DIODE LASER WITH EXTERNAL GRATING-BASED CAVITY

CALIFORNIA POLYTECHNIC STATE UNIVERSITY, SAN LUIS OBISPO DEPARTMENT OF PHYSICS DIODE LASER WITH EXTERNAL GRATING-BASED CAVITY A Senior Project SUBMI...
Author: Stuart Casey
2 downloads 0 Views 319KB Size
CALIFORNIA POLYTECHNIC STATE UNIVERSITY, SAN LUIS OBISPO DEPARTMENT OF PHYSICS

DIODE LASER WITH EXTERNAL GRATING-BASED CAVITY

A Senior Project SUBMITTED TO THE DEPARTMENT OF PHYSICS in partial fulfillment of the requirements for the degree of BACHELOR OF SCIENCE

By Cristian Alonso Heredia San Luis Obispo, California 2004

DIODE LASER WITH EXTERNAL GRATING-BASED CAVITY

A SENIOR PROJECT APPROVED FOR THE DEPARTMENT OF PHYSICS

By

c by Cristian Alonso Heredia 2004

All Rights Reserved.

Contents

List Of Figures

v

Abstract

1

1 Introduction

2

2 Theory 2.1 Diode Lasers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2 Rubidium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3 Littrow Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

3 3 5 6

3 The Apparatus

7

4 Analysis 4.1 Current Tuning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2 Temperature Tuning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3 Piezo Tuning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

8 8 8 9

5 Procedure

11

6 Conclusion

12

References

13

iv

List Of Figures

1.1

Rb Cell . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

2

2.1 2.2 2.3 2.4 2.5 2.6

Band structure of semiconductor . . . . . Lateral confinement of the laser current in Light emission from a semiconductor. . . Light emission from a semiconductor. . . 87 Rb D1 and D2 transition lines . . . . . . Blazed grating . . . . . . . . . . . . . . .

. . . . . .

3 4 4 5 5 6

3.1

Apparatus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

7

4.1 4.2 4.4 4.5 4.3 4.6 4.7

Current Tuning . . . . . . . . Concentric Circles . . . . . . Temperature vs. Wavelength Piezo Tuning . . . . . . . . . Temperature Tuning . . . . . Etalon at 40V . . . . . . . . . Etalon at 60V . . . . . . . . .

6.1

Fluorescence of Rb atoms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. a . . . .

. . . . . . .

. . . . . . . . . . . . semiconductor laser. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. 8 . 8 . 9 . 9 . 10 . 10 . 10

v

Abstract We have developed an external cavity for a diode laser via the Littrow configuration. In conjunction with this configuration and varying diode temperature we were able to fine tune a 785 nm diode laser down to 780 nm. The laser output has been diagnosed at several stages using a real-time, computer driven spectrometer, providing insight into the operation of both a free running and stabilized diode laser. At this wavelength we were able to observe the fluorescence of the Rb D line at 780nm.

1

Chapter 1 Introduction

Most free running diode lasers are fine for the everyday use; such as a laser pointer in

was a diffraction grating which was used to find the Littrow configuration (see Section 2.3).

the classroom setting. However, these types

My first task was to find the above men-

of lasers would not be adequate for use in the

tioned configuration, once found the intensity

laboratory setting.

of the output laser beam would increase. This

Diode lasers are semiconductor based lasers.

could be measured by using a power meter,

A diode laser is simply a light emitting diode

or simply placing an infrared card in front of

(LED) with a feedback cavity. Diode lasers un-

the output beam and witnessing a dramatic

like LED’s will not operate below a threshold

increase in intensity. After this was accom-

current. Consequently, only when the thresh-

plished the laser was to be placed at the proper

old current is reached will the diode commence

wavelength; using the thermoelectric cooler to

lasing. The first diode lasers appeared in CD

modulate frequency. The USB spectrometer

players, and were massed produced by frontier

was used to identify the laser wavelength. With

electronic franchises.

this set up I could alter the frequency until the

As a consequence of the mass production

spectrometer read the proper wavelength. Al-

of diode lasers one can always find a surplus

though, the true test would be to take the out-

on the market. They tend to lase in the range

put beam and traverse it through the Rubid-

of 780 nm to 790 nm, serendipitous, Rb has a

ium gas and observe fluorescence. This turned

resonance in this range. In addition of being

out to be the crux of the endeavor.

a low cost laser, with the aid of frequency stabilization, diode lasers have been known to be locked at a specific absorption line for hours. Exploiting this coincidence was my endeavor. At my convenience I had a couple of diode lasers (785 nm), a Fabry-Perot etalon, some Piezo crystals, a USB spectrometer, and a steady current source. The diode laser was mounted

Figure 1.1: Laser beam is traversing Rb cell and being analyzed by oscilloscope.

into an apparatus that was coupled with a thermoelectric cooler (see Chap. 3) this was used to cool, heat, and keep the diode at a steady temperature. Also, attached to the apparatus

2

Chapter 2 Theory

2.1

Diode Lasers

is located near the middle of the gap for a semiconductor and because the energy gap is small,

Semiconductor lasers are small efficient laser devices with typical dimensions of less than a millimeter. They can have an efficiency of up to 30%, while a typical HeNe gas laser can possess efficiencies as low as 0.025%. Semiconductor lasers operate on wavelengths ranging from 600 to 1550 nm, depending upon the material of the laser medium. Semiconductor lasers generally operate on a continuous wavelength basis. The typical gain bandwidth, corresponding to the recombination emission

appreciable numbers of electrons are thermally excited form the valence band to the conduction band[1]. There are many empty levels in the conduction band; therefore, a small applied potential difference can easily raise the energy of the electrons in the conduction band, resulting in a moderate current. Because electrons’ being thermally excited across the narrow gap is more probable at higher temperature, the conductivity of semiconductors increases rapidly with temperature.

line width, which is of the order of 20 nm, but the laser bandwidth can be significantly reduced by cavity or resonator effects.

Al-

though most semiconductor lasers can be optically pumped, electrical pumping is much more practical. For this reason, all commercial semiconductor lasers are operated by passing an electrical current though the laser medium. The small size of semiconductor lasers is made possible by the extremely large population inversion densities that can be produced on a steady state basis by locating two specially doped semiconducting materials directly adjacent to each other to form a junction and placing a forward-bias voltage between them.

Figure 2.1: Band structure of a semiconductor at ordinary temperature. The energy gap is much smaller than an insulator , and many electrons occupy states in the conduction band.

These two doped materials include one with an excess of electrons (n-doping) and one with an excess of holes (p-doping). The band structure of a semiconductor is

The first useful semiconductor devices were composed of GaAs material that emit in the near infrared at 800 nm.

Since then, laser

shown in Figure 2.1. Because the Fermi level 3

emission wavelengths have been extended fur-

will dominate because of homogenous broad-

ther in to the infrared to 1500 nm, based upon

ening. This particular wavelength can be al-

InP/InGaAsP-layered semiconductor material

tered by modulating the temperature, gener-

for use in optical communications. Laser wave-

ally, the lower the temperature the lower the

lengths have also been extended to 630 nm

wavelength.

based upon GaAs/AlGaAs-layered material for

Light emission and absorption in semicon-

applications such as laser pointers and bar-

ductor is similar to light emission and absorp-

code readers.

tion by gaseous atoms, except that in the discussion of semiconductors discrete atomic energy levels must be replaced by bands. As shown in Figure 2.3, an electron excited electrically into the conduction band can easily recombine with a hole. As this recombination takes place, a photon of energy Eg is emitted.[3]

Figure 2.2: Lateral confinement of the laser current in a semiconductor laser. The laser cavity can be produced in one of two ways. The first is the typical Fabry–Perot cavity, but in this case the mirrors are produced at the ends of the laser gain medium by cleaving the ends of the semiconductor crystal perpendicular to the optical axis of the crystal[2]. Due to the high index of refraction of the material, typically n1 =1.5, the reflectivity at those cleaved interfaces is of the order of 30%. The reflectivity can be altered by adding dielectric coating to the surfaces. In commercial semiconductor lasers the laser cavity length typically ranges from 0.2 to 0.1 mm. This cleaved type of cavity produces a relatively broad spectrum laser output that can consist of many longitudinal modes, although typically a single mode at the wavelength of the highest gain

Figure 2.3: Light emission from a semiconductor. The recombination emission broadening within the laser gain region, the junction, is homogeneous. Therefore, the laser wavelength will generally occur at the peak of the recombination emission profile unless special frequency selective techniques are used to select the desired laser frequency or wavelength[4]. Figure 2.4 shows a typical output of a semiconductor laser versus current flowing through the laser. At low current where the absorption exceeds the gain, the recombination emission from the junction region increases linearly with current. At the threshold current, Ith , typically in mA, where the gain exceeds the losses previously mentioned, laser action begins and

4

the output increases dramatically, as shown in Figure 2.4.

2.2

Rubidium

Rubidium is the 37th element in the periodic table, and possesses an atomic mass of 85.468 amu. The

87

Rb isotope has a 72% natural

abundance. This isotope has a D2 line from the decay of the 52 P3/2 state to the 52 S1/2 state, corresponding to 3.85 x 1014 Hz or 780 nm[5]. See Figure 2.5.

Figure 2.4: Light emission from a semiconductor.

Figure 2.5: 87 Rb D1 and D2 transition lines and their respective quantum states.

5

2.3

Littrow Configuration

In the Littrow configuration first-order light which is diffraction from the grating is coupled back into the laser diode (see Figure 3.1), and directly reflected light forms the output beam[6]. This particularly simple and effective configuration can be used with an inefficient grating to reduce feedback and increase output power, thus, improving overall efficiency. This configuration is analogous to the addition of standing waves on an oscillating rope. The configuration is made possible with the implementation of blazed gratings (see Figure 2.6). Blazing is the technique of shaping individual grooves so that the diffracted envelope maximum shifts into another order. While the diffraction envelope is shifted by the shaping of the individual grooves, the interference maxima remain fixed in position. Their positions are determined by the grating equation, in which angles are measured relative to the plane of the grating. The result is that the diffraction maxima now favors a principal maximum of higher order (|m|>0), and the grating redirects the bulk of the energy where it is most useful[7]. Figure 2.6: In a blazed grating the diffraction envelope maximum at β = 0 and the zerothorder interference at m = 0 are seperated.

6

Chapter 3 The Apparatus

The apparatus consists of a Sanyo diode (DL-7140-201), mounted in a collimating tube (Thorlabs LT230P5-B) fixed to an aluminum block. The tuning diffraction grating consists of 1200 lines/mm on a 10 × 10 × 3mm3 substrate.

The grating is attached to a modi-

fied piece of aluminum block, which provides vertical and horizontal grating adjustment. A piezoelectric transducer disk under one of the pivot arms is used to modify the cavity length for fine frequency tuning, and may also be used to vacillate the frequency with an AC resonating system.

A temperature sensor (Jameco

206981 0.55W @ 25o C thermistor) and Peltier thermoelectric cooler (30 × 30 × 3.3mm3 ) were used for temperature control. A constant current supply (Thorlabs LDC500) was used to power the laser diode. An infrared card (Radio Shack) used to detect the main laser beam output. A USB spectrometer (Ocean Optics USB2000 Miniature Fiber Optic Spectrometer) was used to detect general wavelength. A Fabry-Perot etalon with

Figure 3.1: Littrow configured with fixed output beam.

a free spectral range of 30 GHz was used to detect fine shifts in frequency. A CCD camera was used to detect infrared radiation from Rubidium cell.

7

Chapter 4 Analysis

4.1

Current Tuning

Current primarily changes intensity and is too erratic and course for frequency modulating, this erratic movement is due to mode hopping. Generally, as the current increases so does the intensity, as is apparent in Figure 4.1.

Figure 4.2: Concentric circles outputted from etalon.

4.2

Temperature Tuning

Temperature is a good way to modulate freFigure 4.1: 54 mA versus 68 mA, The 68 mA beam is the more intense of the two light sources, as is expected. Figure’s like Figure 4.1 were acquired from the data of a figure similar to Figure 4.2. A horizontal strip of information is taken from the center of the data. The data contains pixel number and a relative intensity number.

quency, similar to current tuning it is suspect to mode hopping. Unlike Current Tuning, Temperature Tuning can be used to cover broad wavelengths. With respect to the lasers natural wavelength at room temperature, it can cover ± 5 nm, see Figure 4.3.

An obvious

advantage to this is the ability to change frequency without changing alignment, see Figure 4.4.

8

λ vs T @70.1 mA 789 ♦ 788 data ♦ 787 786 ♦ 785 λ(nm) 784 783 ♦ ♦ 782 781 ♦ 780 15 20 25 30 35 40 45 T emperatureo C

4.3

Piezo Tuning

Piezo tuning is ideal for small frequency shifts. When Piezo tuning is performed a breathing of the resonance line can be observed via the Fabry-Perot etalon. The breathing effect, corresponds to a shift in frequency, this shift can be extrapolated back from the Free Spectral Range of the etalon. See Figures 4.6 and 4.7. For example if the concentric circles were to converge all the way to next set of concentric circles, it would correspond to a shift of 0.061 nm with respect to 780 nm and a Free Spectral

Figure 4.4: Temperature vs. Wavelength. Increase in wavelength is almost linear with respect to an increase in temperature.

Range of 30 GHz.

Figure 4.5: 40 V versus 60 V

9

Constant Current 70.1 mA 2500

.307M Ω @ 17.2o C ? .211M Ω @ 25.0o C o 4 .175M Ω @ 30.1 C 2000 .119M Ω @ 44.3o C × 44 × ?? 4 4 × ? ?? 1500 × 4 4 × Relative × × ? 4 Intensity ? 4 ? 1000 × × ? 4 4 ? × 4 ? × × 4 500 ? ? 4 × ? × 4 × ? 4 ? 4 ×× 4 ? × 4 ×4 ?4 44 ×4 ×4 44444444 ×4 ×4 ×4 ?????????????????? ×4 ×4 ×4 ×4 ×4 ×4 ×4 ×4 × ×4 ×4 ×4 ×4 ×4 ×4 ×4 ×4 4 ???????????????????? × ×4 ×4 ×4 ×4 ×4 ×4 ×4 ×4 ×4 ×××××××××××????????? ×4 ×4 ×4 × 04 775 780 785 790 795 λ(nm)

Figure 4.3: Resistance is inversely related to temperature ,ergo, from the graph the higher the temperature the higher the wavelength.

Figure 4.6: Image of the etalon with the voltage set at 40V through the piezoelectric crystal. Note: Figures 4.6 and 4.7 show a direct shift in frequency when a voltage is applied to the piezoelectric crystal. The center spot has changed in intensity, in addition to the shifting of concentric circles.

Figure 4.7: Image of the etalon with the voltage set at 60V through the piezoelectric crystal.

10

Chapter 5 Procedure

Here we have the procedure in which the

to ensure proper wavelength was to wit-

apparatus was setup and primed for data re-

ness radiation from the Rb cell.

trieval and absorption in the Rb D2 resonance

beam was set to traverse the Rb cell, and

line.

the CCD camera was set orthogonally to

1. Placed laser in Littrow configuration. Current was dropped just below threshold current, so that two visible infrared spots were noticeable on an infrared card. Using the two control knobs, one for horizontal alignment and the other for vertical, the two faint beams were converged upon each other to form one beam. Pursuant with combining the individual beams

The

pick up scattered infrared radiation from Rb atoms. Fine adjustments were made with the vertical and horizontal knobs, until absorption and emission were witnessed. Meanwhile cautious not to modulate the knobs too much and, therefore, displacing the laser out of Littrow configuration. 4. Post measurements. After proper wave-

the intensity of the converged beams in-

length was achieved the frequency was

creased. Consequently the threshold cur-

swept by the aid of a Piezo electric crys-

rent of the diode laser decreased.

tal between one of the pivot arms. The

2. Proper wavelength was calibrated. The current was set to at least to at least three times the threshold current, but was set below the manufactured suggested maximum current for the diode laser. The temperature controller was set to cool the diode laser to a wavelength of 780 ±1 nm, with respect to the USB spec-

Piezo crystal was perturbed by exposing it to a triangular wave. The beam was then traversed through an etalon. The frequency shift was noted by calculating the shift in concentric circles outputted by the etalon, and extrapolating back to the relationship of the free spectral range of the etalon.

trometer. 3. Fine tuned wavelength. The USB spectrometer was good for broad wavelengths, but it had become superfluous at this stage in the tuning process since it had an uncertainty of ±2 nm. The only way

11

Chapter 6 Conclusion

A more complete understanding of how the properties of current, temperature, and external cavity govern a diode laser has been achieved. The current does not affect the laser much except for intensity, and some small fluctuations in frequency. Note: this is enough to perturb the laser from lasing at the proper resonant frequency. Temperature tuning is fine for large changes in frequency, but it is not the best method for fine tuning due to mode hopping. Once resonant frequency is obtained Piezo tun-

Figure 6.1: Fluorescence of Rb atoms, the fluorescence is caused by the frequency stabilized diode laser beam passing through the cell.

ing appears to be the best method of frequency shifting since it involves changing an external cavity length by small increments, on the order of nanometers, and in turn the frequency is recalibrated slightly. This latter method has many advantages, in addition to the cavity not being altered by macroscopic changes, or the laser being susceptible to mode hopping, the signal sent to the Piezo crystal can be altered accordingly to achieve a multitude of results.

12

References

[1] M. Ott. Capabilities and reliability of led’s and laser diodes. Swales Aerospace: Technology Validation Assurance Group, 1997. [2] C. J. Hawthorn, K. P. Weber, and R. E. Scholten. 10.1063/1.1419217, 2001.

American Institute of Physics,

[3] R.A. Serway and R. J. Beichner. Physics for Scientist and Engineers with Modern Physics. Harcourt College Publishers, Florida, 2000. [4] W.T. Silfvast. Laser Fundamentals. Cambridge: Cambridge University Press, 1996. [5] D. A. Steck. Rubidium 87 D line data. New Mexico: Los Alamos National Laboratory, 2001. [6] D. W. Preston and C. E. Wieman. Doppler-free saturated absorption spectroscopy: Laser spectroscopy. California : California State University, Hayward. and Colorado: University of Colorado, Boulder, 1998. [7] F.L Pedrotti and L.S. Pedrotti. Introduction to Optics. New Jersey: Prentice-Hall, 1993.

13

Suggest Documents