, so alignment is easier and the ends are less prone to damage. The following graph shows the output spectrum of two such tapered fibers:
The input pulses came from a Ti:sapphire laser (λ = 850 nm) with a length of 200-500 fs and a repetition rate of 76 Mhz.
D. Pictures
Another group is using tapered fibers to generate frequency combs, that of Harald Giessen in Bonn, Germany [web 5]. Here is a picture of the spectrum light from a frequency comb:
References
A. Papers
B. Web Pages
Acknowledgements
I would like to thank Professor Harold Metcalf for direction and advise with this project. Dr. John Noé in the Laser Teaching center gave me a great deal of assistance in learning how to couple the laser into the fiber. I would also like to thank Oleg Kritsun for allowing me to use his setup for testing my alignment skills, and Professor Harald Giessen for being a great source of information on tapered fibers and frequency combs.
My goal for this project was to learn how to couple laser
light into an optical fiber [web 1]. This is particularly
useful in experiments involving lasers because the laser itself is not always
conveniently located: so optical fibers can be used to transport the beam
across a room, or to the other side of a building. I first learned
how to couple a laser into a fiber in the Laser Teaching Center, and I then
applied this knowledge to couple a laser into a fiber that went between a
room in Professor Metcalf's lab to a room in Professor Orozco's lab.
In addition, I researched an interesting application of fiber optics: the production of a "frequency comb," a beam of coherent white light generated using femtosecond laser pulses and fibers with unique optical properties. One particularly interesting experiment involved using only a tapered fiber to generate a frequency comb.
In addition, I researched an interesting application of fiber optics: the production of a "frequency comb," a beam of coherent white light generated using femtosecond laser pulses and fibers with unique optical properties. One particularly interesting experiment involved using only a tapered fiber to generate a frequency comb.
Part I: In the Lab
A. Setup
In the Laser Teaching Center, this was the setup:
B. Equipment
I used the following equipment:
- HeNe laser, wavelength 633 nm
- 2 mirrors: mirror 1 for x/y alignment, mirror 2 for angular adjustment
- PAF fiber port
- Single-mode fiber
- Multi-mode fiber
- Thorlabs photo-detector, with 10K load resistor
- Digital Multimeter
To measure the power of the laser, I used the photo-detector, which gives a current output proportional to the power coming in. I used a 10K resistor across the terminals of the digital multimeter to convert the signal to a voltage in the proper range. Thus on the multimeter I read a voltage that was proportional to the power of the beam. I first measured the power coming directly out of the laser as a reference. Then, as I was aligning the beam into the fiber, I aligned the output end of the fiber with the power meter to measure the amount of beam that was getting coupled.
The PAF fiber port, shown below, is what couples the laser beam into the fiber.
Step 1: Align the beam in the coupler without the fiber plugged in. I adjusted the mirrors until the beam went through the coupler without reflecting off the walls. I also used a cross drawn on a piece of cardboard held at various distances from the coupler to ensure that the beam going through perpendicular to the plane of the lens.
Step 2: Plug the fiber in the coupler and see if anything comes out. With the lights off, I could see a very faint spot projected on a piece of paper (the signal was too faint for the power meter to read). Then I used the knobs on the mirrors to maximize the output of the beam. Adjustment of the mirrors must be done iteratively, because you are basically walking up a "hill" in a four dimensional phase space of two directions and two angles. At this point it is best to have the fiber backed off of the coupler, not screwed all the way in, as shown:
Step 3: Once the fiber was in the center of the beam, I slowly inserted the fiber into the coupler until it was screwed all the way on, maximizing the output with the mirrors each time. The output of the fiber was then bright enough to be measured by the power meter, but the output was only about one or two percent of the input.
Step 4: The final parameter to adjust is the distance from the fiber to the lens. The fiber needs to sit exactly at the focus of the lens for maximum coupling. I slowly adjusted the z-position of the lens with the screw on the fiber port, each time maximizing the output with the mirrors.
D. Data
The following table shows my results:
| Total Output [V] | Output of HeNe [V] |
Coupling % |
|
| Alignment 1 |
1.95 ± 0.15 |
3.16 ± 0.05 |
62.2 ± 5.2 |
| Alignment 2 |
1.9 ± 0.1 |
3.20 ± 0.05 |
59.4 ± 4.1 |
The first alignment took several days of trying. It took me a while to figure out the proper method for aligning the mirrors and the coupler, based on the coupler's manual and previous web reports on fiber optics (see References below). The second alignment took about an hour, and based on others' experiences, this is about as fast as one could reasonably expect to do this. Aligning a laser into a fiber is not a trivial thing to do!
E. Application
The next step was to apply this hard-earned skill to a real experiment. One of Professor Metcalf's students, Oleg Kritsun, was doing an experiment where he needed to take light from a Titatium-Sapphire laser in one room and use it for an experiment in another. An optical fiber went between the two rooms. The process of coupling the laser to the fiber was the same, except for one thing: there was only one mirror, so I used the mirror to adjust the angles, and the x and y screws on the coupler for the position. This made things slightly more difficult because the screws on the coupler are much more sensitive than the mirros, but the end result was the same.
Things were more complicated because the two ends of the fiber were in different rooms. The output end was again pointed at a power meter, and a video camera was trained on the power meter. Next to the coupler was a moniter, so I could watch the power meter as I adjusted the beam alignment. Since the whole apparatus was on an optical bench, the alignment was stable as long as no one hit the coupler or the mirror.
Part II: Frequency Combs
A. Background
A frequency comb (a.k.a. supercontinuum, frequency chain) is white light that has "the brightness of a laser with the bandwidth of a lightbulb." This light is produced by putting femtosecond laser pulses through special sections of optical fiber that produce a broad range of frequencies, in many cases from ultraviolet to infrared.
B. Applications
There are a wide variety of possible applications of light with such characteristics. They include:
- Frequency metrology. Once the frequencies of the "teeth" in the frequency comb are known, the light can be used to compare the frequencies of light from a known source and an unknown source. Thus the frequency comb is used as a ruler to precisely measure the difference between to frequencies of light. For example, T. Hänsch's in Munich has done spectroscopic experiments with Hydrogen [e.g. f1s2s = 2 466 061 413 187 103 (46) Hz], Cesium, and Indium using frequency combs [web 2, web 3].
- Atomic clocks. Since atomic clocks rely on the measurement of transition rates of electrons in an atom, the increased accuracy gained with frequency combs allows for more accurate measurement of time. Current atomic clocks use radio frequencies, whereas frequency combs measure optical (visible) frequencies.
- Other areas of atomic physics that rely on precise optical frequency measurements.
Frequency combs were first produced using microstructure photonic crystal fibers (PCFs) [paper 1]. This fiber is basically a silica core surrounded by large air holes that run the length of the fiber. Unfortunately, these are very expensive to make and are quite delicate. T. A. Birks et al [paper 2] discovered another way to generate a frequency comb, using a tapered fiber, shown below:
The tapered fibers were made from conventional multimode telecommunication fiber by heating and stretching in a flame. The transition from untapered to tapered is about 35 mm, and the taper waste has a diamater of about 2µm with a length of 90mm. They can be produced in the lab in a matter of minutes, and are not as delicate PCFs. In addition, the untapered fiber has a much larger core than a PCF, so alignment is easier and the ends are less prone to damage. The following graph shows the output spectrum of two such tapered fibers:
The input pulses came from a Ti:sapphire laser (λ = 850 nm) with a length of 200-500 fs and a repetition rate of 76 Mhz.
D. Pictures
Another group is using tapered fibers to generate frequency combs, that of Harald Giessen in Bonn, Germany [web 5]. Here is a picture of the spectrum light from a frequency comb:
References
A. Papers
- J.K. Ranka, R. S. Windeler, and A. J. Stents, "Visible continuum generation in air-silica microstructure optical fibers with anomalous dispersion at 800 nm," Opt. Lett. 25, 25 (2000).
- T.A. Birks, W. J. Wadsworth, and P. St. J. Russell, "Supercontinuum generation in tapered fibers," Opt. Lett. 25, 1415 (2000).
B. Web Pages
- The web pages of students who have already done Optics Rotation Projects using fiber optics proved extremely useful. They include those of Xueqing Liu, Xiyue Miao, Haijiang Gong, and Jing Wang.
- The "frequency chain" pages of T. W. Hänsch's group contain many interesting applications of frequency combs, some of which are mentioned above.
- MenloSystems sells a product based on the use of frequency combs from a PCF for measurement and synthesis optical frequencies.
- T. A. Birks and his collaborators, also of the University of Bath, are now involved with a commercial company called Blaze Photonics that specializes in the production of PCF.
- The pages of Harald Giessen's group in Bonn show some spectacular pictures of frequency comb light, but they have not yet published any papers on the subject.
Acknowledgements
I would like to thank Professor Harold Metcalf for direction and advise with this project. Dr. John Noé in the Laser Teaching center gave me a great deal of assistance in learning how to couple the laser into the fiber. I would also like to thank Oleg Kritsun for allowing me to use his setup for testing my alignment skills, and Professor Harald Giessen for being a great source of information on tapered fibers and frequency combs.