Department of Physics, Middlebury College 1992-93
Modern Physics Laboratory


XVII. Optogalvanic and Absorption Spectroscopy
Using a Tunable 670 nm Diode Laser


Discussion

Diode lasers operating in the infrared, and more recently in the visible red, have found many commercial applications, including use in fiber optic communications systems, laser printers, and remote sensing, and in consumer products such as compact dis k players, home security systems, and remote control devices. Because of their low cost, modest power requirements, and reliability, diode lasers are rapidly replacing older gas lasers such as the HeNe laser and tunable dye lasers in many scientific appl ications in the red and infrared region of the optical spectrum.

In this experiment you will use the new Toshiba TOLD 9211, visible red, 670 nm diode laser as a tunable, narrow linewidth, laser light source for high resolution optogalvanic and absorption spectroscopy. The narrow linewidth (100 MHz, or equivalently, 0.0007 Å at 6700 Å) of this laser will allow you to measure the absorption profile of the 21P1-->31D2, 6678 Å line of He with considerably better resolution than you could using the SPEX 1704 1 m monochromator, which has a resolution of only 0.04 Å in the same spectral region. The resolution comparison between the 1 m spectrometer and the diode laser is even more striking when you consider that the TOLD 9211 is about the size of a pea and costs less than $30.

Only a brief sketch of diode laser operating characteristics can be given here. Reviews of diode laser techniques in atomic spectroscopy can be found in Refs. 1 and 2. The optogalvanic effect in high resolution spectroscopy is reviewed in Refs. 3 and 4. The interested student is encouraged to browse through these references for more details.

(1) Diode Lasers.

Diode lasers capable of operation at a number of wavelengths in the range from 6500 Å (visible red) to 1.6 Ám (near infrared), are now commercially available at low cost. This discussion will highlight the properties of the TOLD 9211, 670 nm diod e laser, but diode lasers that operate at other wavelengths can be characterized in a similar manner.1,2 Complete details of the internal construction of the TOLD 9211 diode laser are of proprietary, but the general layout of components inside the diode package is given in Fig. 1(a). Fig. 1(a) shows clearly the small size of the actual diode laser chip relative to the size of the TOLD 9211 protective c ase. The diode laser chip emits an external laser beam that passes out of the case through a glass window, as well as a beam that travels to the rear of the case to a photodiode chip that provides a convenient internal monitor of the laser output power. The multi-layer structure of the diode laser chip is shown in Fig. 1(b). As current passes through the semiconducting layers of the diode structure, electrons and holes recombine in the active layer of InGaP to emit p hotons. Stimulated emission occurs within the active layer to produce a laser beam that is confined by the lower refractive indices of the adjacent cladding layers, in a process called index guiding. The cleaved facets of the active layer form the end m irrors of the laser cavity. The active layer is made sufficiently thin that the laser operates in its fundamental transverse mode with linear polarization parallel to the active layer. Because the thickness of the active layer is so small, diffractive e ffects are significant, and the laser beam is highly divergent in the direction perpendicular to the active layer, as shown in Fig. 1(b). For the TOLD 9211, the parallel and perpendicular beam divergences, and of Fig. 1(b), are 8o and 31o, respectively.3 For most applications a narrow beam is desired and therefore the diode laser beam must be focussed by an external lens.

The minimal electrical circuit necessary to operate the diode laser and its photodiode is given in Fig. 2(a). Considerably more complicated circuits are necessary to provide for adjustment and modulation of the dio de laser current and to suppress strong electrical transients that can permanently damage the diode. In normal operation the diode laser and the internal photodiode detector are operated in forward and reverse bias configurations, respectively. The elec trical requirements of the TOLD 9211 are modest: a mere 2.3 V applied to the diode laser results in a 50 mA diode current and a 4 mW laser output beam. This corresponds to a 3-4% efficiency for conversion of electrical energy into laser light energy. Th is is more than 100 times as efficient as the electrical-to-optical energy conversion process that takes place in the HeNe laser of Expt. I.

As the forward current through the diode is increased, there is a sudden increase in laser output power, indicating laser action, at a threshold current in the range of 40 to 50 mA, as shown in Fig. 2(b). As the fo rward current is increased above threshold, the laser output power increases sharply and the spectral output gradually changes from multimode emission to single mode emission, as shown in Fig. 3(a). The approximately 1 .3 Å spacing between longitudinal laser modes is determined by the distance between the end facets and the index of refraction of the active lasing layer of the laser diode. The relatively large spacing between longitudinal modes is due to the very short optical length of the diode laser cavity.

When operating in a single longitudinal mode, the diode laser wavelength can be changed by varying the forward current and temperature of the diode. Unfortunately for the purposes of high resolution spectroscopy, a given diode laser cannot be tuned c ontinuously throughout a large spectral region at will. Instead, a given diode laser can only tune continuously over small wavelength regions as indicated by the "staircase" of lines in Fig. 3(b). Clearly, for a given diode laser there are many wavelengths at which single mode laser emission is not possible. In order to produce single mode laser light at the wavelength of a specific atomic transition it is necessary to purchase many diode lasers at once and to hope t hat the unavoidable nonuniformities inherent in the fabrication processes result in at least one diode having a longitudinal mode structure that allows single mode lasing at the desired transition. Although there is an element of luck involved in such a venture, the purchase of at least five diodes will almost always guarantee success.

(2) Optogalvanic Spectroscopy.

In optogalvanic spectroscopy, the current passing through a gas discharge is monitored as a laser light source is tuned through the frequencies of allowed transitions for excited atoms in the discharge, as shown in Fig. 4(a). When the laser resonantly excites an atom from a low-lying state to a state of higher excitation, as in the example of the 21P1-->31D 2, 6678.151 Å transition of He shown in Fig. 4(b), an atom is excited to a less bound state, thereby increasing the probability that the atom will be ionized by discharge collisions and contribute to a n increase in the discharge current. This small change in discharge current can be detected with great sensitivity if the laser beam is chopped and the periodic variation of the discharge current at the chopping frequency is detected by a lock-in amplifi er. It should be emphasized that in contrast to other spectroscopic methods you have used in other experiments, optogalvanic spectroscopy does not require a photomultiplier tube or photodiode detector to obtain atomic transition spectra, because the gas discharge itself serves as a resonant photodetector.

There are many subtleties concerning the optogalvanic effect that cannot be discussed here, such as the quite common case that the gas discharge current is found to decrease  when a laser is adjusted to resonance with an atomic transition of excited discharge atoms. (Can you explain why this might happen?) The interested student will have to consult Refs. 3 and 4 to learn more about the puzzling dynamics of the optogalvanic effect in gas discharges.

Discussion

(1) In this step you will familiarize yourself with the operation of the LDC-3722 Laser Diode Controller and observe fluorescence from iodine molecules (I2) as the diode laser is tuned manually through absorption lines o f I2 in the range of wavelengths accessible to the TOLD 9211.

(a) Turn on cooling water to the LDM-4412 laser diode mount and check for a steady water flow. Turn on power to the LDC-3722. Set the thermoelectric cooler (TEC) to operate at 0 oC and the diode laser current (LASER) to 40 mA. Turn on the thermoelectric cooler and the diode laser current using the ON buttons in the TEC MODE and LASER MODE sections of the LDC-3722 front panel, respectively. The laser should emit a dim red beam. Push the PP D button in the LASER DISPLAY section of the front panel to determine the power output of the diode laser according to the internal photodiode of the TOLD 9211. Be sure that the correct photodiode sensitivity parameter has been entered into the LDC-3722 memory. The diode laser power should be about 1 mW. Push the LASER button in the ADJUST section of the LDC-3722 front panel and slowly adjust the laser current until 5 mW of diode laser power is obtained. It should not be necessary to exce ed a diode current of 60 mA to obtain 5 mW of laser output power. The current limit setting of the LDC-3722 will shut off the diode laser current if you attempt to exceed 60 mA.

(b) Mount the I2 absorption cell in front of the diode laser as shown in the photograph of Fig. 5. Turn out the room lights and let your eyes adapt to the darkness. As you r partner slowly varies the diode laser current, look at the I2 cell from the side and watch for a faint, red fluorescent glow along the axis of the diode laser beam. With some practice you will be able to find several I2 absorption lines and to return to them by setting the diode laser current with the LDC-3722.

(c) Lower the diode laser temperature to -10 oC and try to observe additional I2 fluorescence lines. When you have convinced yourself that you can see the fluorescence and that y ou can scan through fluorescence peaks by adjusting the diode laser current, put a brief note in your laboratory notebook and remove the I2 cell before going to the next step.

(2) Here you will measure multimode and single mode spectra of laser light emitted by the TOLD 9211 diode laser. These measurements will give you useful information concerning the range of operating currents necessary for single mode operation of th e diode near the 6678 Å He absorption line.

To obtain diode laser emission spectra, direct the laser beam to the SPEX 1704 spectrometer using mirror assembly M. Place diffuser plate D at the entrance slit so that the laser beam does not pass directly into the spectrometer. Set the diode laser temperature to 0 oC and obtain laser emission spectra for currents in the range 35 to 60 mA. Because every diode laser has slightly different operating characteristics, you will have to take a few spectra to determine the center wavelength of the laser emission lines and the range of currents over which the transition from multimode to single mode laser line emission takes place.

Obtain at least four diode laser emission spectra that demonstrate clearly the transition from multimode to single mode emission, such as those in Fig. 3(a). Plot these four spectra for inclusion in your laboratory report. For each spectrum record the diode laser output power as provided by the TOLD 9211 internal photodiode.

(3) Block the light from the diode laser and remove the diffuser plate D. Mount the chopper and HeNe discharge tube as shown in Fig. 6. Set the chopper to a frequency of 700 Hz. Turn on the HeNe discharge and ad just mirror M so that light from the discharge is present at the entrance slit of the spectrometer. Scan a 20 Å region near the 6678 Å He line to obtain a spectrum like that in Fig. 7(a). The two peaks in Fig. 7(a) are the 6678.151 Å, 31D2-->21P1 He and the 6678.2764 Å, 2p4 -->1s2 Ne emission lines.6

Your goal in this part of the experiment is to obtain optogalvanic and absorption spectra of the 6678 Å He line by tuning the diode laser through the He line absorption profile. To do this, unblock the diode laser beam and allow a small portion of the laser beam to pass through the entrance slit. Adjust the diode laser current and temperature to bring the laser wavelength to the 6678 Å He line. Use the spectrometer in short scans around the 6678 Å region to monitor the laser line as you make changes in diode laser parameters. An example of a combined HeNe discharge and diode laser spectrum, when the diode laser has been tuned to the 6678 Å He line, is shown in Fig. 7(b) . The spectrum of Fig. 7(b) was taken with a TOLD 9211 diode laser set to a temperature of 0.8 oC and current of 46.40 mA.

Once the diode laser wavelength appears to be tuned near the He line on the spectrometer, you will want to set up a laser current scan to find the He absorption line. The most convenient way to do this is to sweep the diode laser current using the Ex ternal Modulation input of the LDC-3722. When in low current mode (0-200 mA), the LDC-3722 gives a 20 mA/V current response to a voltage applied to the External Modulation input. This means the diode laser current is swept through a 20 mA range when a 1 V peak-to-peak signal is provided at the External Modulation input. The current sweep is centered at the current value displayed on the front panel of the LDC-3722. In this experiment, a swept voltage is generated using the triangular waveform of an HP 3310 A Function Generator. Since the output voltage of the function generator is typically too large for the External Modulation input of the LDC-3722, a voltage divider is used as indicated in Fig. 6. A typical curr ent sweep of the TOLD 9211 should have a range of 4 mA, therefore a 200 mV peak-to-peak triangle wave should be present at the External Modulation input. As you will show in part (4), near 0 oC and 45 mA diode current a 4 m A sweep of the current of the TOLD 9211 corresponds approximately to a 25 GHz, or 0.4 Å, tuning range of the diode laser output.

With a current sweep width of 4 mA, adjust the temperature of the diode laser in increments of 0.1 oC with different current settings near the value you used to bring the laser to the 6678 Å He line. Observe the HP 3310 A sweep voltage and the optogalvanic signal simultaneously on the digital storage oscilloscope. With some patience - and luck - you will find a temperature and current setting that will allow you to obtain an optogalvanic signal such as the one sho wn in the oscilloscope photograph of Fig. 8(a). The upper trace in Fig. 8(a) shows the 200 mV peak-to-peak triangle wave sweep signal of 10 s period; the 4.0 mA current sweep range w as centered at a diode laser current of 46.40 mA. The lower trace shows the optogalvanic signal from the voltage divider output of the HeNe discharge. The diode laser temperature in Fig. 8 was 0.8 oC. The diode laser parameters for Fig. 8(a) can only serve as a guide for your experiment since every diode laser has slightly different tuning characteristics.

While searching for an optogalvanic signal it may also be helpful to have your lab partner observe the diode laser beamspot on mirror M. The absorption by the 6678.151 Å, 31D2-->21P1 He transition is so strong that the beamspot becomes very dim when the laser passes through the absorption line. The lack of an optogalvanic signal, but the clear presence of an absorption at the He line, could indicate that something is amiss in the optogalvanic signal detection circuit.

If a patient search for an optogalvanic or absorption signal is unsuccessful, consult with the instructor. You may need to install a different diode laser to reach the 6678 Å He line. A new diode laser is easy to install and the fumes from sold ering a new diode into the diode laser mount may give you the just the mental disposition you need to succeed in your second try.

Assuming suitable optogalvanic and absorption signals are obtained, make a record of your work in the following steps.

(a) Use the oscilloscope camera to record oscilloscope traces of the sweep voltage and optogalvanic signal similar to those in Fig. 8(a).

(b) Put photodiode PD1 in place of mirror M and adjust the lock-in amplifier controls to obtain a high quality absorption spectrum on the oscilloscope such as the one shown in Fig. 8(b). Use the oscillosc ope camera to photograph traces of the sweep voltage and the absorption signal displayed on the oscilloscope.

(c) Finally, remove the sweep signal from Ch.1 of the oscilloscope and record simultaneously the optogalvanic and absorption signals as shown in the photograph of Fig. 8(c). Make a photographic record of the oscilloscope trace and plot the oscilloscope trace on the HC-100 plotter.

(5) To measure the width of the He line you observed by optogalvanic and absorption detection, it is necessary to know the change in laser frequency produced by a change in diode laser current. A crude measure of this could be obtained by setting th e diode laser to the current values corresponding to the end points of the 4.0 mA wide scans of step (3), and then measuring the laser wavelength using the 1 m spectrometer. By measuring the difference in wavelength produced by the two diode laser curren t settings, one could calibrate the tuning characteristic, in GHz/mA, of the diode laser.

A better method is to use a Fabry Perot interferometer that provides transmission maxima, equally spaced in frequency, for recording as the absorption or optogalvanic spectrum is being acquired . The spacing between the transmission maxima of the int erferometer is called the free spectral range (FSR). The free spectral range of the Coherent 216B Spectrum Analyzer shown in Fig. 6 is 300 MHz. The recording of frequency "markers" every 300 MHz, as the diode laser cu rrent is scanned, allows one to calibrate the current tuning of the diode laser.

(a) Turn off the HeNe discharge and insert the 300 MHz interferometer, as shown in Fig. 6. The diode laser focus should be adjusted to produce a minimum beamspot size near the center of the interferometer . The alignment of the interferometer is a bit tricky, but you will observe an interference pattern at photodiode detector PD2 after some effort. With PD2 connected to an oscilloscope, turn on the Coherent 251 Spectrum Analyzer Controller, and make adju stments to the interferometer to obtain broad, Fabry Perot transmission peaks on the monitor oscilloscope. The transmission peaks will be of rather poor quality because the diode laser linewidth, 100 MHz, is comparable to t he 300 MHz FSR of the interferometer. Look ahead to the Expt. XVIII using the ring dye laser to see what the transmission peaks of a Fabry Perot interferometer can look like for a very narrow linewidth (▓1/2) MHz) laser light source. Be sure the diode l aser is not scanning when adjusting the interferometer as this will make the trace on the monitor oscilloscope very difficult to interpret.

(b) Turn off the power to the 300 MHz interferometer controller. Connect PD2 to a lock-in amplifier and its output to Ch. 2 of the digital storage oscilloscope. Begin scanning of the diode laser at the 6678 Å He line and turn on the H eNe discharge with the diode laser sweep monitored on Ch. 1. If the diode laser is still sweeping the 6678 Å He line you will obtain an oscilloscope trace similar to Fig. 9(a). In Fig. 9(a) the upper trace is the diode laser current sweep signal and the lower trace shows the 300 MHz interferometer signal at PD2. The expected pattern of transmission maxima separated by the 300 MHz FSR of the interferometer is clearly modified b y the absorption process that occurs in the HeNe discharge at the 6678 Å He line. You can see now that it might have been wiser to put the interferometer before the HeNe discharge tube, rather than after it. But it is good to be aware of what can h appen when various components of a spectroscopy experiment interact in an unexpected way. Make a photographic record of this oscilloscope trace.

(c) Fig. 9(a) demonstrates that the 300 MHz interferometer is functioning properly and that the diode laser is scanning the 6678 Å He line. To eliminate the effect of absorption by the He line, simpl y turn off the HeNe discharge to obtain a trace similar to Fig. 9(b). Save and plot a record of your digital storage oscilloscope trace.

(d) By counting 30 or 40 interferometer fringes in the tuning range containing the 6678 Å He line, determine the tuning calibration, in GHz/mA, of the TOLD 9211. A value of 6.3 GHz/mA was obtained from the trace of Fig. 9(b), which is typical of the TOLD 9211.

Convert your tuning calibration in GHz/mA to Å/mA at 6678 Å.
(e) Measure the full width at half maximum (FWHM) of the He absorption peak in step 4(c). Express your FWHM in units of GHz and Å.

(f) The natural width of the 6678 Å He absorption line is much less than the result you have obtained in this experiment. Most of the measured width of the He absorption line in this experiment is due to Doppler broadening produced by the random thermal motion of He atoms in the discharge. The Doppler width (FWHM) of an absorption (or emission) line of frequency o is giv en by7

= 2*(ln2)^1/2 (o/c)  (2kT/m)^1/2

where m is the mass of the absorbing (or emitting) atom, T is the temperature, and k is Boltzmann's constant. Evaluate for the 6678 Å He line, assuming the He atoms are at room temperature, and compare it to the value obtained in this experiment.

References Beeson
1. C.E. Wieman and L. Hollberg, Rev. Sci. Instrum. 62,1 (1991).
2. J.C. Camparo, Contemp. Phys. 26, 443 (1985).
3. B. Barbieri, N. Beverini, and A. Sasso, Rev. Mod. Phys. 62, 603 (1990).
4. J.E.M. Goldsmith and J.E. Lawler, Contemp. Phys. 22, 235 (1981).
5. TOLD Series Laser Diode Product Guide, manual (Toshiba, Irvine, CA, 1991).
6. American Institute of Physics Handbook, edited by D.E. Gray, 3rd ed. (McGraw-Hill, New York, 1972), p. 7-38.
7. A. Corney, Atomic and Laser Spectroscopy  (Oxford University Press, Oxford, 1977), p. 248.
8. J. Weiss, Fun with Diode Lasers or: An Attempt at Laser Spectroscopy Using a Diode Laser, B.A. thesis, Middlebury College, 1991.

Table of Contents