Volume 1l, number 4
THE E F F E C T S OF THE I N S E R T I O N OF A CW, L O W - P R E S S U R E CO 2 L A S E R I N T O A TEA CO 2 L A S E R C A V I T Y A. G1RARI3* Gen-7k)c lnc., 2625 Dalton. Quebec, Canad#, GlP 3S9
The single mode performances of a CW TEA (double discharge) CO2 hybrid laser are reported. Single longitudinal and single transverse mode emission have been obtained in the short laser pulse regime (200 mJ, 150 ns fwhm) and in the long laser pulse regime (70 mJ, 2.5/asec fwhm) on at least 30 lines of both the 10.6 gm and 9.6 ~m bands in a tar field, diffraction limited, half angle of divergence of 1.11 mrad.
1. Introduction The insertion of a continuous wave, low-pressure CO 2 laser into a TEA CO 2 laser cavity can be used to alter the temporal evolution of a normal TEA laser pulse and to considerably change its characteristics. One of the major results of this combination is the complete removal of any spontaneous mode-locking that normally occurs in a TEA CO 2 laser and this leads to single longitudinal mode emission either in short or in long pulse operation. This represents a seriotts advantage to the use of a TEA CO 2 laser for applications where an accurate value of the peak power must be obtained such as in the study of non-linear optics. Moreover, for applications where coherent detection is to be used, the narrow transmission bandwidth due to the longitudinal mode selection essentially turns the laser into a very suitable source compatible with coherent detection. Consequently, such TEA lasers can be useH for Doppler measurements: hence. the importance of elimination of mode-locking (large transmission bandwidth, random nature). Moreover, the use of very long pulses without mode-locking should improve the accuracy in Doppler measurements. This paper reports the results of a preliminary study of the pulse shape and energy of a hybrid CW CO 2 TEA CO 2 laser as a function of the operating conditions of both the CW and the TEA laser.
* Also a Ph.D. student with INRS-Energie, Universitd du Qudbec, C.P. 1020, Varennes, Qudbec, Canada. 346
2. Basic principle of laser intensity buildup in the hybrid laser Studies of tile dynamics of the laser intensity buildup in a TEA CO 2 laser have been made by a number of researchers such as Gilbert [ 1]. The formation of a TEA CO 2 laser pulse is illustrated by the solid curve in fig. I. As the laser medium is excited by the current pulse, the appearance of gain in the optical resonator means that the intensity I of the electromagnetic field inside the cavity begins to grow as indicated in the upper part of the fig. 1 (solid curve). Starting from the level of spontaneous emission, the field builds up exponentially with time. As the field intensity becomes large enough, it gives rise to so many, stimulated emission induced transitions from the upper laser level to the lower laser level that the population density of these two levels tends to equalize and the gain is reduced. When the net round trip gain falls to one, the peak value of the laser field is reached. As the net round trip gain becomes smaller than one, the field intensity begins to decrease more or less rapidly depending on the resonator losses. In fact, when N 2 is present in the gas mixture, collisions between unexcited CO 2 molecules and vibrationally excited N 2 molecules (u = 1) give a repumping of the upper laser level [ 1,2]. With N 2 concentrations of the same order of magnitude as that of the CO2, the repumping of the CO 2 by the N 2 maintains the laser field for substantial periods of time (more than one microsecond) and a long tail can be observed after the giant laser pulse. This long tail can be considered as a relaxed pulse since it represents a case where the laser action is sustained by con-
Volume 11, number 4
"8o~'~ o ~_~,
Fig. 1. Laser field intensity and net round trip gain buildup as a function of time in the case of a normal TEA CO2 laser without a CW CO 2 laser inside the cavity (solid curves), with a CW laser under the lasing threshold (dashed curves), and with a CW laser above the lasing threshold (dotted curves).
tinuous pumping during the laser pulse, while for giant pulses the pumping action during the laser pulse is generally considered negligible. In particular, the relaxed pulse dynamics is the subject of a different study reported separately . If a low-pressure CW CO 2 laser section is introduced inside the optical resonator, the dynamics o f the TEA CO 2 laser pulse will be affected. If the gain in the CW laser is too small to overcome the resonator losses, the situation will be modified to that illustrated by the dashed curves in fig. 1. The points to be noted are that the buildup in the resonator will begin earlier as the threshold gain G t is reached earlier and the rate of rise of the laser intensity will be higher resulting in a laser pulse which will occur earlier. In the case where the CW laser gain is above the threshold value
there will be a laser field inside the resonator prior to the excitation of the TEA section. This situation is illustrated by the dotted curve in fig. 1. Since the initial signal in the TEA resonator is of the order of watts rather than where spontaneous emission is the only source of signal, the buildup time for the TEA laser field will be greatly reduced. In fact, it should be so greatly reduced that the signal intensity will reach a large value and as a result the stimulated emission transitions will be sufficiently strong to prevent any further increase in gain. In such a case, the giant laser pulse will be eliminated while the resultant single mode pulse, having a relatively slow risetime, will never reach a very high value and will be of long duration. In order to better visualize how mode-locking takes place in a TEA laser and how it can be eliminated by the use of a CW section, it is necessary to consider the gain profile of the laser medium. Due to the operation at atmospheric pressure, the laser transitions are greatly broadened by collisions. The bandwidth of a typical TEA CO 2 laser transition is of the order o f 4 GHz/atm. Wtlen such a medium is placed in a 3 meter-long optical resonator which has resonant frequencies spaced 50 MHz, the laser tends to oscillate on many different frequencies simultaneously corresponding to the different resonance frequencies of the resonator. Fig. 2 illustrates the situation by indicating on the frequency axis all the resonator resonance frequencies and the broad gain profile of the TEA laser. The narrow gain profile of the CW low-pressure CO 2 laser is superin> posed at the center of this gain profile. In absence of this CW laser, it can be seen that there are over 20 resonator frequencies for which the gain coefficient differs by less than 5 %. Thus, in the absence of the CW laser, the TEA laser field will build up approximately equally for all these frequencies, and the presence of a large number of oscillations at different frequencies will give rise to a modulation of the total laser output, the type of modulation being quite random as the different oscillations are initially triggered by noise. With a CW laser giving an additional peak in the gain spectrum of the amplifying medium in the resonator, it can be readily seen from the difference between the solid and the dashed curve of fig. 1 that the only resonator mode whose frequency lies within the narrow bandwidth of the CW oscillator, will give rise to a very rapidly increasing field, This field will not only be so strong as to make the other modes insignificant but 347
Volume 11, number 4
I-~-- 5 0 M H z I t 4GHz
Fig. 2. Gain profile of the TEA COs laser medium as a function of frequency, with some resonator frequencies (solid lines) for which the gain coefficient differs by less than 5 %. At the center of the gain profile is superimposed the narrow gain profile of the CW low-pressure CO2 laser. will actually prevent ttlem from ever becoming important by depopulating the upper laser level of the transition before these other modes can reach a significant intensity. It is, therefore, not necessary to operate the CW laser above the threshold value to eliminate the spontaneous mode-locking and by this way it is possible to generate giant pulses and relaxed pulses either one separately or as a combination with no significant amount of spontaneous mode-locking. Furthermore, by modulating the CW laser current so that between TEA laser pulses its gain is above threshold, providing the necessary information for locking to the reference oscillator in a coherent system, but decreasing its current before the TEA laser is excited in order not to have a too high background signal, a single mode pulse would still be obtained.
3. Experimental apparatus The laser used is a pin-grid double-discharge TEA CO 2 type (Gen-Tec Model DDL-5 or DD-250 for high repetition rate operation at 300 pps). Its active volume is approximately 360 cc (length 75 cm; height of discharge 3 cm; width of discharge approximately 1.6 cm). The spatial gain profile of the TEA CO 2 laser has 348
been measured. The peak small-signal gain is 0.023 cm -1 at 100 J/liter o f electrical excitation, a mixture o f T 0 ~ He and CO2/(CO 2 + N2) = 0.7. As commonly found by other researchers  using approximately the same electrode structure, the gain is flat between the grid and the solid electrode, increasing a little on the side o f the grid due to the preionisation. On the other axis, the gain is uniformly decreasing and attained the half value on 1.6 cm approximately. During the experiments, the excitation energy was maintained constant at 18 J (50 J/liter) using a 0 . 0 2 / I F capacitor charged at 42 kV. The optical resonator consists of a 10 m radius o f curvature total reflector and a flat, partially transparent mirror 2.56 m apart. A different resonator has also been used consisting of a plane, totally reflecting diffraction grating blazed for 10.6/~m or 9.6/~m, and a 10 m radius of curvature, partially transparent mirror 2.56 m apart. Single transverse mode (TEM00) operation has been achieved by inserting an adjustable iris near the plane mirror. For short laser pulse generation the reflectivity of the partial reflector was first set at 65 % while the He concentration was varied from 90 % to 65 %. The CO2/(CO 2 + N2) ratio was set at 0.7. For long laser pulse generation a reflectivity of 85 % was used, the He concentration was set at 90 % and the CO/(CO 2 + N2) = 0.142.
Volume 11, number 4
9o~ 0.5 MW
6s% ~.9 M W
Fig. 3. Short laser pulse shapes in single longitudinal and single transverse (TEMoo) mode for different tte concentrations; 65 % R, CO2/(CO2 + N2) = 0.7. The laser peak power is indicated for each pulse set.
A parametric study has been conducted for the operation of the Gen-Tec water-cooled CW CO 2 laser. The best conditions for operation near but under the last-
ing threshold were found as follows: gas mixture He: CO 2 :N 2 = 70:10:20, pressure = 6 tort, current = 7 mA (stabilized to 0.0t %). In order to separate the TEA 349
Volume 11, n u m b e r 4
cavity from the low-pressure CW section, a three-inch diameter NaC1 Brewster window was used. The output of the TEA laser was measured simultaneously with a Ge photon-drag detector giving the pulse shape on the screen of a 250 MHz bandwidth oscilloscope, and a pyroelectric joulemeter (Gen-Tec Model ED-200) giving the total energy in the laser pulse. The line emission of the laser was recorded by a spectrum analyzer (Optical Engineering, Model 16-A).
Table 1 Variation of the delay between the beginning of the discharge current pulse and the beginning of the hybrid laser output pulse (delay) as a function of the CW laser pump current Ucw). The experimental conditions are as follows for the CW laser: He:CO2:N 2 = 70:10:20 at 6 torr, and for the TEA laser: 65 % R, 70 % tle, CO2/(CO2 + N2) = 0.7.
CW C O 2 Laser Operating
4. Experimental results
4.1. Single mode emission in the short laser pulse regime Single longitudinal and single transverse (TEM00) mode emission and elimination of spontaneous modelocking have been achieved in the short pulse regime with the resonator consisting of the 10 m radius of curvature total reflector and the 65 % R fiat partial mirror. Typical results are shown in fig. 3. The best results were obtained at 70 % He and CO2/(CO 2 + N 2) = 0.7 which gives a single mode pulse of ~ 200 mJ in 150 ns (fwhm) with a peak power of more than one megawatt. It is to be noted that the delay between the beginning of the current pulse and the beginning of the TEA laser output pulse was decreased by increasing the CW laser current. The experimental results are shown in table 1. Single mode output has also been obtained on at least 30 lines of both the 10.6/~m and 9.6/xm bands centered around the P(20) transition. The divergence angle has also been measured in the short pulse regime on single mode and single line operation [P(20) on the 9.6/xm band] using a method described by Siegman . The best fitting curve gives a beam waist value of
wo(Emax/e2) = ~~.75 -+ 0.25 nun, and a far field, diffraction-limited, half angle of divergence of
B 1= ),/~rw0 = 1.11 -+0.10mrad ,
no CW Laser
Under Losing Threshold Over Lasin9 T h r e s h o l d
regime. The resonator consists of the I 0 m radius of curvature total reflector and the 85 % R flat partial mirror. Typical results are shown in fig. 4. We recorded a laser pulse energy of-~ 70 mJ in a pulse width (fwhm) o f ~ 8/Jsec for the giant pulse and ~ 2.5/xsec for the relaxed pulse. By considering the results attained in the previously published study , it should be possible to obtain single mode emission with relaxed pulses up to ahnost 100/xsec fwhm. This feature should improve drastically the accuracy in Doppler measurements. However, we are beginning an experimental study of possible frequency sweeping of the single mode TEA long pulse, due to the temperature dependence of the refractive index of the gas mixture. The experiment is done by measuring the difference in optical frequency between the single mode output from the hybrid CW TFA CO 2 laser and a stabilized CW CO 2 laser, obtained by optical heterodyning in a Hg: Cd:Te detector I6]. The results should, therefore, be compared to those obtained in TEA lasers using an intracavity SF 6 cell [7,8].
2 where ~ = laser wavelength.
4.2. Single mode emission in the long pulse regime Single longitudinal and single transverse (TEM00) mode emission and elimination of spontaneous modelocking has also been achieved in the long laser pulse 350
It has been shown experimentally that combining a low-pressure CW CO 2 laser and a TEA CO 2 laser in the same optical resonator, it is possible to eliminate
Volume 11, number 4
by using different resonator configurations like the unstable one  and to improve the output performance by using a transverse pulsed, low-pressure laser which should have the same gain on a shorter length. Another way to decrease the divergence of the beam should be attained by using a helical TEA CO 2 laser instead of a double discharge one [ 13,14], with less single mode performances due to the use of resistors as the source of discharge [ 15 ]. Furthermore, it has been shown that single mode emission can be obtained in long laser pulses, and this particular feature is believed to have a successful application in tile field of coherent detection with a better accuracy than the usual techniques employed up to now. The author wishes to thank his thesis director, Dr. J.A. Beaulieu, Defence Research Establishment Valcartier, Valcartier, Qudbec, Canada, for his advice and valuable discussions and also G. Vaudreuil and J.C. Drouin for their technical assistance.
Fig. 4. a) Single transverse (TEMoo) long laser pulse shape with spontaneous mode-locking. 85 % R, 90 % He, CO2/(CO2 + Nz) = 0.142. b) Long laser pulse shape in single longitudinal and single transverse (TEM0o) mode. 85 % R, 90 % He, CO2/ (CO2 + N2) = 0.142. the natural tendency of the TEA CO 2 laser to generate spontaneously mode-locking pulses. Furthermore, as it has been shown experimentally, the delay between the beginning of the current pulse and the beginning of the TEA laser output pulse can be adjusted by varying the CW laser current. This is believed to be of major interest in the synchronization of laser pulses from two different TEA CO 2 lasers. This technique is believed to be superior to other techniques such as the use of long bleachable gas cells [9,10] or the use of Fabry Pdrot reflectors [ 11 ] in that it is much less critical on adjustment than the Fabry---P6rot technique and it does not require to operate the TEA laser near threshold on only one transition as in the saturable gas technique. However, we are considering improving the divergence value obtained with the C W - T E A laser
[ 1] J. Gilbert, J.L. Lachambre, F. Rheault and R. Fortin, Can. J. Phys. 50(1972) 2523 2535.  G.C. Vlases and W.M. Moeny, J. Appl. Phys. 43 (1972) 1840 1844.  A. Girard and A.J. Beaulieu, in the Corresp. Section of the IEEE J. Quantum Electron. (June 1974).  M.C. Richardson, A.J. Alcock, K. Leopold and P. Burtyn, 1EEE J. Quantum Electron. QE-9 (1973) 236 243.  A.E. Siegman, An Introduction to Lasers and Masers, (McGraw-Hill Book Co., 1971) pp. 312--314. [61 J.A. Weiss and J.M. Schmur, Appl. Phys. Lett. 22 (1973) 453-454.  A.W. Stiehl, Res. Lab. Electron. M.I.T. Quart. Prog. Rep. 106 (1972) pp. 63 69.  A.W. Stiehl and P.W. Hoff, Appl. Phys. Lett. 22 (1973) 680 682.  T.A. DeTemple and A. Nurmikko, Opt. Commun. 4 (1971) 231-233. 110] A. Nurmikko, T.A. De Temple and S.E. Schwarz, Appl. Phys. Lett. 18 (1971) 130-132.  C.R. Hammond, D.P. Juyal, G.C. Thomas and A. Zembrod, J. Phys. E: Scient. Instr. 7 (1974) 45-48.  A.E. Siegman, paper X.3 pres. VIII INtern. Quantum Electron. Conf. San Francisco, June 10 13 (1974).  R. Fortin, M. Gravel and R. Tremblay, Can. J. Phys. 49 (1971) 1783-1793.  A. Girard, H. P6pin and J.G. Vall6e, Can. J. Phys. 51 (1973) 1705--1708.  A. Gondhalekar, E. Holzhauer and N.R. Heckenberg, Phys. lett. 46A (1973) 229-230. 351