Midinfrared molecular gas lasers optically pumped by a continuously tunable infrared optical .... mTorr of N2O pumped by the OPO is (1014 molecules/cm3).
Midinfrared molecular gas lasers optically pumped by a continuously tunable infrared optical parametric oscillator H. Charles Tapalian, Chris A. Michaels, and George W. Flynn Department of Chemistry and Columbia Radiation Laboratory, Columbia University, New York, New York 10027
~Received 3 September 1996; accepted for publication 24 February 1997! We report the observation of midinfrared super radiant pulses emitted from molecular gases pumped by a narrow-band, continuously tunable infrared optical parametric oscillator. Molecules pumped either into a vibrational combination band or a vibrational overtone state decay by means of super radiant emission to a lower vibrationally excited state. Emission was observed in N2O at 4.5 and 8 mm, and in C2H2 at 13.6 mm and 15.7 mm. Pulse energies as high as 0.5 mJ and pulse widths as short as 7 ns make this an excellent technique for the production of intense midinfrared light. © 1997 American Institute of Physics. @S0003-6951~97!00317-3# The advent of powerful narrow-band continuously tunable infrared ~IR! lasers has renewed interest in the study and applications of mid-IR optically pumped molecular gas lasers ~OPML!. We have observed mid-IR super radiant emission ~classic Dicke super radiance!1,2 from samples of N2O and C2H2 pumped by a tunable IR optical parametric oscillator ~OPO!. The emission is a cooperative effect and occurs between excited rotational-vibrational states following the OPO pumping of either combination band or overtone vibrational transitions ~see Fig. 1!. Line-tunable emission wavelengths ranging from 4.5 to 16 mm have been observed. Much work was performed in the 1970’s and early 1980’s on the development of OPMLs.3,4 Some of these optically pumped lasers are operated by V-V energy transfer5 while others exhibited super radiant ~mirrorless! emission.6 A significant motivation for this early work was the possibility of using OPMLs to perform laser radioactive isotope separation.7 Most of these projects, however, used linetunable lasers, primarily CO2 lasers, as pump sources. This limited the number of molecular states accessible for OPMLs due to the required resonance between the pump and the absorption lines. Continuously tunable pump sources, on the other hand, provide a perfect match between the pump and transition frequencies. Low sample pressures ~,100 mTorr! can therefore be used since pressure broadening of the gas is not required for resonance with the pump beam. Although high pressure, continuously tunable CO2 lasers were formerly used as pump sources, their tuning range was very limited ~;10 cm21 on each branch of the emission band!.8,9 By utilizing an OPO continuously tunable from 1.5 to 4.0 mm, many potential OPML molecules are accessible with a single pump source. Recently, mid-IR lasing was observed from HBr following overtone pumping by an IR OPO.10 A Continuum Mirage 3000 narrow-band mid-IR pulsed ~KTP! OPO is used as the pump source in our experiments. With a bandwidth of approximately 500 MHz, the OPO is capable of exciting a single rotational state within a vibrational transition band of an arbitrary molecule. Output energies range from 1 to 10 mJ depending on where in the 1.5– 4.0 mm tuning range the laser is set. Sealed gas cells ranging from 10 cm to 3 m in length and filled with 10 mTorr to 50 Torr of gas were used without mirrors as OPML media. Emission occurs in both directions along the pump beam Appl. Phys. Lett. 70 (17), 28 April 1997
axis. Monitoring of short wavelength ~up to approximately 5 mm! IR pulses was accomplished using room temperature unamplified InAs photodetectors while emission at longer wavelengths was monitored by liquid nitrogen cooled, amplified HgCdTe photodetectors. The InAs detector response time was fast enough to observe the temporal profile of the pump and emission pulses. A variety of longpass and bandpass optical filters was used to separate the pump laser beam from the emitted beam. Amplified pyroelectric joulemeters provided an absolute measure of the output energy. A highresolution 3/4 m monochromator was used to separate and identify the wavelengths of the individual emission lines. The primary requirement for the occurrence of this type of emission is that a population inversion be induced in the system. In order for this to be accomplished, absorption of the pump light must be large enough that a significant fraction of the ground state population is transferred to the excited state. In addition, the lower vibrational level of the emitting molecular transition must be high enough above the ground vibrational state that only a small fraction of the total
FIG. 1. Representative OPML energy level diagrams using N2O as an example. ~a! Pumping the combination band 100 1 yields emission on the 100 1→100 0 transition and ~b! pumping the overtone 000 2 yields emission on the 000 2→000 1 transition. Both transitions emit in the 4.5–4.6 mm region. At room temperature, both the 100 0 and 000 1states are effectively empty before pumping of 100 1 or 000 2 by an optical parametric oscillator ~OPO!.
0003-6951/97/70(17)/2215/3/$10.00
© 1997 American Institute of Physics
2215
equilibrium population is present in this level. N2O was selected as a suitable OPML candidate due to its relatively strong absorption bands within the OPO tuning range and due to the large dipole transition moments of the symmetric and asymmetric stretch vibrational modes which provide the appropriate gain conditions for super radiant emission to occur. Several rovibrational pump transitions were explored, each providing emission on one of the stretch transitions. Pumping the 00 0 0→10 0 1 transition at 2.9 mm and 000 0→000 2 transition at 2.2 mm produced emission on the asymmetric stretch transitions 100 1→100 0 and 000 2→000 1, respectively. Since the absorption coefficient for 000 0→100 1 is approximately 35 times larger than that for 000 0→000 2,11 the emission intensity is much greater when pumping the former transition. Even though a single j state is excited in 100 1 or 000 2 by the OPO, super radiant emission is allowed and occurs on two transitions, D j 561, corresponding to the P and R branches of the 100 1→100 0 or 000 2→000 1 transitions. P branch emission occurs at 4.6 mm and R branch emission at 4.5 mm. Transmitted pump energy measurements performed using 200 mTorr N2O in a 3-m long gas cell indicate that approximately 24% of a 1.7 mJ, 2.88 mm input pump beam tuned to the 000 0→100 1, P~13! transition was absorbed. An average emission energy in the direction of the pump pulse of 0.11 mJ ~including both the P and R branch emissions! was measured for this same N2O pressure. The resulting photon conversion efficiency is N emit E emit /h n emit 5 50.4360.1. N abs E abs /h n abs This high efficiency is not surprising since the gain coefficient for the emission transition 100 1→100 0 is quite large. From the number of absorbed pump laser photons ~24% of 1.7 mJ at an N2O pressure of 200 mTorr5631015 photons! and the irradiated volume ~60 cm3!, the density of excited molecules in 100 1, J512 is approximately 1014 molecules/cm3. This corresponds to 41% of the molecules in 000 0, J513 being excited to 100 1, J512 @the total number of pump photons contained within the Doppler width ~200 MHz! of the absorbing transition is approximately ~200/ 500!~2.531016)51016 photons for a 1.7 mJ pump energy and a 500 MHz laser bandwidth#. The small signal gain for the 100 1, J512→100 0, J513 transition can be calculated from the density of molecules pumped to 100 1, J512 by the OPO ~ 1014 molecules/cm3 for a 200 mTorr sample! and the absorption cross section for the 000 0, J513→000 1, J512 transition. On a per molecule basis, the absorption for this transition is identical to that for 100 0,J513→100 1, J512 in the harmonic oscillator limit, an excellent approximation for N2O at these vibrational energies. From the known linestrengths for N2O,11 this absorption cross section is 5.3310215 cm2/molecule. Thus the gain coefficient for a 300-cm-long sample cell with 200 mTorr of N2O pumped by the OPO is ~1014 molecules/cm3! ~300 cm!~5.3310215 cm2/molecule!5159. This gives a small signal gain of e 159'1069 under these conditions. Total output energies and peak powers ~including both P and R branches! exceeding 0.5 mJ and 70 kW, respec2216
Appl. Phys. Lett., Vol. 70, No. 17, 28 April 1997
FIG. 2. Time-resolved N2O super radiant emission pulse from the 100 1→100 0 transition measured using an unamplified InAs detector. The gas cell used for this measurement was 14 cm long and filled with 10 Torr N2O. The time scale is relative to an electronic trigger; however the delay between the optical parametric oscillator ~OPO! pump pulse and the N2O emission pulse was approximately 10 ns. From the figure, the pulse width is approximately 7 ns.
tively, were observed at moderate N2O pressures ~500–800 mTorr! in a 3 m cell. A 14-cm-long cell filled with 10 Torr N2O provides similar output. The pump threshold for the emission of pulses in a 3 m cell at moderate N2O pressures was measured to be as low as 50 mJ. Emission pulse widths as narrow as 7 ns were measured using an InAs detector with a 0.25-mm-square active area. A representative pulse is shown in Fig. 2. The emission pulses were delayed relative to the OPO pump pulses by 10–100 ns; the larger the pulse intensity, the shorter the pulse delay. These pulse delays are characteristics of super radiant emission.12 Quenching of the 100 1→100 0 emission process was observed to occur between 50 and 150 Torr N2 O, depending on the pump conditions. Collisions between molecules resulting in a combination of vibrational and rotational relaxation processes, and thus a loss of gain, are responsible for the quenching. Conditions for emission, therefore, require that the delay time between the pump and emission pulses be less than several times the relaxation period for the 100 1 excited molecules. Emission at longer wavelengths in N2O was accomplished by pumping the third overtone of the symmetric stretch vibrational transition, 000 0→300 0. The pump wavelength for this transition is 2.6 mm and the absorption coefficient is comparable to that for the 000 0→000 2 transition. Cascade emission at approximately 8 mm was observed on the 30 0 0→20 0 0 and 200 0→100 0 transitions, resulting in the populating of several j states in the 100 0 vibrationally excited state. The 8 mm wavelength of this emission, however, is out of the range of the InAs detector, so an accurate temporal profile of the pulses could not be determined. Even longer emission wavelengths were attained using C2H2 as the OPML gas medium. When pumping the ~00000 0 0 ) 0 →(01011 1 1 ) 0 transition at 3.0 mm, emission occurs on both the ~01011 1 1 ) 0 →(01011 0 0 ) 1 and the (01011 1 1 ) 0 →(01000 1 1 ) 1 transitions at 13.6 and 15.7 mm, Tapalian, Michaels, and Flynn
respectively. Q, P, and R branch emission lines were observed using a type C HgCdTe photodetector. ZnSe windows were used on the cell in order to pass the long wavelength light. Total unidirectional output energies of up to 20 mJ were measured for 1 mJ of pump energy and 100 mTorr C2H2. In summary, we have demonstrated that the simple addition of a sealed gas cell placed at the output of a narrowband near-IR OPO is sufficient to produce mid-IR super radiant pulses. The range of accessible wavelengths of these OPOs is, thus, effectively extended further into the IR on a line-tunable basis. Potential OPML emission media and their associated emission wavelengths ~in mm! include CO~4.7!, CO2~4.3!, HCN~3.0 and 14.0!, H2O~6.3!, NO2~6.2!, O3~9.6!, OCS~4.8!, SO2~7.3!, NH3~6.1!, CF4~7.8!, CH4~3.3 and 7.7!, and N2O3~5.5!. The simplicity and economy of this arrangement provides the potential for a number of experiments which can be performed using intense, tunable IR pump pulses. The authors wish to thank Professors Richard M. Osgood and Irving P. Herman for their insightful discussions with regard to optically pumped gas lasers and superradiance. This work was performed at Columbia University and supported by the Department of Energy ~DE-FG02-88ER13937!. Equipment support was provided by the National
Appl. Phys. Lett., Vol. 70, No. 17, 28 April 1997
Science Foundation ~CHE-94-19465! and the Joint Services Electronics Program ~U.S. Army, U.S. Navy, and U.S. Air Force; DAAH04-94-40057!. R. H. Dicke, Phys. Rev. 93, 99 ~1954!. M. S. Feld and J. C. MacGillivray, in Topics in Current Physics: Coherent Nonlinear Optics, Recent Advances, edited by M. S. Feld and V. S. Letokhov ~Springer, New York, 1980!, pp. 7–57, and references therein. 3 C. R. Jones, Laser Focus 14, 68 ~1978! and references therein. 4 P. K. Gupta and S. C. Mehendale, Hyperfine Interact. 37, 243 ~1987! and references therein. 5 See, for example, H. Kildal and T. F. Deutsch, Appl. Phys. Lett. 27, 500 ~1975!. 6 See, for example, H. R. Schlossberg and H. R. Fetterman, Appl. Phys. Lett. 26, 316 ~1975!. 7 J. A. Horsley, P. Rabinowitz, A. Stein, D. M. Cox, R. O. Brickman, and A. Kaldor, IEEE J. Quantum Electron. 16, 413 ~1980!, and references therein. 8 B. K. Deka, P. E. Dyer, and R. J. Winfield, Opt. Lett. 5, 194 ~1980!. 9 B. K. Deka, P. E. Dyer, and R. J. Winfield, Opt. Commun. 33, 206 ~1980!. 10 H. C. Miller, J. Dan, T. Radzykewycz, and G. Hager, IEEE J. Quantum Electron. 30, 2395 ~1994!. 11 L. S. Rothman, R. R. Gamache, A. Barbe, A. Goldman, J. R. Gillis, L. R. Brown, R. A. Toth, J. M. Flaud, and C. Camy-Peyret, Appl. Opt. 22, 2247 ~1983!. 12 For general references regarding super radiant pulses, see L. Allen and J. H. Eberly, Optical Resonance and Two-level Atoms ~Wiley, New York, 1975!; A. V. Andreev, V. I. Emel’yanov, and Yu A. Il’inskii, Cooperative Effects in Optics: Superradiance and Phase Transitions ~IOP, London, 1993!. 1 2
Tapalian, Michaels, and Flynn
2217
