IEEE PHOTONICS TECHNOLOGY LETTERS, VOL. 20, NO. 4, FEBRUARY 15, 2008
303
Frequency Manipulation of THz Bound-to-Continuum Quantum-Cascade Lasers Joshua R. Freeman, Christopher Worrall, Vasilis Apostolopoulos, Jesse Alton, Harvey Beere, and David A. Ritchie
Abstract—The laser emission frequencies of existing designs are controlled by uniform thickness transformations. Experimental results from a total of ten wafers are presented demonstrating a predictable shift in the emission wavelength concurrent with an anticipated alteration in electrical transport. Wafers based on a design around 2.9 THz produced a frequency shift of 0.35 THz. Furthermore, a 2.5% thickness reduction in our 2-THz design saw a 0.1-THz frequency shift and changes in electrical transport predicted by simulations. Index Terms—Frequency control, intersubband transitions, quantum cascade laser (QCL), semiconductor laser, terahertz (THz).
I. INTRODUCTION ERAHERTZ (THz) quantum-cascade lasers (QCLs) [1] operate on an intersubband transition allowing the energy of the transition to be engineered over a range of frequencies, presently from 4.8 [2] to 1.6 THz [3]. This ability to control the frequency of the QCL coupled with high output powers and narrow linewidth makes THz QCLs highly suited to heterodyne spectroscopy applications [4], of particular importance in the THz region where many molecules and compounds have distinct absorption lines [5] and there is a lack of compact highpower sources. For such applications where a specific frequency is required, it is desirable to have a method to alter the emission frequency of a QCL to match that of the application. Small shifts in emission frequency can be achieved with the distributed feedback technique already demonstrated with THz QCLs [6], [7]. The range of this technique, however, is limited by the width of the gain curve. An alternative approach is to modify the gain curve by changing the design of the structure. The design of a THz QCL for a specific frequency, however, is not trivial and it would be convenient to adjust a known design in a simple and predictable way rather than restart the design process. Here we present a
T
Manuscript received July 27, 2007; revised October 4, 2007. This work was supported in part by the Department of Trade and Industry through the Technology Programme Opto-Electronics and Disruptive Electronics Theme, in part by Project Long Wavelength, Ultra-low Threshold Current THz Quantum Cascade Lasers, in part by Project TP/3/OPT/6/I/16377, and in part by the European Community through the PASR 2004 Project TERASEC. J. R. Freeman, C. Worrall, V. Apostolopoulos, H. Beere, and D. A. Ritchie are with the Semiconductor Physics Group, Cavendish Laboratory, University of Cambridge, Cambridge CB3 0HE, U.K. (e-mail:
[email protected]). J. Alton is with Teraview Ltd., Cambridge CB4 0WS, U.K. Color versions of one or more of the figures in this letter are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/LPT.2007.913340
Fig. 1. Emission frequency of ten devices taken from eitht wafers against period thickness. The connected (red) points are the prediction from simulations; square (blue) marks are wafers with the three devices from V193 marked as triangular points. The inset shows a typical X-ray rocking curve from a QCL wafer, the spacing of the satellite peaks is used to determine the period length.
method to achieve this by adjusting each well and barrier by the same fractional amount, causing a change in the period length of a bound-to-continuum THz QCL resulting in a predicable shift in the emission frequency. There has been some use of this technique for mid-infrared (IR) QCLs [8], however, this is the first systematic theoretical and experiment study. The range over which we tune the frequency highlights the robustness of the bound-to-continuum design. The effect of this technique on other design types, such as phonon depopulation designs [9], should be a matter for further investigation. II. EXPERIMENT All wafers were grown on a Veeco ModGen II molecular beam epitaxy reactor on semi-insulating GaAs substrates. The period length of each wafer was determined by high-resolution X-ray diffraction, as shown in the inset to Fig. 1. The spacing of satellite peaks around the [004] lattice reflection was used to determine the period length in each structure [10] with an accuracy of 0.5%. The wafers were processed into 250- m-wide 3-mm-long single plasmon waveguides; the method is described in [11]. For the measurement, devices were mounted in a helium flow cryostat. Electrical measurements were made in three terminal configurations, making use of the two bottom contacts to remove the voltage contribution of the highly doped n channel. Spectra from the devices were taken using a Bruker IFS66v/S FTIR spectrometer with 7.5-GHz resolution.
1041-1135/$25.00 © 2008 IEEE
304
IEEE PHOTONICS TECHNOLOGY LETTERS, VOL. 20, NO. 4, FEBRUARY 15, 2008
III. SIMULATIONS In the following simulations, the QCL alignment condition was taken as the field at which the anti-crossing between the upper and injector states occurs as this resonance characterizes a maxima in conduction. As noted in [11], our lasers exhibit negligible frequency shifts with field suggesting that below full alignment there are two field domains present.1. It is the aligned domain that we are identifying by using this simulation method. Each structure is generated from the original design by all layers being multiplied by a constant to thicken or thin the structure. The modified structure is modelled self-consistently at the alignment field to determine the emission frequency of the laser. IV. 2.9-THz DESIGNS For this study, we present eight wafers exhibiting period variation of a 2.9-THz design [13]. The publayers, lished design has Al Ga As–GaAs starting from the injection barrier, in nanometers 3.8/14.0/0.6/9.0/0.6/15.8/1.5/12.8/1.8/12.2/2.0/12.0/2.0/11.4/ 2.7/11.3/3.5/11.6. AlGaAs layers are shown in bold and the 12.0- and 11.4-nm-thick quantum wells are doped with Si at a cm , this design has total period thickness level of of 128.6 nm. The samples presented here have period lengths between 124 and 135 nm with individual layer thicknesses in the same ratios as above, the period length of each sample was determined by high resolution X-ray diffraction. In Fig. 1, we show results from the eight wafers and simulations; all samples were taken from the center of the wafer. The simulation of the structures show a clear, near-linear trend suggesting that thicker structures emit at lower frequencies. There is very good agreement between data from wafers and simulations, the frequency of the wafers span a range of 0.35 THz. Included with the eight wafers in Fig. 1 are three samples taken from a single wafer labelled V193; here we make use of the slight growth variations over a 3-in wafer. The three samples were taken from the center (c1), midway between center and edge (c2), and from the edge (c3) of the wafer. Sample c1 nm, c2 was nm, had a mean period of nm, showing that at the very edge of and c3 was the wafer the structure becomes slightly thinner. The data from these samples fits well with that from the other wafers. Predictions of electrical properties were also obtained from simulations. The simulations suggest that thicker structures should have a smaller splitting between the upper and injector states and align at a lower electric field. This smaller upper-injector splitting in thicker structures can be attributed to thicker barriers which reduce coupling not only between upper and injector states but all states in the miniband, reducing the overall conductivity. A change in alignment voltage should be visible by a shift in the bias where all periods are aligned . and the structure is producing maximum optical power, Experimental results, as shown in Fig. 2 for the samples c1, c2, and c3, agree with this behavior. Pane (c) shows that the thinner structure, c3, has a larger leakage current at low bias and a . The bias is 0.32 V higher voltage at threshold and 1Domain
formation has been observed in GaAs–AlGaAs mid-IR QCLs [12].
Fig. 2. Data from three devices taken from the center (c1), mid-way (c2), and edge (c3) of a wafer. (a) Continuous-wave spectra obtained from the devices near maximum power. (b) Comparison of the current at maximum power and the calculated injector-upper state splitting. (c) Electrical and optical measurements bias occurs for each for the devices; the dashed lines show where the V device.
higher for c3 than c1 compared to a predicted increase 0.20 V. also shifts with thickness as predicted from The current at the change in calculated injector-upper state splitting, shown in (b). The spectral comparison of the three devices, (a), shows the emission frequency of c3 has increased by 0.125 THz from the thicker sample, c1, close to the 0.130 THz predicted by modelling. Threshold bias, nevertheless, is more difficult to estimate from the simulations as it is affected by the alignment bias and the injection efficiency. The similarity of the data for c1 and c2 highlights the uniformity of growth across the 3-in wafer. V. 2.0-THz DESIGNS To investigate this frequency tuning method in another design, two further wafers were grown [14]. V305 was grown to a 2.0-THz recipe given in [11] and V306 was grown with all wells and barriers reduced in thickness by 2.5%. The thicknesses were confirmed by high-resolution X-ray diffraction measurements. Spectra taken from the devices are show as an inset in Fig. 3. Laser spectra were taken at 10 K to allow for temperature stability at high pump power. The emission frequency differs by 0.10 THz between the two structures in good agreement with a predicted shift of 0.15 THz. A comparison of electrical characteristics at 4K are shown in Fig. 3, highlighting similar differences to those shown in Fig. 2. The current–voltage ( – ) for V306 shows more current flowing at low bias and a larger discontinuity in the differential resistance at threshold. Both these differences point to the
FREEMAN et al.: FREQUENCY MANIPULATION OF THz BOUND-TO-CONTINUUM QCLs
305
REFERENCES
Fig. 3. I –V and light current–output (I –L) plots for V305 and V306 at 4.4K are shown. The I –V of V306, the thinner structure, is characterized by a more conductive miniband, and larger dynamic range. The inset shows a spectral comparison of the two wafers near maximum power. The dashed lines highlight dif. ferences in bias at maximum power, V
thinner structure being more conductive, as predicted by simulations. The bias at which the thinner structure reaches has increased by 0.18 V in line with a prediction of 0.25 V. VI. CONCLUSION Changing the period length of a THz bound-to-continuum QCL can be used to controllably alter the emission frequency and the transport properties of a known active region. Data was presented from ten wafers demonstrating how this technique has been used to tune the emission frequency of two separate bound-to-continuum designs operating around 2.0 and 2.9 THz. In the case of our 2.9-THz design, we tuned emission over a range of 0.35 THz.
[1] R. Köhler, A. Tredicucci, F. Beltram, H. E. Beere, E. H. Linfield, A. G. Davies, D. A. Ritchie, R. C. Iotti, and F. Rossi, “Terahertz semiconductor heterostructure laser,” Nature, vol. 417, pp. 156–159, 2002. [2] A. Tredicucci, L. Mahler, T. Losco, J. Xu, C. Mauro, R. Koehler, H. E. Beere, D. A. Ritchie, and E. H. Linfield, “Advances in THz quantum cascade lasers: Fulfilling the application potential,” Proc. SPIE, vol. 5738, pp. 146–158, 2005. [3] C. Walther, G. Scalari, J. Faist, H. Beere, and D. Ritchie, “Low frequency terahertz quantum cascade laser operating from 1.6 to 1.8 THz,” Appl. Phys. Lett., vol. 89, no. 23, p. 231121, 2006. [4] H. W. Hübers, S. Pavlov, A. Semenov, R. Köhler, L. Mahler, A. Tredicucci, H. Beere, D. Ritchie, and E. Linfield, “Terahertz quantum cascade laser as local oscillator in a heterodyne receiver,” Opt. Express, vol. 13, pp. 5890–5896, 2005. [5] P. Siegel, “Terahertz technology,” IEEE Trans. Microw. Theory Tech., vol. 50, no. 3, pp. 910–928, Mar. 2002. [6] L. Mahler, R. Kohler, A. Tredicucci, F. Beltram, H. E. Beere, E. H. Linfield, D. A. Ritchie, and A. G. Davies, “Single-mode operation of terahertz quantum cascade lasers with distributed feedback resonators,” Appl. Phys. Lett., vol. 84, no. 26, pp. 5446–5448, 2004. [7] S. Golka, C. Pflugl, W. Schrenk, and G. Strasser, “Quantum cascade lasers with lateral double-sided distributed feedback grating,” Appl. Phys. Lett., vol. 86, p. 111103, 2005. [8] R. Köhler, C. Gmachl, F. Capasso, A. Tredicucci, D. L. Sivco, and A. Y. Cho, “Single-mode tunable quantum cascade lasers in the spectral range of the CO laser at = 9:5 10:5 m,” IEEE Photon. Technol. Lett., vol. 12, no. 5, pp. 474–476, May 2000. [9] Q. Hu, B. Williams, S. Kumar, H. Callebaut, and J. Reno, “Terahertz quantum cascade lasers based on resonant phonon scattering for depopulation,” Phil. Trans. R. Soc. Land., vol. 362, pp. 233–249, Feb. 2004. [10] K. Bowen and B. K. Tanner, High Resolution X-Ray Diffractometry and Topography. New York: Taylor & Francis, 1998. [11] C. Worrall, J. Alton, M. Houghton, S. Barbieri, H. E. Beere, D. Ritchie, and C. Sirtori, “Continuous wave operation of a superlattice quantum cascade laser emitting at 2 THz,” Opt. Express, vol. 14, pp. 171–181, 2006. [12] S. L. Lu, L. Schrottke, S. W. Teitsworth, R. Hey, and H. T. Grahn, As quantum “Formation of electric-field domains in GaAs/Al Ga cascade laser structures,” Phys. Rev. B, vol. 73, p. 033311, 2006. [13] S. Barbieri, J. Alton, H. E. Beere, J. Fowler, E. H. Linfield, and D. A. Ritchie, “2.9 THz quantum cascade laser operating up to 70 K in continuous wave,” Appl. Phys. Lett., vol. 85, no. 10, pp. 1674–1676, 2004. [14] H. E. Beere, C. H. Worrall, S. Whelan, D. A. Ritchie, J. Alton, S. Barbieri, and C. Sirtori, “Tuning the emission frequency of a 2 THz quantum cascade laser by altering the tatal thickness of the structure,” J. Cryst. Growth, vol. 301–302, pp. 935–940, 2007.
0
