The continuous tuning range of an external-cavity diode laser can be extended by making small corrections to the external-cavity length through an electronic feedback loop so that the cavity resonance condition is maintained as the laser wavelength is tuned. By maintaining the cavity resonance condition as the laser is tuned, the mode hops that typically limit the continuous tuning range of the external-cavity diode laser are eliminated. We present the design of a simple external-cavity diode laser based on the Littman–Metcalf external-cavity configuration that has a measured continuous tuning range of without an electronic feedback loop. To include the electronic feedback loop, a small sinusoidal signal is added to the drive current of the laser diode creating a small oscillation of the laser power. By comparing the phase of the modulated optical power with the phase of the sinusoidal drive signal using a lock-in amplifier, an error signal is created and used in an electronic feedback loop to control the external-cavity length. With electronic feedback, we find that the continuous tuning range can be extended to over . This occurs because the electronic feedback maintains the cavity resonance condition as the laser is tuned. An experimental demonstration of this extended tuning range is presented in which the external-cavity diode laser is tuned through an absorption feature of diatomic oxygen near .
© 2006 Optical Society of America
External-cavity diode lasers (ECDLs) have proven to be useful laboratory tools because of their spectral coverage and tunability. Some of the many applications involving ECDLs include differential absorption lidar,[, , ] molecular spectroscopy,[, ] and sources for stabilized wavelength references.[] The typical performance of an ECDL when used with a standard Fabry–Perot diode laser as the laser source includes of noncontinuous tuning, a linewidth of , and output powers of . However, one limiting factor to the usefulness of ECDLs is the limited continuous tuning range due to mode hops.
The operating wavelength of the ECDL is determined by both the wavelength of the optical feedback and the external-cavity resonance condition. An ECDL in the Littman–Metcalf external-cavity configuration[] is shown schematically in Fig. 1. Light from the diode laser is incident on an optical grating that spatially separates the different wavelength components shown schematically as λ(1) and λ(2) in Fig. 1. A retroreflector is used to pick one of these wavelength components and direct it back to the laser diode via a second reflection from the optical grating. Provided that the resonance condition of an integer number of half-wavelengths fits within the external-cavity optical path length, the optical feedback will be maintained by the external cavity. The schematic drawing on the left in Fig. 1 shows the ECDL operating at λ(1). Continuous tuning of the ECDL is achieved by first rotating the retroreflector to provide optical feedback at a different wavelength while simultaneously changing the external-cavity optical path length so that the same integer number of half-wavelengths is maintained within the cavity at the new wavelength. This is shown schematically by the drawing on the right in Fig. 1. If the retroreflector is rotated to change the wavelength of the light fed back into the diode laser but the optical path length changes so that a different integer number of half-wavelengths fits within the optical cavity, a mode hop will occur. Mode hops prevent truly continuous tuning of the ECDL to approximately the free spectral range of the external cavity and can limit the ECDLs usefulness in applications such as laser spectroscopy that require large continuous tuning ranges.
One method used to increase the continuous tuning range of an ECDL is to choose the pivot point around which the retroreflector is rotated so that change in frequency of the light fed back into the diode laser is matched to the change in the optical cavity length of the external optical cavity.[, , , , , ] This method requires very precise tolerances for the mechanical parts of the ECDL and is hard to implement. An error in locating the pivot point by as little as is enough to cause the ECDL to mode hop.[] This is well below typical machining tolerances of less than for any given part.[] A second method to increase the continuous tuning range involves always maintaining the resonance condition within the optical cavity while the ECDL is tuned.[, , , ] This can be achieved by making small corrections to the optical path length of the external cavity by applying a small correction current to the diode laser. The small correction current adjusts the carrier density and temperature of the laser diode, which affects the index of refraction and physical path length of the diode laser and allows for the small changes to the optical path length of the external cavity. An electronic feedback scheme is introduced in this paper allowing one to continuously make small corrections to the optical path length of the external cavity through a small correction current so that the resonant condition is always maintained as the laser is tuned. In this way the mode hops can be eliminated and the continuous tuning range of the ECDL can be extended.
The optical feedback from the external cavity will be highest when the ECDL is operated in the resonance condition. As the ECDL is tuned and the external cavity no longer maintains a resonance condition, the amount of feedback is less due to destructive interference in the external cavity. With less feedback to the diode laser to suppress the spontaneous emission within the diode, the power in the side modes will begin to grow, causing the side-mode suppression ratio to lessen. This fact implies that, as the external cavity no longer maintains the resonance condition, optical power in the side modes causes the overall output power of the ECDL to grow. Thus, one way to monitor how well the external cavity is maintaining the resonance condition as the ECDL is tuned is to monitor the output power of the ECDL.
In this paper, we present a method for extending the continuous tuning range of an ECDL through the use of electronic feedback. First, an error signal is generated by monitoring the output power of the ECDL. The error signal is then used to provide a small correction current to the ECDL that allows small corrections to the optical path length of the ECDL to be made. In this way, the resonance condition of the ECDL is maintained as the laser is tuned.
This paper is organized as follows. Section 2 contains a description of the ECDL and the current controller. The implementation of the electronic feedback used to extend the continuous tuning range is described in Section 3. In Section 4, a spectroscopic application based on the extended tuning range of the ECDL is presented. Finally, some brief concluding remarks are presented in Section 5.
2. External-Cavity Diode Laser
A photograph of the ECDL built at Montana State University in the Littman–Metcalf external-cavity configuration is shown in Fig. 2. A commercially available Fabry–Perot diode laser with a center wavelength of in a package is used as the source for the ECDL (Sacher Lasertechnik Model Fabry–Perot 765-30). The diode has a high-reflectivity coating on its back facet and a 1% antireflection coating on its front facet. The light from the diode is collimated using an aspheric lens with a focal length of , a numerical aperture of 0.55, and a broadband antireflection coating for the wavelength range of . The laser diode and lens are mounted in a collimating tube that holds both the laser and the collimating lens. The collimated light is incident on a holographic grating at a grazing angle of . The first-order reflection from the grating is next incident on a retroreflecting roof prism that directs the light back to the diode via a second reflection from the grating, therefore providing the optical feedback necessary to control the operating wavelength of the ECDL. The external-cavity length of the ECDL is , corresponding to a cavity free spectral range of . The ECDL is mounted on a thermoelectric cooler and the temperature is monitored using a thermistor embedded in the ECDL below the diode laser. A commercial temperature controller is used to stabilize the ECDL temperature to within . The ECDL can be tuned by rotating the roof prism about a pivot point mechanically by rotating a 3∕16-100 fine thread screw or electronically by applying a voltage to a piezoelectric transducer (PZT) stack. The maximum voltage that can be applied to the PZT is . The pivot point was chosen using the results from Ref. [] to synchronize the cavity mode and optical feedback as the grating is rotated. However, machining tolerances of (0.005 in.) and uncertainty of the location of the semiconductor chip in the can limit the accuracy of the location of the pivot.
A current controller was built based on the low-noise current controller described by Libbrecht and Hall.[] The current controller serves three purposes. First, it provides a stable dc current to the diode laser. Second, a small-signal high-speed sinusoidal modulation can be added to the dc setpoint to modulate the wavelength of the ECDL. The current controller converts a modulated voltage signal into a modulated current signal and adds this signal to the dc setpoint. Third, a small-signal current can be added to the dc current set point to correct the optical path length of the external cavity. The small-signal correction to the dc set point is proportional to the voltage applied to the small-signal dc input port.
The ECDL has a center wavelength of with over of coarse tuning. The ECDL has a lasing threshold of with a slope efficiency of 0.07 . Tuning of the laser can be achieved by changing the laser drive current producing a small change in the external-cavity length. The measured current tuning response for this ECDL is 1.25 . Tuning the ECDL is also achieved electronically by applying a voltage to the PZT and is shown in Fig. 3. This measurement used a computer to first set a voltage supplied to the PZT and then the operating frequency of the ECDL was measured with a Burleigh WA-1500 wavemeter with a resolution. As the voltage is applied to the PZT, the laser begins tuning with a tuning rate of . However, once the laser has tuned , the changing frequency of the light fed back to the diode laser can no longer maintain the same number of half-wavelengths associated with the optical cavity length and the ECDL goes through a mode hop. A plot of the tuning as a function of voltage for the ECDL is shown in Fig. 4 for a larger applied voltage range. In Fig. 4 we see the mode hops associated with the external-cavity laser as well as two mode hops associated with the cavity formed by the front and back facets of the diode laser. The facet mode hops cause the ECDL to jump approximately . One clear problem with tuning an ECDL is that mode hops limit the actual frequency ranges in which the ECDL can be operated. This becomes a problem, for example, if one wants to use the ECDL for spectroscopic purposes where large continuous tuning ranges are needed.
The output power of the ECDL will change as the laser is tuned. When the external cavity maintains a resonance condition, the optical feedback to the diode laser is a maximum. In this case, the optical feedback will allow suppression of the spontaneous emission yielding a high side-mode suppression ratio. When the ECDL is tuned and the external cavity no longer maintains a resonance condition, the optical feedback to the diode laser will decrease due to destructive interference within the external cavity. This will cause less power to be fed back into the diode and the spontaneous emission will no longer be suppressed. The growth of the spontaneous emission will have two measurable effects, including the growth of side modes, thus decreasing the side-mode suppression ratio, and an increase in the output power of the ECDL. Using a Burleigh wavemeter, an optical spectrum analyzer (OSA), and a photodetector, the output of the ECDL was studied in the following way. First, a computer sets a voltage to the PZT used to tune the ECDL. Next, the operating frequency, an optical spectrum, and the voltage of an external photodiode were recorded. A new voltage was applied to the PZT and these measurements were repeated. A plot of the operating frequency and photodiode voltage as a function of the applied voltage to the PZT is shown in Fig. 5. The solid curve represents the operating frequency of the ECDL while the dashed curve represents the photodiode voltage, which is proportional to the ECDL output power. In Fig. 5 we see that, as the PZT voltage increases, the laser operating frequency decreases as well as the output power of the ECDL until a mode hop occurs, at which time both the operating frequency and the output power experience a discontinuity as the ECDL experiences a mode hop. As the PZT voltage increases further, both the operating frequency and the optical power decrease again until a second mode hop occurs. Because an OSA is used to capture an optical spectrum for each individual setting of the PZT voltage, a study of the evolution of the optical structure can be undertaken. A plot of the optical spectrum is shown in Fig. 6 when the laser is tuned to the left of a mode hop labeled as point A in Fig. 5. In Fig. 6 the side-mode suppression ratio is measured to be . A plot of the optical spectrum is shown in Fig. 7 to the right of the mode hop labeled B in Fig. 5. In Fig. 7 the amplified spontaneous emission is seen near the main laser mode. The effects of the spontaneous emission are to decrease the side-mode suppression ratio to while increasing the optic output power of the ECDL as seen in Fig. 5 by 4%. These results were obtained by tuning the ECDL with an increasing voltage applied to the PZT. Similar results occurred when the ECDL was tuned using a decreasing voltage applied to the PZT.
The growth in the spontaneous emission results from the mismatch between the external-cavity mode and the frequency of the light fed back to the diode laser as the grating is rotated. When the frequency of light fed back to the diode from the external cavity matches the external-cavity mode condition, optical power in the external cavity is a maximum and the optical feedback can suppress the spontaneous emission. As the ECDL is tuned so that the frequency of light fed back into the diode does not match the external-cavity mode, the optical power in the external cavity decreases thus decreasing the amount of light fed back into the laser diode. The decreased optical power fed back into the diode laser allows the spontaneous emission to begin growing, thus causing the increase in the output power of the ECDL while causing the side-mode suppression ratio to decrease.
Looking at the optical spectra for each PZT voltage, a plot of the side-mode suppression ratio as a function of applied voltage to the PZT can be generated. A plot of the side-mode suppression ratio as a function of the voltage applied to the PZT is shown in Fig. 8. The solid curve represents the side-mode suppression ratio while the dashed curve represents the voltage seen by a photodiode monitoring the output power of the ECDL. As the ECDL is tuned via the PZT, the side-mode suppression ratio is increasing as the external-cavity length begins to match the resonance condition and correspondingly the optical power decreases as the spontaneous emission is suppressed.
One way to keep the external cavity at the proper length to maintain the resonance condition as the ECDL is tuned is to monitor the output power of the ECDL. The external-cavity length can be maintained in the resonance condition by adjusting the optical path length of the external cavity so that the output power of the ECDL remains constant. We will use this idea to extend the continuous tuning range of an ECDL.
3. Extending the External-Cavity Diode Laser Continuous Tuning Range
The ECDL will experience a mode hop when the change in the optical path length of the external cavity does not match the changing wavelength fed back into the diode laser due to the first-order reflections from the optical grating. One way to make small corrections to the optical path length of the external cavity is to make small changes to the laser drive current causing a change in the index of refraction due to a change in the carrier density as well as a change in the physical length of the diode due to thermal expansion. We will exploit this method of correcting the optical path length of the external cavity to extend the continuous tuning range of the ECDL.
A schematic of the experimental setup used to extend the continuous tuning range is shown in Fig. 9. The output from the ECDL described above is incident on a first beam splitter. Light reflected from the first beam splitter is incident on a photodetector while light transmitted through the first beam splitter is directed to a second beam splitter. The second beam splitter is used to send part of the light to a Burleigh WA-1500 wavemeter or an OSA via a single-mode optical fiber while the remaining light is sent through a scanning flat-plate Fabry–Perot interferometer. The Fabry–Perot interferometer has a free spectral range of with a measured finesse of 70. The scanning Fabry–Perot interferometer, which has a frequency resolution of , was used to confirm mode-hop-free operation of the ECDL as the laser is tuned. The current controller described above is used to provide a forward current to the ECDL. The dc set point of the current driver was set at throughout all the experiments described in this paper. The drive current resulted in an output power of . A function generator was used to provide a small sinusoidal modulation to the dc set point. The modulation was set at and provided a sinusoidal additional current at with a peak current excursion of . Thus, the sinusoidal modulation will broaden the laser linewidth by = . The broadening of the ECDL laser linewidth does not present a problem when one is trying to measure spectral features of the order of gigahertz. The current is also 47 dB below the dc drive current set point, so intensity variations due to the modulation can be neglected when considering the output power of the ECDL. The signal from the function generator was also used as a reference signal for a lock-in amplifier. The lock-in amplifier compares the phase of the reference signal from the function generator with the phase of the signal generated at the detector at the frequency set by the function generator and outputs a voltage related to this phase difference. The output of the lock-in amplifier is sent to a differential amplifier and this signal is compared to a reference voltage. The output of the differential amplifier is proportional to the difference between the lock-in amplifier and the reference signal and is used as the input to the small-signal correction port of the current driver.
To understand how the electronic feedback loop can be used to extend the tuning range of the ECDL, consider the output power as a function of the tuning voltage applied to the PZT shown in Fig. 5. A small modulation signal is applied to the current set point causing the ECDL to tune. This small tuning will cause a small modulation in the output power of the ECDL. The lock-in amplifier compares the phase of the rf signal used to modulate the drive current with the phase of the modulated signal produced by the modulated output power of the ECDL as seen by a photodiode. The lock-in amplifier produces an output voltage related to this phase difference. The modulated output power of the ECDL will change when the ECDL is tuned near a mode hop due to the jump in optical power as seen in Fig. 5. This change in the modulated signal due to the output power will have a different phase relationship compared with the rf signal used to modulate the drive current, thus causing the voltage output of the lock-in amplifier to change. In this way the voltage output of the lock-in amplifier is used to monitor when the ECDL approaches a mode hop due to a mismatch in the external-cavity resonance condition. The output voltage of the lock-in amplifier can be thought of as an error signal that is next conditioned with a differential amplifier. This conditioning consists of subtracting a dc offset voltage and provides the appropriate gain. The output of the differential amplifier is used to supply a voltage to the small-signal port of the current controller. The voltage is converted into a current and this correction current is added to the current set point. The correction current changes the optical path length of the external cavity to bring the external cavity back into a resonance condition, and in this way mode hops are suppressed and the tuning range of the ECDL is extended.
A plot of the operating frequency as a function of the voltage applied to the PZT is shown in Fig. 10. Without feedback, external-cavity and facet mode hops are clearly visible. When the electronic feedback is connected, the mode hops are suppressed. The ECDL can now tune over as demonstrated in Fig. 10.
The broadening of the laser linewidth due to the sinusoidal modulation needed to create the error signal was estimated to produce a laser linewidth of . This estimation results from realizing that the current tuning response was while the sinusoidal modulation to the current had a peak current excursion of . The above experiments were repeatable down to a maximum current excursion of 0.3 μA corresponding to a linewidth of . This narrow linewidth, while broader than a typical ECDL linewidth, is very narrow compared to spectroscopic features of many molecules and should not limit the usefulness of tunable ECDLs for spectroscopic experiments.
4. Absorption Measurements of Diatomic Oxygen
A useful application of the extended tuning range of the ECDL is spectroscopy and its applications to optical sensors. An experimental setup used for differential absorption measurements of diatomic oxygen in the atmosphere is shown in Fig. 11. The output from the extended tuning range ECDL as shown in Fig. 9 is incident on a beam splitter with part of the light directed to a wavemeter (Burleigh WA-1500) while the remaining light is incident on a second beam splitter. The second beam splitter sends approximately 4% of the light to a reference detector with the remaining 96% of the light passing through to a transmission detector. The light detected by the transmission detector travels an additional when compared with the light collected by the reference detector. Dividing the signal from the transmission detector by the signal from the reference detector produces a normalized transmission measurement. A computer is used to set the voltage of the PZT. The computer then reads the wavemeter, the reference detector, and the transmission detector and records this information. A plot of the normalized transmission is shown in Fig. 12. The circles, triangles, and squares represent three different normalized transmission measurements while the solid curve represents the theoretical predictions using the HITRAN database.[] Good agreement between the measured and expected values is shown.
5. Concluding Remarks
A method has been presented for extending the continuous tuning range of an external-cavity diode laser (ECDL) through the use of electronic feedback. The electronic feedback forces the laser to maintain a maximum output power as the laser is tuned so that the resonance condition within the external optical cavity is maintained. By maintaining the resonance condition within the external optical cavity as the ECDL is tuned, mode hops are suppressed and the continuous tuning range of the ECDL is extended.
We have demonstrated a continuous tuning range of greater than by implementing the electronic feedback scheme presented in this paper. Extending the continuous tuning range will allow ECDLs to be used for many spectroscopic applications, such as molecular spectroscopy, as demonstrated by the diatomic oxygen absorption measurements demonstrated in Section 4.
This work was supported by the U.S. Department of Energy (DOE) under award DE-FC26-04NT42262. However, any opinions, findings, conclusions, or recommendations expressed herein are those of the author(s) and do not necessarily reflect the views of DOE.
1. K. S. Repasky, J. A. Shaw, J. L. Carlsten, M. D. Obland, L. S. Meng, and D. S. Hoffman, “Diode laser transmitter for water vapor DIAL measurements,” in Proceedings of the 2004 IEEE International Geoscience and Remote Sensing Symposium (IEEE, 2004), Vol. 3, pp. 1947–1950. [CrossRef]
2. J. L. Machol, T. Ayers, K. T. Schwenz, K. W. Koenig, R. M. Hardesty, C. J. Senff, M. A. Krainak, J. B. Abshire, H. E. Bravo, and S. P. Sandberg, “Preliminary measurements with an automated compact differential absorption lidar for the profiling of water vapor,” Appl. Opt. 43, 3110–3121 (2004). [CrossRef] [PubMed]
3. M. D. Obland, L. S. Meng, K. S. Repasky, J. A. Shaw, and J. L. Carlsten, “Progress toward a water-vapor differential absorption lidar (DIAL) using a widely tunable amplified diode laser source,” U. N. Singh , ed., Laser Remote Sensing for Environmental Monitoring VI, Proc. SPIE 5887, 205–215 (2005).
4. Q.-V. Nguyen, R. W. Dibble, and T. Day, “High-resolution oxygen absorption spectrum obtained with an external-cavity continuously tunable diode laser,” Opt. Lett. 19, 2134–2136 (1994). [CrossRef] [PubMed]
5. D. Aumiler, T. Ban, and G. Pichler, “High resolution measurements of the pressure broadening and shift of the rubidium 52S1∕2–62P3∕2 line by argon and helium,” Phys. Rev. A 70, 032723(1)–032723(5) (2004). [CrossRef]
6. G. J. Koch, A. L. Cook, C. M. Fitzgerald, and A. N. Dharamsi, “Frequency stabilization of a diode laser to absorption lines of water vapor in the 944-nm wavelength region,” Opt. Eng. 40, 525–528 (2001). [CrossRef]
11. G. W. Switzer, “Semiconductor laser transmitter for water vapor lidar on Mars,” Ph.D. dissertation (Montana State University, Bozeman, Mont., 1998), pp. 37–49.
12. L. Nilse, H. J. Davies, and C. S. Adams, “Synchronous tuning of external cavity diode lasers: the case for an optimum pivot point,” Appl. Opt. 38, 548–553 (1999). [CrossRef]
13. T. Nayuki, T. Fujii, K. Nemoto, M. Kozuma, M. Kourogi, and M. Ohtsu, “Continuous wavelength sweep of external cavity 630 nm laser diode without antireflection coating on output facet,” Opt. Rev. 5, 267–270 (1998). [CrossRef]
14. V. P. Gerginov, Y. V. Dancheva, M. A. Taslakov, and S. S. Cartaleva, “Frequency tunable monomode diode laser at 670 nm for high resolution spectroscopy,” Opt. Commun. 149, 162–169 (1998). [CrossRef]
15. C. Petridis, I. D. Lindsay, D. J. M. Stothard, and M. Ebrahimzadeh, “Mode-hop-free tuning over 80 GHz of an external cavity diode laser without antireflection coating,” Rev. Sci. Instrum. 72, 3811–3815 (2001). [CrossRef]
16. S. Mattori, T. Saitoh, S. Kinugawa, H. Kameyama, S. Ozaki, and J. Shirono, “A mode hopping suppressed external-cavity semiconductor laser using feedback control,” IEICE Trans. Electron. E85-C98–103 (2002).
17. K. G. Libbrecht and J. L. Hall, “A low-noise high-speed diode laser current controller,” Rev. Sci. Instrum. 64, 2133–2135 (1993). [CrossRef]
18. L. S. Rothman, A. Barbe, D. C. Benner, L. R. Brown, C. Camy-Peyret, M. R. Carleer, K. Chance, C. Clerbaux, V. Dana, V. M. Devi, A. Fayt, J.-M. Flaud, R. R. Gamache, A. Goldman, D. Jacquemart, K. W. Jucks, W. J. Lafferty, J.-Y. Mandin, S. T. Massie, V. Nemtchinov, D. A. Newnham, A. Perrin, C. P. Rinsland, J. Schroeder, K. M. Smith, M. A. H. Smith, K. Tang, R. A. Toth, J. Vander Auwera, P. Varanasi, and K. Yoshino, “The HITRAN molecular spectroscopic database: edition of 2000 including updates through 2001,” J. Quant. Spectrosc. Radiat. Transfer 82, 5–44 (2003). [CrossRef]