Download Experiment 4: Semiconductor Lasers | OPTI 511L and more Lab Reports Chemistry in PDF only on Docsity!
R. J. Jones Optical Sciences OPTI 511L Fall 2008
Experiment 4: Semiconductor Lasers 2 x 3 hours
Semiconductor (diode) lasers are by far the most widely used lasers today. Their small size and properties of the light output take some getting used to. Also these devices are
extremely power and static sensitive. The most basic elements of a theory of laser diode
operation can be found in Yariv; for a more extensive and modern treatment see e.g. "Fundamentals of Photonics" by Saleh & Teich. Detailed characteristics of a particular laser diode can be obtained from the manufacturers User's Manual.
The popularity of laser diodes is largely due to a host of desirable characteristics, such as ruggedness, reliability, compactness, light weight, high efficiency, low cost etc. From a laser physics standpoint diode lasers are also of interest because they make it relatively easy to observe two basic aspects of laser operation:
- Exploration of laser characteristics near threshold, including buildup of laser oscillation from noise as excitation increases through the threshold value.
- Measurement of the fundamental laser linewidth.
We will divide our study into four parts:
I. Becoming familiar with the devices. (week 1)
II. Measuring laser linewidth. (week 1 or 2)
III. Observing buildup of laser oscillation from noise using a homemade spectrometer. (week 2)
IV. Using a diffraction grating and a mirror to establish optical feedback. (week 2)
Part I. Laser Characteristics
1. Physical inspection and dimensions.
A. Examine a diode laser (use a dead one) with a binocular microscope.
B. Identify the electrical contacts. Can you determine the direction in which electrical current flows? You may need to find a detailed diagram of a heterojunction laser diode.
C. Adjust illumination to observe the mirror-like cleaved surface where light is emitted.
D. Using a calibrated length and eyepiece reticule, measure the laser cavity length L (separation between cleaved surfaces).
2. Beam shape.
A. Use a diode without a collimating lens. Measure the diode laser beam divergence, and record the uncertainties in your measurements that are consistent with your choice of measurement method. Compare beam divergence measurements with specifications.
Q1: Why is the laser beam elongated? How is the beam elongation oriented with respect to the p-n junction?
B: Assuming the beam is diffraction limited, what is the effective size of the radiating region?
3. Polarization.
Q2: How would one expect a diode laser beam to be polarized?
A. Measure the state of polarization of the laser output beam.
B. What is the relationship between polarization and laser resonator orientation?
input to the OSA. Try to observe the following features, and make sketches in your notebook as appropriate:
A. What is the longitudinal mode spacing of the laser (in units of length)? Be sure you record this number. What is the FSR of the laser (in frequency units)? If the index of refraction of the gain medium is 3.6, what is the laser's cavity length? How would you expect a mode frequency to change if the transverse mode number changes by 1?
B. Observe the rate at which a single cavity mode can acquire most of the laser's output power. Below threshold, many modes will have power. At what current does a single mode have nearly all of the laser power?
C. Look at the amplitude of the spectrum on a log scale. How many modes can you resolve in the entire spectrum when the current is just below threshold?
D. Put the laser at threshold. With the OSA still on a log scale, pick out a single mode at the center of the spectrum, and make the horizontal scale show just a few modes. As you increase the laser current, the central mode will gradually change in wavelength. How much of a shift is there for this mode over the available range of laser currents? Does this mode shift up or down in wavelength as current is increased?
E. Would this OSA be useful for analyzing the frequency spectrum of a gas laser? Remember, the resolution bandwidth of this OSA is 0.07nm.
7. Laser frequency
A. Use the OSA to measure the central optical wavelength of the diode laser as a function of injection current. Is the wavelength a simple function of injection current, or does it exhibit hysteresis, i.e. different values if a given current is approached from above and below?
B. For fixed injection current, measure the laser output wavelength as a function of laser temperature. Verify the manufacturers claim that the wavelength increases with temperature. Be careful not to exceed the rated output power as you cool down the laser.
Q4: You will not see a smoothly changing wavelength as a function of current and temperature. Suggest physical reasons why discontinuous frequency jumps might occur.
Part II. Laser linewidth
The linewidth of an ideal gas laser can be understood in terms of phase noise from spontaneous emission into the lasing mode. Shawlow and Townes derived the following expression
N 2
N 2 − N 1
π h ν Δ( ν cavity )
2
P
where
Δν cavity =
c
2 π L
1 − R
R
is the laser cavity linewidth, P the laser power output and N 1 , N 2 the populations of the lower and upper lasing level.
For semiconductor lasers there is an additional broadening that occurs due to the high density of charge carriers,
where the term 1 ( +α 2 ) typically has a value of 20 or more. This is sometimes referred
to as the "modified Shawlow-Townes limit". A rule of thumb to estimate linewidth of GaAs diode lasers is
( output power )× ( linewidth ) = 100 mW MHz.
For typical laser diode output powers this is in a measurable range.
8. Fabry-Perot measurement of laser linewidth.
Use the 2 GHz FSR Spectra-Physics FPI to measure the linewidth of the diode laser at a few different output powers. Compare this to the modified Shawlow-Townes limit, assuming that the diode laser facets have ~ 30% reflectivity.
Note: During this measurement it is important to minimize optical feedback into the laser. Do this by locating the FPI as far away as you can while still obtaining a visible signal. If you need to use a focusing lens, misalign it so that the lens is tilted with respect to the beam path(use the lens only to make the beam converge, but not focus, at the entrance hole).
10. Buildup of laser oscillation.
Observe the transition from amplified spontaneous emission below threshold, over multimode oscillation to single mode oscillation high above threshold. Examine again the change in laser wavelength versus injection current and laser temperature (to do so you must determine which direction in your display corresponds to increasing wavelength). Record your observations regarding the oscilloscope and the camera monitor. Look also for evidence of mode-hopping.
Based on the oscilloscope signal, make a sketch of the observed frequency spectrum for the different current and temperature conditions. Be sure to note the overall spectrum shape, and the relative changes in central position. The oscilloscope can be calibrated with an estimate of laser mode spacing.
11. Evaluation of the spectrometer performance.
Using the laser mode spacing that you obtained with the OSA for calibration, estimate the wavelength resolution of the diffraction grating/CCD setup.
The theoretical resolution of the diffraction grating (used in first order) is
Δλ th =
N
where N is the number of illuminated grating lines. We can obtain N from the known grating line spacing d and an estimate of the beam size.
Compare the observed and theoretical resolution.
(continued on next page) Part IV. Optical feedback
This part of the lab explores the use of a diffraction grating in single-frequency laser operation. The principles are similar to those found in an upcoming lab, and we begin to explore the techniques here.
You are to follow the TA's instructions for setting up this part of the lab. The setup is almost identical to that of the grating spectrometer from Part III. Here, however, you will take the additional step of sending some of the light that is diffracted off of the grating back onto itself, and thus intentionally sending it back into the laser. You will use a glass plate or beamsplitter to reflect a small portion of light from the grating- diffracted beam to the CCD spectrometer you set up earlier in the lab. The light transmitted through the beamsplitter will be sent to a retro-reflecting mirror, which will direct some of the diffracted light back towards the diffraction grating. The different wavelengths in the beam will be further separated by the grating, and in returning through the focusing optics, will be spatially resolvable at the face of the diode laser. You can now align the retro-reflecting mirror such that only one of the spatially resolved back-reflected diffraction peaks enters the LD.
With proper alignment, the beam that returns to the LD will cause additional light build- up in the LD for one particular axial mode, whichever one is properly retro-reflected back to the laser. This is a form of injection locking, or seeding the laser with light, and is an example of optical feedback.
(1) Look on the CCD monitor, and the oscilloscope. Align the retro-reflecting mirror such that one mode near the peak of the output grows in strength. Then, scan the horizontal tilt of the mirror. Try to observe the effects of feedback as you successively inject one mode after another.
(2) With laser current and retro-reflecting mirror adjusted properly, you should find that the laser can operate in primarily two axial modes simultaneously when you inject the retro-reflected beam. One of these modes will be due to the injected optical feedback, the other will be the primary natural lasing mode for the particular current and temperature with which you are working. Estimate the fraction of the total light power that can be put into the injected mode. Can you force the laser to operate almost entirely in a mode other than its natural lasing mode? How many different modes can you enhance with the injected light?
References:
"Fundamentals of Photonics, Saleh and Teich, Wiley (1991)
