RETURNTP ; HL-Q‘M;JQMMATERIALS: V Sgéee 7“,??nk Crap to i “ \»LVC twig Checkout fr Your regard. FTE53 wi]?m I be charged if 556k”;s P returned after the dat y stamped below. e new. use 1:: 11-; I 1.. 'I'" me'q-L5n-m. I V ,V. l 'fl.“r'. 7—7"? l' J; ILT’ I -’ fl .' ~‘W'Mvzl E H. _h- in, ‘ b .' a '-' .‘ —. «If. 0 If.‘ ' ‘1‘ ‘ ‘I ‘1 ’l I ; ' i a 4 §. ‘ ’ fi V“” I y . ‘V . a v ‘\ 0‘" ‘f.’ I, F a 1.1 I: ' i F}. :1 "- 4'34 4 I! 1 ,a e‘ i L :9 I . ‘ 5 I ~ :1 a; itxw“"’* ‘ \ r. 1 ~77 . 2‘ 4 flax“ :~- = 1. § : .‘ i 53 a} TIME RESOLVED SPECTROSCOPY AND SMALL SIGNAL GAIN IN A FLASH INITIATED, PULSED HF LASER By Paul E. Sojka A DISSERTATION Submitted to Michigan State University in partia] fulfiiiment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Mechanical Engineering 1983 ABSTRACT TIME RESOLVED SPECTROSCOPY AND SMALL SIGNAL GAIN IN A FLASH INITIATED, PULSED HF LASER By Paul E. Sojka An experimental and computer modeling investigation of a pulsed, flash photolysis initiated, H2 + F2 chemical laser was undertaken. Time resolved spectral (TRS) output, time history of small signal gain ($56) and total pulse energy (TPE) were measured. Several experimental trends were noted. ‘ For the TRS results, regular shifts of individual transition initiation, termination and peak intensity times with increasing rotational level are observed. Transition pulse duration increased with rotational level. For the 556 results, regular shifts of positive gain initiation, termination and peak gain times with increasing rotational level were observed. Positive gain duration increased with rotational level. The experimental TRS results were compared with those of other researchers and then with the results of computer simulations. Pulse duration in this work was longer than that reported elsewhere. This was most likely due to weak initiation of the H2 + F2 chain. No reportable rotational lasing was observed. This is in contrast to other work but in agreement with model calculations. In addition to the experimental study, an existing computer model was modified by the substitution of a wavelength dependent threshold gain in place of the previous wavelength independent threshold gain and by the addition of a flash photolysis initiation option. The modified model and a second, simplified, model were used to simulate the TRS and $56 experiments. Two model rate coefficients were varied to investigate the effects of the hot reaction vibrational pumping distribution and of the vibrational deactivation mechanism rate co- efficients on the simplified model TRS and SSF results. The experimental TRS and 556 results were compared to the calcu- lations resulting from the two models. The results of the simplified model, assuming Vibrational-Transla- tional energy transfen.more closely duplicated experiment than did the results of the modified model (assuming Vibrational-Rotational energy transfer). This is in contrast to the currently accepted under- standing of kinetic mechanisms. Conclusions reached in this study were: (l) The time scales of $56 and TRS are not the same, 556 having much longer durations. (2) The trends of initiation, termination and peak gain or intensity times are similar for $56 and TRS. (3) Computer models are capable of accurately predicting the time resolved characteristics of gain and emission. (4) Further work is necessary to determine the form of V-R,T energy transfer. DEDICATION This thesis is dedicated to Janice. Without you, I never would have made it. ii ACKNOWLEDGMENTS I would like to thank all the friends who helped out: Dr. Mary Brake for teaching me about the MSU Computer System, Tom Nees for sub- mitting and retrieving jobs, Drs. warren Jaul and Robert Brown for explaining portions of the computer models used herein and Craig Everlove and Sherry Ison for numerous hours spent plotting results and drawing the figures you see. I would like to thank many of the pe0ple at the Air Force Weapons Lab who were so helpful: Drs. Phil Whitefield and Steve Davis for advice and the loan of equipment, Len Hanko and Jay Benze for technical expertise and many helpful suggestions. Drs. Larry Rapagnani, Dennis Lankford and Bob Shea for assistance understanding the intracacies of the AFWL computer system, and Harvey Nutter and Jim Ryan for their excellent and willing technical support. In addition, I would also like to thank Captain Ralph Hill, Lieutenant Colonel Dave Olson and Major Barry Crane for their patience, support, and ability to clear away red tape. I thank the faculty members who served on my Guidance Committee: Professors Jes Asmussen, Jr., Robert Cukier and James Beck. I also thank Professor Katherine Hunt for filling in when asked, and my thesis advisor, Professor Ronald Kerber, from whom I've learned so much. Finally, I would like to thank my parents, John and Barbara Sojka, for supporting me in so many ways for so many years. iii TABLE OF CONTENTS Page LIST OF TABLES .......................... vii LIST OF FIGURES . . ....... . . . . . . . . . . . . . . . . iix 1. INTRODUCTION .................. . . ..... l l.l Background. . . ............ . . . . . . . l 1.2 Present Hork. . . . . . . . . . . . . . . . . . . . . 6 2. EXPERIMENTAL INVESTIGATION OF AN HF LASER . . . . ...... 13 2.1 Introduction. . . . . . . . . . . . . . . . . . . . . 13 2.2 Experimental Study. ....... . . . . . . . . . . l4 2.2.l Flash Photolysis Laser. . . . . . . . . . . l4 2.2.2 Diagnostics . . . . . . . . . . ..... . 26 2.3 Results of Time Resolved Spectroscopy Studies . . . . 38 2.3.1 Introduction. . . . . . . . . . . . . . . . 38 2.3.2 Time Resolved Spectroscopy Results for the Mixture He:02:F2:H2 = 20.8:l.O:4.6:l.2. 42 2.3.3 Time Resolved Spectroscopy Results for the Mixture He:02:F2:H2 = 22.0:l.O:2.7:l.0. 51 2.3.4 Results of Pure Rotational Lasing Studies . 55 2.3.5 Discussion ................ . 57 2.4 Results of Small Signal Gain Studies ......... 55 2.4.1 Introduction. . . . . . . . . . . . . . . . 55 2.4.2 Results of Small Signal Gain Studies for the Mixture He:02:F2:H2 = 20.8:l.O:4.6:l.2. 70 2.4.3 Results of Small Signal Gain Studies for the Mixture He:OZ:F2:H2 = 22.0:l.0:2.7:l.0. 72 2.4.4 Discussion. . ............... 75 2.5 Results of Total Pulse Energy Studies ........ 80 iv 3. COMPUTER SIMULATION OF AN HF LASER AND COMPARISON WITH EXPERIMENT. . . . . . ............. . ...... 3.1 Introduction ..................... 3.2 Computer Modeling Results of Time Resolved Spectroscopy and Comparison with Experiment ..... 3.2.1 Introduction ................ 3.2.2 Comparison of VT Modeling Results for Time Resolved Spectroscopy with Experiment: Initial Rate Package. . . . . ....... 3.2.3 Comparison of VT Modeling Results for Time Resolved Spectroscopy with Experiment: Modified Vibrational Pumping Distribution . 3.2.4 Comparison of VT Modeling Results for Time Resolved Spectroscopy with Experiment: Modified Vibrational Pumping Distribution and Modified V-T Deactivation ....... 3.2.5. Comparison of VRZOJ Modeling Results for Time Resolved Spectroscopy with Experiment. 3.2.6. Summary of Time Resolved Spectroscopy MOdeling Results. . . . . . . . . . . . . . 3.3 Computer Modeling Results of Small Signal Gain and Comparison with Experiment. . . ..... . . . . 3.3.l Introduction. . . . . 3.3.2 Comparison of VT Model Small Signal Gain Results with Experiment ...... 3.3.3 Comparison of VR20J Model Small Signal Gain Results with Experiment . . . . . . . . . . 3.3.4 Summary of Small Signal Gain Modeling Results . ..... . . . . . . . . . . . 4. SUMMARY AND CONCLUSIONS ....... ‘ ............ APPENDIX A: COMPUTER MODEL FORMULATION ............. APPENDIX B: DETERMINATION OF INITIAL HF CONCENTRATION DUE TO PREREACTION ................. APPENDIX C: MODEL RATE COEFFICIENTS ............... APPENDIX D: APPENDIX E: ERROR ANALYSIS FOR PROBING SMALL SIGNAL GAIN OFF LINE CENTER ................ RAN DATA FOR TIME RESOLVED SPECTROSCOPY PLOTS USED IN FIGURES 2.17 AND 2.21 ......... V Page 81 81 88 88 88 91 93 95 97 104 104 105 110 113 120 126 138 140 144 146 Page APPENDIX F: TRS AND SSG PHOTOGRAPHES OF OSCILLOSCOPE DATA . . . . 149 REFERENCES ............................ 249 vi Table 2.1 LIST OF TABLES Page Summary of TRS Results from References [14, 17, 19, 42, 43, 52] ................. 63 Current Rate Coefficients in H2 + F2 Systems ....... 141 VT Model Rate Coefficients for the H2 + F2 Chemical Laser ...................... 143 Data for TRS presented in Figure 2.17 .......... 147 Data for TRS presented in Figure 2.21 .......... 148 SSG photographes, He:02:F2:H2=20.8:1.O:4.6:1.2, 102 torr. 149 TRS photographes, He°02:F2:H2=20.8:1.O:4.6:l.2, 36 torr . 152 TRS photographes, He'OZ:F2:H2=20.8:1.O:4.6:1.2, 102 torr. 158 TRS photographes, He202:F2:H2=20.8:1.O:4.6:1.2, 331 torr. 165 SSG photographes, He:02:F2:H2=22.O.1.O:2.7:l.0, 102 torr. 172 SSG photographes, He°02:F2:H2=22.O:1.0:2.7:1.0, 331 torr. 179 TRS photographes, He:02:F2:H2=22.0:1.0:2.7:1.0, 36 torr . 187 TRS photographes, He:02:F2:H2=22.O:1.O:2.7:1.O, 102 torr. 199 TRS photographes, He:02:F2:H2=22.0:1.O:2.7:1.0, 331 torr, R0=O.81 .................... 217 TRS photographes, He:O :F :H =22.0:1.0:2.7:l.O, 331 torr, R0=O.97 .................... 233 vii NNNNN NNNNN Nanci-boo .10 .11 .12 .13 .14 .15 LIST OF FIGURES Page Laser ceH schematic ................... 15 Laser cell window configurations (a) laser cell window configuration for the mixture He:02:F2:H2= 20.8:l.O:4.6:1.2 (b) laser cell window configuration for the mixture He:02:F2:H2= 22.0:1.0:2.7:1.0 ................... 16 Pulse discharge circuit schematic ............ 19 Trigger circuit block diagram .............. 20 Laser cell gas handling schematic ............ 21 Hydrogen radial “sting" injection configuration ..... 23 System photographs (a) mixing tube _ (b) laser cell ...................... 24 Rotameter panel photograph ................ 25 Time resolved spectroscopy experimental configuration . . 27 Small signal gain experimental configuration ....... 28 Total pulse energy experimental configuration ...... 29 Data photographs (a) typical TRS intensity trace (b) typical SSG trace .................. 31 Photographs of experimental setup (a) photograph of monochromators 1 (b) photograph of oscilloscopes ............. 32 CH probe laser used in small signal gain tests ...... 34 Flashlamp intensity measurement experimental configuration ...................... 36 viii Figure 2.16 2.17 2.18 2.19 2.20 2.21 2.22 2.23 3.1 Page Typical flashlamp intensity trace ............ 37 Time resolved spectral output: He:02:F2:H2= 20.8:l.O:4.6:1.2 (a) 36 torr pressure (b) 102 torr pressure (c) 331 torr pressure ....... . ........... 41 Spectral time history: He:O :F :H = 20.8:l.O:4.6:1.2, 36 torr pfeséuré ............ 43 Spectral time history: He:0 :F :H = 20.8:l.O:4.6:1.2, 102 torr Eregsuge ........... 44 Spectral time history: He:O :F :H = 20.8:l.O:4.6:1.2, 331 torr BreEsu3e ........... 45 Time resolved spectral output: He:02:F2:H2= 22.0:1.0:2.7:1.0 (a) 36 torr, R (b) 102 torr, 90 (c) 331 torr, R0 (d) 331 torr, R0 Small signal gain time history: He:O :FzzH =20.8:1.0:4.6:l.2 (a) 3 = 1-6 band, 102 torr (b) v = 2-1 band, 102 torr ................ 71 = 0.97 = 0.97 = 0.97 - = 0.81 ................. 53 signal gain time history: He:02:F2:H2=22.0:1.O:2.7:l.0. = 1-0 band, 102 torr (b) v = 2-1 band, 102 torr = 1-0 band, 331 torr = 2-1 band, 331 torr ................ 73 Time resolved spectral output: comparison of VT model results and experiment using rate package VT (standard rates) He:O :F :H = 20.8:l.O:4.6:1.2 (a) experimental results, 36 torr2 2 2 (b) experimental results, 102 torr (a) experimental results, 331 torr (d) model results, 36 torr e) model results, 102 torr (f) model results, 331 torr ............... 89 ix Figure 3.2 3.3 3.4 3.5 3.6 3.7 Page Time resolved spectral output: comparison of VT model results and experiment using rate package VT2, He:0 :F . = 20.8:l.O:4.6:1.2 (a) experimental Tesélté, 36 torr (b) experimental results, 102 torr (c) experimental results, 331 torr (d) model results, 36 torr (e) model results, 102 torr (f) model results, 331 torr ............... 92 Time resolved spectral output: comparison of VRZOJ and VT model results (with rate package VT3) with experiment, He: 02: = 20.8:l.O:4.6:1.2 (a) experimental2 regulés, 102 torr (b) VT model results using rate package VT3, 102 torr (c) VR20J model results, 102 torr ............ 94 Individual derivative contributions for mechanisms reducing population inversion on P( (5), He: 0 H2=20.8:1.0:4.6:1.2 (a) VT model résults fog rgte package vT3,102 torr (b) VR20J model results, 102 torr ............. 103 Small signal gain time histories: comparison of experimental and VT model results using rate package VT (standard rates), He 0 :F 2 = 20.8:l.O:4.6:1.2 Ea) experimental results,v FOFb afidH 102 torr b) model results, v = 1-0v band, 102 torr (c) experimental results, v = 2-1 band, 102 torr (d) model results, v = 2-1 band, 102 torr ........ 105 Small signal gain time histories: comparison of experimental and VTF model results using rate package VT2, He: 0 =20.8:1.0:4.6:1.2 (a) experimental 4e561t§,v = 1 -0 band, 102 torr (6; model results, v =1 -0 band, 102 torr (c experimental results, v = 2- 1 band 102 torr d) model results, v =2-1 band, 102 torr ........ 107 Small signal gain time histories: comparison of experimental and VTF model results using rate package VT3, He: 0 = 20.8:l.O:4.6:1.2 (a) experimental 4e561t§,v = 1-0 Oband 102 torr (b) model results, v = 1 -0 band 102 torr (c) experimental results, v = 2-1 band, 102 torr (d) model results, v = 2-1 band, 102 torr ........ 108 Figure . Page 3.8 Small signalgain time histories: comparison of experimental and VRZOJ results (a) experimental results, v = 1-0 band, 102 torr (b) model results, v = 1-0 band, 102 torr (c) experimental results, v = 2-1 band, 102 torr (d) model results, v = 2-1 band, 102 torr ........ 111 A.1 Total flashlamp intensity vs time ............ 134 A.2 Reflectivity vs wavenumber for the output coupler with nominal 97% reflectivity ....................... 136 A.3 Reflectivity vs wavenumber for the output coupler with nominal 81% reflectivity ....................... 137 xi a. . v I~:‘. e" I’:r- 5 ‘l \.O‘ a.b CHAPTER 1 INTRODUCTION . 1.1 Background A chemical laser is defined as a laser whose population inversion is produced by energy liberated during a chemical reaction [1]. The first chemical laser emission was observed by Kasper and Pimentel [2] in 1965 in an exploding mixture of hydrogen and chlorine. The first hydrogen-fluoride (HF) chemical laser emission was observed by Kompa and Pimentel [3] in 1967. Both lasers were initiated by flash photo- lysis of’molecular fluorine. The chemical lasers of most interest are those operating on diatom-radical exchange reactions of the form: * A + BC = AB + C AH is unnecessary to exactly uniformly initiate the entire lasing volume: If the medium is initiated in a small portion of its volume the reaction will propagate throughout the whole medium volume due to collisions and diffusion and due to thermal effects [15]. Two disad- vantages of non-uniform initiation are poor beam quality due to index of refraction gradients within the mixture and a lack of repeatability in the laser output. A lesser advantage of the HF system is the large stimulated emission cross section. Research on HF chemical lasers also contributes to the understand- ing of two other chemical laser systems of interest: DF and DF/COZ. The DF laser emits in an atmospheric transmission window providing good beam propagation at lower altitudes and has slower rates of deactivation than HF. The chemical kinetics of the DF molecule are greatly complicated by its high density of states. This high density of states increases the number of relevant reaction pathways and makes it very difficult to measure the relevant reaction rates. However, one can apply results gained from studying HF lasers to the DF and DF/CO2 laser systems by making use of the isotOpe effect between HF and DF and by making use of surprisal analysis techniques for analagous reactions in the HF and DF systems [16]. In this work, initiation of the HF laser pulse was by photolysis of molecular fluorine using flashlamps. The initiation reaction pro- ceeded as: F + hvp = 2F (1-4) 2 with up being an ultraviolet photodissociating photon of frequency up. After generating an initial fluorine atom concentration, the chain re- action proceeds as stated above. It is known that highly diluted H2 + F2 systems initiated flash photolytically permit detailed analyses of the kinetic mechanisms [17]. Electric discharge initiation, on the other hand, produces unwanted charged species, complicating the chemistry [4, 18]. For the conditions of this study, the medium absorbing the ultra- violet flash photolysis photons can be characterized as optically thin. This is equivalent to stating that the photolysis signal passes through the medium with its intensity nearly constant: absorption by F2 is minimal. In that case, flash photolysis has the advantage of genera- ting a homogeneous fluorine atom concentration as a function of time over a wide range of fluorine partial pressures and mixture total pres- sures, even when mixed with non-absorbing gases [19]. There are no charged species involved. Consequently, lasers initiated by flash photolysis are easier to model numerically than those using another form of initiation and are thus well suited to studies gaining insight into competing kinetic mechanisms. One potential complication is the possibility of "hot" F atom formation during the photolysis process. "Hot" atoms are those with a thermal energy significantly greater than 3/2 ka. If such atoms are formed, it is felt they could significantly alter the vibrational state product distribution of Equations (1.2) and (1.3)[20]. Furthermore. since the ultraviolet light required for ini- tiation is difficult to produce, and since it is difficult to couple more than a small fraction of the ultraviolet light produced by the flashlamps into the medium, the initiation efficiency is poor and the ’0 a 0. .al- 'I- :-0. ‘- A a\-F '5 I‘M“ ‘8 ‘3 .. , .f‘~ 5r- rate of fluorine atom production is less than that of other initiation schemes. A further disadvantage of photolytic initiation arises if the medium is not optically thin. This occurs if significant absorption of the photolysis signal occurs as it traverses the medium. This would be caused by an overlarge absorption path length or an increase in the pressure of the absorbing species (F2). Either of these would lead to inhomogeneous absorption and initiation of the medium, and hence, poor beam quality and lack of repeatability. Additional disadvantages of HF lasers are the very efficient de- activation processes of HF itself and other collision partners [21] and the poor beam transmission through the atmosphere [22]. In addition, there are very strong analogies between the mechanisms in HF and DF lasers. Since lasing on DF occurs in an atmospheric transmission window and is of great interest, research on HF lasers continues. 1.2 Present Work Evaluation of the performance of chemical lasers requires an understanding of the chemical, radiative and relaxation kinetic mechan- isms that pump and deactivate the energy levels associated with lasing. The detailed representation of the mechanisms necessary for accurate prediction of laser performance requires a mix of experimental and computer modeling investigations. The main limitation in the comparison of computer models with ex- periment is the lack of a well defined and characterized experiment. n;- - .V '- co. J... 1". "p- 7 Such an experiment should have sufficient diagnostics to monitor as many time histories of HF(v,J) populations as is feasible. Previous experimental endeavors consist of time resolved spectra recorded at only one pressure, using one mixture composition and employing only one value of outcoupling. This is insufficient. A more comprehensive study is necessary. This work presents the results of a comparison of such a set of detailed experimental observations with computer simulations to faci1-. itate the understanding of the kinetic mechanisms involved in the H2 + F2 laser. To that end, experiments have been performed to fully characterize a pulsed H2 + F2 flash photolysis laser at selected, well controlled operating conditions. A systematic experimental study was conducted where: (l) the time-resolved small signal gain on eleven lines, (2) the total P-branch laser emission, and (3) the total pulse energy were measured for an H2 + F2 laser. Specific objectives of this study were: (1) To measure small signal gain and time resolved spectra at three cavity pressures (36, 102, and 331 torr). (2) To measure small signal gain and time resolved spectra for two mixture compositions (He:02:F2:H2 = 20.8:110:4.6:l.2 and 22.0 1.0 2.7 1.0). (3) To measure time resolved spectra using two different output couplers (nominal 81% and 97% reflectivity). A minimum of three detailed data sets were to be developed from the results of objectives 1-3. #1 ii 'Al- .1! 0.. '\I .'l ‘- 1" 'I1‘ .*‘-“ I ‘u 1 I l 1 I 1 g \ J 'a,’ . '11 1‘ 1 '5‘ ‘I '~ I 8 Finally, all experimental results were to be compared with the re- sults of a computer modeling study. This very comprehensive computer model includes V-V, V-R,T and R-R,T relaxation channels, and P-branch and pure rotational lasing [23]. The model was modified by inserting the latest available rate data [24-27], and adding a wavelength depen- dent threshold gain. Many investigators have already completed studies of P-branch time resolved spectra. Several initiation schemes have been used including laser photolysis [28, 29] ultraviolet spark photolysis [30], electron beams [12], electric discharge and flash photolysis. A multitude of reacting species have been used. In the case of flash photolysis initiation this included: F20 + H2 [31], UF6 + H2 [18, 32], XeF4/SbF5 + HZ/CH4 [33], MOF6 + H2 [18, 34, 35], NF6 + H2 [36], and F2 + H2 [11, 17-19, 35, 37-43]. Individual transition intensity traces are shown in several of these studies: References 18, 31, 33 and 41 show "cold band" lasing while References 14, 17, 42 and 43 show both "hot band" and "cold band" lasing. In the case of electric discharge initiation, reacting species used included: H2 + SF6 [44, 45], CH4 + SF6 [44, 46], C3H8 + SF6 [47-49], HI + SF6 [50], H2 + freons [51] and H2 + F2 [14, 51-54]. Several studies utilizing electric discharge initiation show individual intensity traces for cold band lasing [44-46, 50, 51, 55, 56] while two show individual intensity traces for hot band and cold band lasing [17, 52]. The literature available on pure rotational time resolved spectra is not as extensive as that on P-branch time resolved spectra. Nonetheless, several studies have been completed. Pure rotational lasing has been observed in mixtures of H2 + CF4 [57], H2 + BF3 [58], CH4 + SF6 [46] and H2 + SF6 [49, 59, 60] initiated by electric dis- charge. Pure rotational lasing has also been observed in flash photo- lytically initiated mixtures of CF3I/CF3Br with C2H2/CH3C2H [61] and in flash photoelimination of hydrocarbons [62]. Both of these groups present traces of individual transition intensities. There have been no reports of pure rotational lasing in H2 + F2 laser mixtures to date. There are few reports of small signal gain measurements. Two authors give values for peak gain on a single transition in SF6 + H2 mixtures initiated by electric discharge [63, 64]. Another study presents a gain averaged over all transitions available from a multi- line probe laser [15]. Small signal gain has been measured versus position in CH HF lasers (cf. Reference 65), The only study to date reporting time history of small signal gain is for the D2 + FZ/CO2 system of Reference 66. Reference 67 reports P-branch and pure rotational gains in CO2 lasers. A number of studies are available which compare computer modeling predictions to experimental results. These include References 23, 39-41, 68-70. These works compare only P-branch time resolved spectra. This work compares time resolved spectra and time history of small signal gain at three pressures and two mixtures. This is the first 10 time such a well diagnosed experiment and such a detailed computer model have been compared. The results, in conjunction with computer simulations, will be used to evaluate weaknesses in the four kinetic relaxation modes. These four relaxation modes of interest are Vibrational to Translational (V-T), Vibrational to Rotational (V-R), Vibrational to Vibrational (V-V), and Rotational to Rotational and Translational (R-R,T). Vibrational to Translational relaxation HF(v,J) + M = HF(v-1,J) + M E m 3500 cm"1 (l 4) is assumed to occur with the product rotational levels being in thermal equilibrium at the translational temperature. The resulting energy defects for HF V-T processes are approximately 3500 cm']. Similar to V-T, V-R relaxation HF(v,J) + M = HF(v-l, J+AJ) + M E m 150 cm“ (1 5) also assumes a portion of the vibrational energy is transferred into rotational energy of the product molecule. However, the product mole- cule rotational energy contribution is much higher for V-R than for V-T. In fact, for V-R, if the product rotational state is assumed to minimize the reaction energy defect, high rotational levels will result with very little contribution to translational energy. Thus, V-T and V-R have essentially complementary fractions of rotational and trans- lational energy in their product molecules: V-T product molecules have nearly all the energy transferred from vibration going into translation, while V-R molecules have nearly all the energy transferred from vibra- tion going into rotation. 11 Current kinetic understanding implies that a combination of V-R and V-T energy transfer represents the vibrational deactivation mode in HF lasers. This combination is denoted V—R,T, the vibrational energy transferred to rotation and to translation in the product molecules are of the same order of magnitude. In contrast to V-T, V-R and V4R,T, all of which are mechanisms for removing vibrational quanta, Vibrational to Vibrational energy transfer is concerned with rearranging the distribution of vibrational energy while conserving the number of vibrational quanta. HF(v],J]) + HF(v2,J2) = HF(vl-Av,J]) + HF(v2+Av,J2) E m 450 cm" (1 6) It is assumed that there is no change in rotational state for V-R exchange. The last relaxation mode of importance is R-R,T. This mechanism is assumed to consist of single quantum rotational exchange (R-R) HF(v],J1) + HF(v2,J2) = HF(v,J1-l) + HF(v2,J2+l) (1 7) and single and double quantum R-T. HF(v,J) + M = HF(v,J-AJ) + M J = 1,2 (l 8) As in the case of vibrational energy transfer, R-T removes quanta and R-R conserves quanta while rearranging the distribution of rotational energy. Current practice is to assume that R-T contributes approxi- mately two thirds of the total R-R,T rate. The other one third is comprised of R-R. In summary, the goal of this work was to examine in a systematic manner the dominate kinetic mechanisms in a pulsed H2 + F2 laser: i.e. to develop a collection of consistent time resolved data to use in 12 evaluating the weaknesses in kinetic models which are in turn used to guide laser system development. CHAPTER 2 EXPERIMENTAL INVESTIGATION OF AN HF LASER 2.1 Introduction Since the advent of the HF chemical laser in 1967, there has been much research devoted to answering fundamental questions regarding laser performance. However, after sixteen years of effort, there are still areas to be explored. This work attempts to address several areas of concern. One such area is the impact of lasing on pure rotational transi- tions. This is important in large scale devices as even a small fraction of the total output power being emitted at pure rotational transition wavelengths could severly damage optical components [71]. It is thus necessary to know the extent of lasing on pure rotational transitions. Another area of concern is in the examination of the time history of small signal gain (SSG) on P-branch transitions. This is pertinent to applications involving laser amplifiers, the gain being the single most important parameter in studies of laser amplifiers. There have been reports of wavelength averaged [15] and peak [63, 64] gains measured for HF lasers and measurements of gain time histories for CO2 lasers [66]. This is the first study to combine measurements of TRS and SSG and a computer simulation, all on a single device. 13 14 In the remainder Of this chapter, an experimental basis for answering these questions will be presented. In the next chapter, the results of computer simulations Of the experiment will be introduced and compared with the experimental results of this chapter. Further conclusions will be drawn from the comparison of the experimental and simulation results and explanations will be proposed for the areas of concern listed above. 2.2 Experimental Study 2.2.l Flash Photolysis Laser A schematic of the flash photolysis laser cell used in this study is shown in Figure 2.1. Referring to this figure, the laser cell con- sisted of a 100 cm long aluminum cavity Of 10 cm x 10 cm cross section. Two 1.6 cm diameter inlet ports were located in one cavity sidewall 10 cm from one cavity end and two 1.6 cm diameter outlet ports were sym- metrically located in the Opposite cavity sidewall 10 cm from the Opposite cavity end. The gas mixture was fed in through the inlets, flowed 80 cm longitudinally through the cavity and exhausted through the outlets. A 10 cm diameter dump port was Opened after each run to ra- pidly exhaust the cavity of combusted gases. The ultraviolet light necessary for initiation Of the gas mixture was coupled into the cavity through 11 cm x 11 cm x 2 cm synthetic quartz windows (Suprasil II). Two different window configurations were used (see Figure 2.2). In one configuration, five quartz windows were mounted on the top side Of the cavity and five quartz windows were mounted on the bottom side of the cavity (Figure 2.2a). In the other configuration, the five quartz windows mounted on the bottom side of 15 1912 VIEW W \wcm DUMP PORT \mtm OUT LETS Figure 2.1 Laser cell schematic 16 53(11 1 11 1 210m SUPRASIL n WINDOWS ( ) BOTH SIDES a [H 112» “T- :1 .. .. 3.7-3. _ _ i _ .. 42$: .x'u D ‘ ". ‘\ II :09; o \ 1‘ c. ‘é’w- 1 ‘3'-1‘-T“T--T*-T"“ 1_l 5.08 cm DIA. COP; WINDOW BOTH ENDS (b) uliuiuzlcm NO PURSE SUPRASIL I! WINDOWS H. "MOE 1q_1 ifiql hip-Janu-J-u-J 'J--1--J-qbd ‘.fi 0 I). \' c. 1- n4 0 \ l ". b L C. - q ------ § J ------ :- db 1 1 o '_L p ,_N_ Z. .JLI 11 I 55 12 ALUMINUM BLOCK 7.87"! NA. COP: WINDOW IOTN ENDS Figure 2.2 Laser cell window configurations (a) laser cell window configuration for the mixture He:02:F2:H2= 20.8:l.O:4.6:1.2 (b) laser cell window configuration for the mixture He:02:F2:H2= 22.0:1.0:2.7:1.0 17 the cavity were replaced by one 55 cm x 11 cm x 2 cm aluminum plate (Figure 2.2b). In each case, the quartz windows were sealed to the cavity with viton o-rings and clamped in place by aluminum brackets. A Can window at the brewster angle was attached to each end Of the cavity by an aluminum mount. Two different brewster window configura- tions were used (see Figure 2.2). In one case, internal brewster mounts were used to hold the 5.08 cm diameter CaF2 windows in place. The window spacing was 81 cm and there was no window path to purge (see Figure 2.2a). In the second case, external brewster mounts were used to hold the 7.67 cm diameter Can windows in place. The window spacing was 121 cm and the window paths were purged with helium (see Figure 2.2b). All window-to-brewster mount and brewster mount-to-cavity sur- faces were sealed with viton O-rings. The entire cavity-quartz window-brewster window assembly was pas- sivated by exposing it to increasing concentrations of fluorine (in- itially 10%, increased in 10% increments, to 50%, F2 in He, total pressure 400 torr) for periods Of one-half hour. The optical resonator was of a stable configuration and external to the cavity. The maximum reflector was a 15.2 cm diameter, 500 cm radius of curvature copper mirror. Two flat dielectric output couplers were used. Corresponding to Figure 2.2a, an output coupler with maxi- mum 81% reflectivity (see Figure A.3) was used. The mirror spacing was 121 cm for this case. Corresponding to Figure 2.26, an output coupler with maximum reflectivity Of 97% (see Figure A.2) was used. For this case, the mirror spacing was 180 cm. The ultraviolet light necessary for laser initiation was supplied by a high voltage discharge through four flashlamps (ILC: 56 cm long, 18 0.9 cm diameter, filled with 50 torr Xe). The lamps were mounted in pairs directly above and below the cavity, outside the quartz windows. The high voltage pulse was produced by charging four capacitors (Maxwell: 31161, 0.7 microfarads, 45 kV maximum) to between 30 kV and 35 kV and discharging two of them through each pair of flashlamps using spark gaps (Maxwell: 40060, 75 kV maximum). The spark gaps were si- multaneously triggered by a high voltage trigger generator (Maxwell: KV50-805). A schematic of the upper half Of this circuit is shown in Figure 2.3. The lower half of this circuit was identical to the upper half. The signal from the trigger generator to the spark gaps was con- trolled in the following manner. The square wave output from a rotating wheel chopper (Ithaco: 383) and the manual triggering signal from the laser control panel were input toia pulse generator (Hewlett-Packard: 214A). The pulse generator served as an “and gate", firing the high voltage trigger generator. The high voltage trigger generator then fired the spark gaps as mentioned, simultaneously triggering the fast risetime oscilloscope (Tektronix: 400 M12, 7844). A block diagram of the trigger circuit is shown in Figure 2.4. Figure 2.5 is a schematic of the gas handling system. The gases used were: (1) electrolytically pure oxygen (99.98% minimum purity) (2) fluorine (98.2% minimum purity) (3) helium and hydrogen (98% minimum purities). Batch analysis of the fluorine showed the following inpurities: HF <0.4 molar percent Air 1.69 molar percent 19 u_meo;um u_:ue_o mmcmcumpu mm_=a m.~ mezm_d COP-203 hzwccao Ir ><4w¢ $02.22!.— R1 4 up: 5)] ¢2(4=m>\[1 a: g] 55:59, E o: m... 3. am 8 ._T o» o illlleIIIII. 59>... ~95? coo... fill: . u 5%.. 53.5 no 2. u " 20¢..— < Tllléll . _ . I Q‘GXm-(flm Ill—J 1A1 na._<> c2153.. ....... who 553.8: :E u>4<> to 55 ...... ® mi; :85 ..... . r\u._ monEO 02.._.h_>~ommg me_h m.~ oczmwm man J mmmmw 4. .3243 85.00 cmu¢==OSO 02¢ 531—32.: if.“ .- ’b C I. .l ' l Iii. | NV . I ' ‘ V At 3.. E .l. A.“ h F 1- h «a ‘1 \. l .1! F .V 1 ‘21 ans Q.‘ I. l\h .' o 1 on. gnu I u s 30 a sample shown in Figure 2.l2a. Photographs of the monochromators and the oscilloscope are shown in Figure 2.13. Time resolved spectroscopy measurements of pure rotational lasing intensities were also performed. The experimental apparatus was that of the P-branch lasing case with three exceptions: (l) The Can brewster windows were replaced with NaCl brewster windows having much improved transmission in the pure rotational lasing region of the spectrum, (2) The l50 grooves/mm, 4.0 micron blaze monochromator diffraction grating was replaced by a 75 grooves/mm, l6.0 micron blaze monochromator diffraction grating, and (3) The GezAu detector and biasing system was replaced by a liquid nitrogen cooled HngTe detector (SBRC: 40742, 200 nsec risetime) and applicable biasing circuit. See Figure 2.9 for a schematic of this setup. A slightly different system was used during the SSG experiments. The laser cell of Section 2.2.1 was modified by removing the optical resonator. A commercially available continuous wave (CW) HF laser (Helios: CLI) was employed to probe the laser medium. This laser utilized an electric discharge in a mixture of SFG, 02, He and H2 to produce laser emission. The probe laser resonator consisted of a gold coated 200 cm radius of curvature maximum reflecting mirror and a grating output coupler which allowed single line operation. Transitions available for probing from this laser were: P](3), P1(4), P](5), P](6), P](7), P](8), P](9), P](l0), P2(5), P2(6), P2(7). P2(8). This probe laser is discussed further in Reference 73. Pictures of the probe 31 v» .. . .r-- . ,‘ ' 1 ,- . . .. ; .l'" I i - 2 -2 " . , o. a‘ .’.‘L’.;‘ ' g , . g :7. 2‘ ‘ .' .. S ». ' 1’. 1.; ' u I 5 I . t . Ly ‘4 . Figure 2.l2 Data photographs (a) typical TRS intensity trace (b) typical SSG trace Figure 2.l3 Photographs of experimental setup (a) photograph of monochromators (b) photograph of oscilloscopes (b) 33 laser and its control panel are shown in Figures 2.14 and 2.8 respectively. . The small signal gain diagnostics were performed as follows. The probe laser was tuned to a specific transition using the probe laser resonator grating. The output was then chopped by the rotating wheel chopper and steered onto a beam splitter. The resulting two output beams were designated as the reference beam and the signal beam. They will be described separately (see Figure 2.10). The signal beam was that portion of the probe laser beam that was reflected off the beam splitter. It traversed a path through the laser medium collinear with the path of the laser emission described in the TRS section above. The time dependent chemistry occurring within the laser medium perturbed the signal beam. The resulting perturbed signal beam then traveled to the detector following the path described for the TRS beam. The reference beam was that portion of the probe laser beam that was transmitted through the beam splitter. Upon exiting the beam split- ter, the reference beam was focused using a Can lens and steered into the EMI shielded room. Once inside the room, the reference beam was focused again by another CaF2 lens and steered into a monochromator diffraction grating-detector-bias circuit system identical to that in the signal beam's path. A dual beam, EMI shielded oscilloscope (Tektronix: 7844) with two fifty ohm vertical amplifier plugins (Tektronix: 7A19) monitored the two signals simultaneously. Pictures were taken of each data run with a sample shown in Figure 2.12b. Pictures of the monochromatic- diffraction grating-detector-bias circuit are shown in Figure 2.l3a. 34 .14 CW probe laser used in small signal gain tests Figure 2 l.‘ 'vl (:1 '5‘ ti- '\ 35 The total pulse energy diagnostic was performed on the laser cell while it was in the configuration described in the TRS section. How- ever, the output beam was routed into a calorimeter (Scientech: 364, 10.2 cm diameter) immediately after exiting the output coupler (see Figure 2.11). It was also necessary to experimentally characterize the flashlamp emission properties. It was important to know flashlamp pulse duration and wavelength distribution as a function of time for the modeling studies. Measurement of these parameters was accomplished in the following manner (see Figure 2.15). The Can brewster window was removed from one end of the cavity. A quartz lens was positioned inside the cavity, focusing the ultraviolet light pulses from the flashlamps onto the entrance slits of a monochroma- tor (GCA-McPherson: 218, 0.3 m). A diffraction grating (GOA-McPherson: 2400 grooves/mm, 0.2 micron blaze) dispersed the desired wavelength component through the exit slit onto a photomultiplier tube (RCA: 4832). Voltage was supplied to the photomultiplier tube (PMT) and the resulting signal displayed on an EMI shielded, fast risetime oscillo- scope (Tektronix: 7844) with a one megaohm vertical amplifier (Tektronix: 7A16A). Flashlamp intensities vs time at 0.1 micron in- tervals in the region 0.25 micron to 0.40 micron were recorded (see Figure 2.16). 36 EXPERIMENTAL CONFIGURATION FOR MEASUREMENT OF 1,311) DISCHARGE CIRCUITRY F LASHLAMPS Ch3 n1 MONOCHROMAIOR QUARTZ WINDOWS > AVITY FAST RISETIME OS CIL LOS COPE Figure 2.15 Flashlamp intensity measurement experimental configuration “Intensity 37 Figure 2.16 Typical Flashlamp intensity trace 38 2.3 Results of Time Resolved Spectrosc0py Studies 2.3.1 Introduction Time resolved spectra measurements were recorded at three pressures and two mixtures using two values of outcoupler reflectivity. Results for the two mixtures will be discussed separately. Some general con- siderations will be presented first. It is widely known that mixtures of H2 and F2 will spontaneously form HF. For laser mixtures this is known as prereaction and has been reported by many authors (cf. Reference [64]). The HF formed by pre- reaction rapidly equilibrates, depositing population in low rotational levels in the ground vibrational state. This population, already pre- sent when initiation of the remainder of the H2 + F2 mixture takes place, is undesirable for several reasons. (1) The equilibrated popu- lation acts as an absorber on several low 0 transitions in the v = 1-0 band, increasing their threshold gains. These increased threshold gains are a loss mechanism robbing the laser pulse of a portion of its energy [74]. (2) Hydrogen fluoride is a very efficient self deactivator. It is the most efficient rotational deactivator [68] and one of the most efficient vibrational deactivators [75]. Hydrogen fluoride population, formed prior to initiation, reduces total pulse power, energy and dura- tion by increasing the relaxation rate of the nascent population dis- tribution [76]. (3) Finally, the laser pulse loses additional power and energy because any HF formed and deactivated prior to pulse initiation cannot contribute to pulse energy and pulse power. For these reasons, it is desirable to minimize prereaction in a laser. For this study, there is another important reason: (4) Because of the large HF Einstein coefficient for stimulated emission (and consequently, the Pp '\o Vi. a.' 'h ’n a I ‘5. \ his 1"!1 ’ - i V 39 large absorption coefficient) and the long length of the active medium (53 cm), gain measurements were extremely sensitive to initial HF con- centration. In fact, a concentration of 0.5 mtorr of HF could be detected by probing on the P](3) transition. See Appendix B for a de- scription of this technique. This sensitivity to initial HF concentration necessitated a much more stringent prereaction requirement for this study than in other works. Suchard [14], for instance, claimed prereaction of less than five per- cent initial mixture F2' If this criterion were used, up to 295 mtorr HF could be present initially for the lowest pressure case. This would have prevented measurements using the gain detection system by totally absorbing the probe laser signal. Historically, to minimize prereaction, 02 is added to the H2 + F2 mixture [1, 17, 30, 42, 77, 78]. The rate of formation of HF is reduced by the presence of 02, in some cases insuring stability [78]. It is believed that 02 removes chain carriers (H and F atoms) from the mix- ture [79]. Unfortunately, Taylor, et a1 [79] have shown that in the absence of prereaction, increasing 02 concentration reduces laser per- formance by removing the H and F atom chain carriers. Thus, 02 concen- tration has to be balanced between the minimum amount required to reduce prereaction to an acceptable level and the maximum amount tolerable to laser performance. Reduction of prereaction was deemed to be of primary importance so 02 was added until the initial concentration of HF was below 1 mtorr. This necessitated an unusually high concentration of 02. In addition to problems associated with prereaction, HF lasers suffer problems with parasitic oscillations (cf. Reference 70). A 39 parasitic oscillation is defined as undesirable stimulated emission occurring within the laser medium. Normally, this means lasing between surfaces other than the cavity mirrors. This can be a large loss mechanism as emission proouced by parasitic oscillation is not emitted in the output beam; parasitic power and energy are lost. Two attempts were made to reduce parasitics. One attempt was to cant the cavity end surfaces such that single pass oscillation could not exist between the Brewster window mounts. This should have eliminated parasitic oscillations in the lasing axis direction. Another attempt was to coat all cavity interior surfaces with absorbant black paint (3M: Nextel). This was expected to eliminate parasitic oscillations between surfaces normal to the lasing axis [39]. The disadvantage of this technique was that the ultraviolet initiation photons hitting the cavity interior surfaces were essen4 tially all absorbed. Sinceinitiation'was low with the absorbant black paint in place, it was removed. Removal of the paint helped increase initiation efficiency, always desirable in a flash photolysis case [1]. While precautions were taken to minimize parasitic oscillations, there is no absolute evidence that parasitic oscillations did not exist. There is the possibility that undetected parasitic oscillations did exist because the system exhibits high gain and because the laser cavity walls and windows are not perfectly transmitting. In addition, Suchard, et al. [39] have discussed the existence of circumferential and axial modes of parasitic oscillations and shown them to be important under certain conditions. They have shown that axial grazing modes can be 40 particularly detrimental to laser performance. Calculations similar to those of Suchard, et a1. [39] for axial grazing modes were performed. The resulting parasitic oscillation threshold gain for this mode is 0.019 cm". This is lower than the experimental peak gain reported here for all transitions except P](7) and P](6). Potential parasitic threshold gains presented here are comparable to gains measured here implying parasitic oscillations may well be important. For the time resolved spectroscopy runs, the data taking procedure was as follows: (1) The cavity was purged with He for approximately two minutes to flush out residual HF. (2) The dump valve was closed and the cavity was evacuated to less than 1 torr. (3) Metered flows of He and 02 were admitted. At this point, in rapid succession, (4) the F2 flow was turned on, the capacitors were charged, the H2 was added, and flows were permitted to stabilize. (5) The capacitors were discharged, pulsing the flashlamps and initiating lasing. (6) All gas flows except the He were turned off and the dump valve was opened. Step (1) was then repeated. The duration between shots was timed to be five minutes. This allowed sufficient time to remove all combustion products from the cavity between runs. All the time resolved Spectra data was taken in the form of pictures of oscilloscope traces. A sample is shown in Figure 2.12a. All the data for one mixture at one pressure were then reduced to a plot like Figure 2.17. In this figure, the horizontal lines show the measured duration of each transition, displayed horizontally at the v and J corresponding to the transition's lower level. The time of each transition's peak intensity is denoted as a closed circle. To aid in the visualization of pulse development, selected transition intensity 'I . ,_:e-———=‘— 5 no 150 I60 100 m (pad .1 3 :- o 8 g. 31 8 g F‘= . F—_—_= . F . g__:*-:_-="_ 0 0 :0.” so 00 50 Figure 2.17 Time resolved spectral output: He:0 20.8:l.O:4.6:1.2 (a) 36 torr pressure (b) 102 torr pressure (c) 331 torr pressure (C) Zzesz = 42 time histories were plotted in Figures 2.18, 2.19 and 2.20. Figures 2.18, 2.19 and 2.20 show relative intensities only. Using the aforementioned data taking procedure, day-to-day re- peatability was about 20% for peak intensities, about 15% for pulse durations and about 20% for pulse energies. Shot-to-shot repeatability was better: 10% for peak intensities, 10% for durations and 15% for pulse energies. Most of this variation can be attributed to two causes: (1) The variation of capacitor charging voltage from run-to-run, and (2) The lack of repeatability in cavity mixture and pressure. The first would cause a change in initiation strength from shot-to-shot by varying the photolysis energy supplied to the mixture. This would perturb all pulse characteristics. 'The second would cause a change in the chemistry and relaxation processes, also perturbing all pulse characteristics. 2.3.2 Time Resolved Spectrosc0py Results for the Mixture He:02:F2:H2 = 20.8:l.O:4.6:1.2 For all runs, there is a strong trend of shifting of the transition peak intensity times with rotational level in all bands. There is also a strong trend toward shifting of the transition initiation and termina- tion times with rotational level in all bands. In addition, one observes an increase in pulse duration in all bands with an increase in rotational level up to Pv(6). Pulse duration shows a decrease at transtions above Pv(6). Finally, there is also a general decrease in peak intensity with increasing rotational level. 43 6 (35'; ‘A 6 ‘_L 7 6 A :2 Sf :3 6K 5 I” I a d ‘5' 7 a ._ — V15/\._ )6 u ‘ co 3.. 8 II. 7 A A. 05 — Am 50 100 1 0 200 250 The (pace) Figure 2.18 Spectral time history: He:0 20.8:l.O:4.6:1.2, 36 torr p €535.33 = Intensity (relative unit.) 44 0 20 40 60 80 100 1'1. ( hue) Figure 2.19 Spectral time history: He:0 :F :H = 20.8?1.6:4?5:1.2, 102 torr pressure lntenelty (reletlve unite) 45 UC‘ .77 17> ‘7>'> 177$?) U F579 U. [ r ( ID 26 10 ‘0 $0 10 20 )0 so so Tile (ueec) rte- (ueec) Figure 2.20 Spectral time history: He:0 :F :H = 20.8:l.O:4.6:1.2, 331 torr pressure 2 2 2 46 The observed monotonic trends of shifting transition initiation, termination and peak intensity times are strong evidence of a rotational distribution approximating Boltzmann. Furthermore, peak intensity with- in a given band generally appears on the transition with the maximum upper level population in that band consistent with a Boltzmann distribution of rotational levels at about 400 K. Nevertheless, nonequilibrium rotational populations are evidenced by simultaneous lasing on adjacent transitions. There appears to be cascading linking the transitions P2(3) and P](4), P3(4) and P2(5) and P5(3) with P4(4) and P 5). 31 Cascading is a phenomena due to lasing where the stimulated emission flux on one transition assists in the inversion buildup, and subsequent lasing, of a second transition. The two transitions are linked by a common, intermediate energy level: HF(v = l, J = 3) for instance. When lasing occurs on the P2(3) transition, population is radiatively transferred from HF(v = 2 J = 2) to HF(v = l, J = 3). This population transfer helps to build up an inversion between HF(v = l, J = 3) and HF(v = 0, J = 4), leading to lasing on the P](4) transition. P2(3) and P](4) are then said to be cascade linked transitions. For each of these pairs of transitions linked by cascading, a peak in the upper transition stimulated emission intensity is closely followed by a peak in the lower transition stimulated emission intensity. This is particularly noticeable for the 36 torr pair P 2(3) and P](4). Initially P (4) rises just after P 2(3) rises, falls just after P 2(3) falls, and rises again just after P 2(3) rises for the second time. Also note that lasing is observed on transitions in the v = 6-5 band. No such lasing was observed for the 36 torr case. This behavior ‘s‘s. 'rie's I re‘ ,0 .24. . .2, 4,- v .1 'A or 73385 ? I .IC 1“] 51-5 V ‘7 r .139. "h "~91“ . w, m‘ :1 mi 9 I 1' I‘Lfl \Hu 0 ‘l 47 is probably a result of a more favorable competition at 102 torr between the net result of pumping plus deactivation and the cavity loss mechan- isms. As the pressure increases for a fixed gas composition, the net inversion produced by pumping minus deactivation increases. From the results presented in Figure 2.17 for the 102 torr case, it appears that for this case, the inversion exceeded the threshold gain due to cavity losses while for the 36 torr case it did not. This increased rate of stimulated emission on the v = 6-5 band for the 102 torr case would help to explain the anomalous behavior of the P5(5) transition. This transition has a significantly higher peak than P5(4), hence, it does not follow the trend of decreasing peak intensity with increasing rotational level. This could be explained by cascading from P6(3). Another possible explanation is that the assumed pumping distribution is in error. An alternative pumping distribution is suggested in Chapter 3. The v = 6-5 band behavior also appears anomalous for the 331 torr case. Here, none of the trends observed for the other levels are obeyed. Because of the high rate of hot pumping into v = 6 [8], one would expect stronger lasing. However, because hot reaction pumping into v = 7 is much less than the pumping into v = 6, V-V exchange and reverse V-R;T (R-V) energy transfer may be transferring a significant amount of population between v = 6 and v = 7. This would diminish the v = 6-5 inversion. Furthermore, since V-R,T energy transfer is thought to 2 7 scale as 'V ' [25], fast V-R,T relaxation will also contribute to the reduction of the v = 6-5 inversion. These mechanisms may account for the behavior of the v = 6-5 band. 48 The unique time history of the v = 6-5 band may also be explained by cascading. It appears that lasing on P6(4) is strongly dependent on the behavior of lasing on P5(5). P6(4) does not initiate lasing until P5(5) reaches its peak intensity. Thus P5(5) lasing would assist in P6(4) lasing by improving the population inversion by cascading. Recall that cascading occurs when the lower transition lases, removing population from the lower level of the upper transition, which then lases. In addition, P6(4) terminates shortly after P5(5) terminates. This implies that when the mechanism assisting the population inversion terminates, so does the laser pulse. If the behavior of P6(4) were neglected, P6(5) and P6(6) would follow the observed trends well. These two transitions exhibit behavior in agreement with the trends for peak intensity and termination times. They disobey the trend for initiation time only slightly. It is clear that the intensity, initiation and termination times, and duration of P](5) do not fit the observed trends. A possible ex- planation is partial absorption of the P](5) signal by a foreign species. This absorption would tend to decrease the signal intensity while leaving the peak position unchanged. This would effectively shorten the pulse duration. This behavior is consistent with that observed for P](5): An initiation time later than predicted by the trends, a termination time earlier than predicted by the trends, a peak intensity lower than pre- dicted by the trends but a peak position in agreement with the trends. Possible candidates as absorbers are impurities in the F2 supply (SF6, 002, 02, N2 or CF4) or products generated during lasing (OH or H20). Oxygen and nitrogen can be eliminated immediately since they have no transitions which absorb in the 2.7 um region[80]. ad- D.\.i ‘r r: ‘\ Ah\ ? 49 Sulfur hexafluoride can also be eliminated from consideration. First, it has no infrared active transitions near 2.7 pm [31], and second, the reported absorption coefficient at the P1(5) wavelength is 1.3 x 10'5 cm"1 torr'1 [82]. Coupled with the estimated SF6 impurity in the F2 supply and the 100 cm absorption pathlength accounted for, this would yield only 7.0 x 10-7% absorption of the P1(5) signal over the entire cavity length. Carbon tetrafluoride can also be eliminated. It has only continuum absorption in this region of the spectrum [83]. This continuum absorp- tion would affect all transitions in this spectral region equally and would not be responsible for the anomalous behavior of a single line. The hydroxl radical has infrared active transitions in the 2.7 pm . region of the spectrum, however none are within 3 cm.1 of the reported P](5) transition wavelength [84]. Since the linewidths of both 0H and HF should be considerably less than 3 cm"1 , even for the 331 torr case, 0H should have negligible absorption for the P](5) transition. Carbon dioxide also has infrared active transitions in the 2.7 um region of the spectrum [80]. There are three transitions near the measured wavelength of P](5): 3741.4598 cm'1 [85]. These are transi- tions at 3741.445 cm'], 3741.471 cm“ and 3741.368 cm" corresponding to 41-0440, 2221-1220 and 1o°1-oo°o bands transitions occuring in the 14 respectively. Although these three transitions lie within the P](5) line profile, they probably do not contribute significantly to P](5) absorp- tion. First, the 1441-0440 and 2221-1220 transitions are hot bands and will have very little population in their lower levels for tempera- tures encountered here. Their absorption will thus be negligible [86]. Second, the remaining transition, 10 1-00 0, is a combination band with 50 a very low absorption coefficient. Approximately 100 torr of CO2 would be needed in the laser cavity to produce the absorption necessary to diminish P](S) [86]. Consequently, C02 absorption does not appear to be the cause of anomalous P](5) behavior. Water vapor has an infrared active absorption line near 3741.4598 cm']. Fraley and Rao [87], report an OOl+000 transition at 3741.3088 cm']. This is within the line profile of P1(5) and could account for the observed absorption. Other investigators have credited water vapor absorption with per- turbing P1(5) intensities. Ultee [88] and Jacobson, et al_[89] credit water vapor absorption with being responsible for anomalous behavior of their reported P](5) intensities, although they do not list the water vapor transition responsible. Galochian, et a1 [29] and Greiner, et a1 [50] also report P](5) intensities inconsistent with the remainder of their observations. They do not suggest a cause. Due to the evidence presented, it is likely that the anomalously low P](5) intensities observed are due to water vapor absorption of the signal. For the 331 torr case, the behavior of P](5) may seem inconsistent with the partial water vapor absorption of the signal as was suggested for the 36 torr and 102 torr cases. However, careful inspection of the P](5) trace reveals the peak intensity is barely in accord with the trend of increasing peak intensity with increasing rotational level up to J = 6 within a given band. An increase in peak height for this tran- sition would still be consistent with the trend observed for the 331 torr case. Thus, partial water vapor absorption of P1(5) is still a viable explanation for the observed behavior. 51 2.3.3 Time Resolved Spectroscopy Results for the Mixture :H He:O = 22.0:1.0:2.7:l.0 2‘F2 2 For this mixture composition, there is monotonic shifting of both transition peak intensity times and transition termination times with. increasing rotational level. These trends are well obeyed except for the 81% outcoupler reflectivity 331 torr case. In addition, transition initiation time increases with rotational level within a band. This is generally obeyed for all pressures with the following exceptions: P3(5) starts before P3(4), P4(4) starts before P4(3), and P5(4) starts before P5(3). Pulse duration also increases with an increase in rotational level. This behavior is strongest for the 331 torr case with 97% outcoupler reflectivity and for the 36 torr case. It is obeyed up to Pv(6) for the remaining two cases. Cascade linked transitions are apparent for all three cases utiliz- ing the 97% reflectivity outcoupler. There is no apparent cascading for the 331 torr case using the 81% reflectivity output coupler. Furthermore, cascading is more evident at 36 torr than at 102 torr and more important at 102 torr than at 331 torr. This is based on the observation that the number of cascade linked transition pairs decreases with pressure. It should be noted that no lasing is observed on v = 6-5 band transitions. The absence of v = 6-5 lasing is most likely explained by the steep dropoff of the output coupler reflectivity with increasing wavelength. See Figure A.2. It is possible that the resultant in- crease in threshold gain with wavelength is such that the v = 6-5 transition population inversions are insufficient to attain threshold. .52 Figure 2.21 Time resolved spectral output: Hezozzesz = 22.0:l.0:2.7:l.0 (a) 36 torr, R = 0.97 (b) 102 torr, 90 = 0.97 (c) 331 torr, R0 = 0.97 (d) 33] torr, R = 0.81 0 53 ON. P :83 is» .81. I: L: A. m m 3. e. e. II. 0 l I O O lit a a I.“ O Ilnlllnlu Aev . . Aev 1 > 98.5 a .98.: ‘3. 3. 2. 8 a.“ o 8. 8o 8... I l . O O I. .‘ Mu o a In _. ...|.o do a n Illoi 1‘ 1. Adv Amy C . .N.N oesmea i» re. . a '3 av :pn - I1 U e...' A fl 03v J 0 ‘1‘ 0V 9 ‘ e :.SJ Ito 54 As mentioned in Section 2.3.2, the observation of increasing pulse duration with increasing rotational level, shifting of initiation time, and monotonic shifting of peak and termination times with increasing ro- tational level, are evidence of a near Boltzmann distribution of rotation- al population within a given vibrational level. Since these trends are also observed here, it is likely that a near Boltzmann distribution of rotational levels exists for this case too. It should also be noted that P](5) is absent. This again, is pro- bably attributable to water vapor absorption. In this case, it appears that absorption is so strong, P](5) never attains threshold. Upon comparison of the two cases at 331 torr, it is evident that a change in output coupler reflectivity has no effect on the number of transitions lasing, the number being the same in each case. Hence, for this case, output coupler reflectivity has little effect on an indivi- dual transition's initiation time or on an individual transition's ter- mination time. 'Outcoupler reflectivity causes a slight effect on pulse duration: of the nineteen transitions observed to lase, ten have longer durations using the 81% reflectivity output coupler, seven have longer durations using the 97% reflectivity output coupler, and two have dura- tions essentially unchanged. This behavior runs counter to the expected result that an increase in the value of output coupler reflectivity would lower threshold gain, leading to increased pulse durations and an increase in the number of transitions lasing. There appears to be no explanation for this lack of agreement. c~' , 1 v’tuif *1 .i, t" 5 Tn llle c ‘.‘7"1 .‘v1/ K 4., ‘1':ec 55 2.3.4 Results of Pure Rotational Lasing Studies As noted in Section 2.1, the portion of laser power emitted at pure rotational wavelengths is of prime importance to researchers and develop- ment engineers interested in high power applications. Pure rotational lasing has also been shown to be linked to V-R,T energy transfer [61, 62, 80]. For these reasons, attempts were made to measure the time solved spectroscopy of pure rotational lasing transitions. The experimental configuration used to investigate pure rotational lasing was discussed in 2.2.2. Two changes were made: (1) The 4 micron blazed monochromator diffraction grating employed in the P-branch lasing studies was replaced by a 16 micron blazed grating. (2) A long pass interference filter was inserted into the optical path immediately prior to the monochromator entrance slit. This filter had a 9 micron cut on point with <0.l% transmission below 9 microns and nominal 50% trans- mission from 9 microns to 20 microns. The interference filter served to eliminate all P-branch lasing signals, passing only radiation due to pure rotational lasing transitions. Initial testing was done with the mixture He:02:F2:H2 = 22.0:1.0: 2.7:1.0° At first, the 81% and 97% peak reflectivity output couplers were used. No rotational lasing was observed for any of the three pressure cases using these two output couplers° This could be due to the drop in output coupler reflectivity above 4 microns. This decrease in reflectivity continued out into the middle IR such that at 16 microns, the center of the pure rotational transition wavelengths, the reflecti- Vity was below 20%. A 20% reflectivity would yield a threshold gain of (3.012 per cm. It appeared that pure rotational lasing transitions never exceeded threshold gain. 56 To correct for this, an output coupler with approximately 95% re- flectivity at 16 microns was substituted for those used previously. This lowered the threshold gain for pure rotational lasing to approximately 0.0011 per cm. This output coupler had low reflectivity in the P-branch region of the spectrum: Nominally 40%. The monochromator was tuned to transitions originating on J = 13,14,15 and 16 for vibrational levels v = 0, l and 2. These levels were chosen by noting the results of other researchers who had reported pure rotational lasing [57, 58, 60, 91], most "OtBPIY Pimentel and coworkers [62,90]. Note that this rotational lasing was not observed from mixtures of H2 + F2, but from boron trihalides, vinyl I or CF Br + 3 3 hydrocarbons and in optically pumped HF. Transitions above J = 16-15 fluoride, l.l-difluoroethene, ClFx+ H2, freons + H2, CF could not be detected due to detector wavelength limitations. For this mixture, no rotational lasing was recorded for either the 36 torr or 331 torr cases. Intermittent pure rotational lasing was de- tected at 102 torr for the transition v = 1, J = 15-14. Lasing was observed to occur only three times out of twenty-four trials. This is not sufficient evidence to report TRS for pure rotational lasing. How- ever, this investigation does imply the existence of pure rotational lasing. An investigation of pure rotational lasing was also conducted for the mixture He:02:F2:H2 = 20.8:l.O:4.6:l.2. The same transitions were examined as for the other mixture. All three output couplers were used. Pure rOtational lasing was not observed for any of these three pressure cases. II- 6. I? .; ‘09 r"; vi. '521‘ a A . \FF. '5'. 57 The conclusions to be drawn from the examination of pure rotational lasing are: (1) For higher F2/H2 ratios, hence stronger initiation, pure rotational lasing is unimportant and probably does not exist. (2) For more strongly diluted systems with Fz/H2 approaching stoichiometric, rotational lasing may exist. (3) If output coupler reflectivity is low, near 25% for pure rotational wavelengths, pure rotational lasing is negligible for all compositions reported here. 2.3.5 Discussion Several trends become apparent upon comparing the runs presented in Sections 2.3.2 and 2.3.3. (1) As expected, individual transition pulse durations show a decrease with an increase in mixture pressure. All but two transitions exhibit this behavior for the mixture composition He:02:F2:H2 20.8:l.O:4.6:l.2. For the mixture composition He:02:F2:H2 22.0:l.0:2.7:l.0, the trend is observed for transitions which terminate on rotational levels abOve J = 4. No pattern is evident for transitions which terminate on rotational levels J = 3 and 4. (2) Individual transition pulse durations decrease with an increase in initial percentage of mixture F2. This behavior is observed for both the 36 torr and 102 torr cases, there being only one exception to each. It is not observed for the 331 torr case. (3) The number of transitions lasing within a given band increases with an increase in initial percentage of mixture F2. This is striclty true for the 36 torr case. It is also true 'e 31- I") 58 for the 1-0, 3-2, 5-4 and 6-5 bands at 102 torr and for the 1-0, 4-3, 5-4 and 6-5 bands at 331 torr. Two more trends become apparent upon comparing the raw data used to generate the figures in Sections 2.3.2 and 2.3.3. (4) As one would expect, individual transition peak intensities increase with an increase in mixture pressure. There are no exceptions when comparing the 36 torr and 102 torr cases, but there are four exceptions when comparing the 102 torr and 331 torr cases on rotational levels above J = 4. (5) Individual transition peak intensities increase with an increase in initial percentage of mixture F2. As in (3) above, there are no exceptions for the 36 torr case. This behavior is generally true for the 102 torr and 331 torr cases. However, there are several sets of intensities which do not follow this pattern for both cases. The increase of individual transition peak intensity and the decrease of individual transition pulse duration with increasing pressure are probably due to binary scaling of the chemical pumping and relaxation kinetic processes. Binary scaling would yield faster pumping and relaxation causing increased rates of fuel and oxidizer consumption and increased rates of product deactivation. This, in turn, is probably responsible for the shortened individual transition pulse durations. The increased rate of chemical pumping would also allow the population inversions generated to compete more favorably with threshold gain loss mechanisms and with stimulated emission allowing higher intensity build- ups. This would result in larger transition peak intensitites. f..l. as o r. ' .Qe ‘ “ n,~' ‘1‘: 59 We conclude the remaining trends, (3), (4) and (5), are due to an increase in the initial percentage of mixture F2° However, the effects of increasing initial mixture F2 are probably distorted by the change in outcoupler reflectivity between the two mixtures. A discussion of these two effects follows. Only two variable are changing during the comparison of the runs at the two mixture compositions. The first is the percentage of initial F2 in the mixture. The second is the nominal output coupler reflectivity. These changes in initial mixture F2 percentage and output coupler reflectivity should affect pulse behavior in opposite ways. One of the effects of increasing the initial percentage of mixture F2 to increase the rate of chemical pumping. This leads directly to an increase in the consumption rates of fuel and oxidizer and a consequent shortening of the pulse duration. Another consequence is more favorable competition between inversion buildup and losses through deactivation and radiation. Hence, one notes: (1) an increase in the number of transitions lasing and (2) increased peak intensities. The conclusion is that an increase in the initial percentage of mixture F2 is partly responsible for observations (3) and (4). It is not clear what the effect of increasing initial mixture F2 percentage would be on observation (5). Changing the output coupler reflectivity has the opposite effect on pulse behavior for (3) and (4). Increasing the output coupler reflectivity decreases the threshold gain, lowering cavity losses, allowing more favorable competition between pumping and losses, .yielding a greater number of transitions lasing and leading to increased 60 transition durations. This is not observed. There are two possible explanations. The first, as stated in Section 2.3.3, is that the effect of the change in outcoupler reflectivity under these mixture conditions is minimal. The second is that the effect of increasing the initial percentage of mixture F2 dominates the effect of the outcoupler reflectivity. There are further observations to be made. One is that for all pressures and all mixture compositions, the time resolved spectra results indicate a nearly thermalized, or near Boltzmann, distribution of rotational levels. This is due to observed shifts in pulse initiation, pulse termination and pulse peak intensity times with increasing rotation- al level. Additional evidence is the increase of pulse duration with an increase in rotational level. As previously stated in both Sections 2.3.2 and 2.3.3, these nearly thermalized rotational distributions are evidence of a rotational relaxation mechanism that is fast compared to chemical pumping and V-V and V-R,T relaxation. Rotational relaxation may be much faster than chemical pumping and vibrational transfer and relaxation, but it is slower than the stimu- lated emission buildup time. Kerber, et a1. [70] report the stimulated emission buildup time to be on the order of 2L/c. For this work, 2L/c = 9.5 x 10'9 sec while model results show the rotational relaxa- tion time to be 5.0 x 10‘8 sec at 331 torr. The rotational relaxation times at 102 torr and 36 torr are longer than that at 331 torr. ear! ‘Ap- be .1 . ,e “L31 In... 1 wt . x I. - C/e PA ('1.- .1 K l.l 1.x) (1') 61 Another observation is that P](5) lasing is present only for the mixture He:02:F2:H2 = 20.8:l.O:4.6:l.2. The proposed increase in pumping percentage rate with increased initial mixture F2, presented above, is consistent with this. It is possible that this increase in pumping rate would overcome the absorption losses which are present for P](5). A final observation is that the relative importance of stimulated emission as a population transfer mechanism is increasing with an increase in mixture pressure. This is more evident for the mixture composition He:02:F2:H2 = 20.8:l.O:4.6:l.2. For this mixture, cascade effects increase with pressure. More transitions are linked by cascading and cascading is more dominant at higher pressures, especially for higher vibrational bands: compare v = 6-5 at 331 torr and 102 torr for this mixture composition. Peak intensities for individual transitions also increase with pressure. As expected, population transfer due to stimulated emission increases with pressure. The latter two effects are related to the increase in cascadinq. As mentioned in Section 1.2, there are six other studies reporting time resolved spectra for H2 + F2 systems. Of these six, four [14, 19, 42, 43] utilized flash photolysis initiation and two [17, 52] used electric discharge. Only the results of one of the electric discharge works will be compared: those of Parker and Stevens [17]. Basov, et a1. [52] do not present a complete set of spectra, but instead present only selected transitions time histories. A meaningful comparison between that work and the results presented here is not possible. Con- sequently, only the work of five authors will be presented: References 14, 19, 42, 43, 17. These results will be compared individually with f!- ‘1- ..1 an IV "I‘ § 1 62 the results of the present work. All pertinent information for each of these five works is presented in Table 2.1. Suchard, et al. [14] present results for a 75 cm long, 1.2 cm diameter tube filled with 50 torr of a He/FZ/H2 mix of composition HezF2:H2 = 60:1:1. The resonator configuration was a 310 cm radius, 98% reflector separated by 100 cm from a 0.2 cm hole outcoupler. A photolysis pulse 55 nsec long initiated the reaction, dissociating ap- proximately 1% of the initial F2. Greiner [19] presents results for a 15 cm long, 0.7 cm diameter tube filled with 45.5 torr of a Oz/Fz/H2 mix of composition 02:F2: H2 = 1:10:ll.0. A photolysis pulse 18 nsec long initiated the reaction, dissociating an estimated 0.75% of the initial F2. This resulted in a pulse 8 nsec long with individual transitions having a mean duration of 2 usec. Lasing was observed only for the lowest four bands: 4-3, 3-2, 2-1, 1-0. There was some evidence of cascading: Greiner [19] gives no information on the configuration of his optical cavity, but does note that the lasing mixture was originally cooled to 200K to 218K. The work of Suchard [42] is very similar to that of Suchard, et a1. [14]. According to Suchard [42], the differences are a different op- tical cavity configuration, a different gas composition and a different cavity length. The values from Reference 42 are a 43.5 cm cavity, a mixture consisting of He:F2:H2 = 80:2:1 and an optical cavity consist- ing of a 90% reflectivity outcoupler separated by 90.5 cm from the 98% reflectivity, 310 cm radius mirror, also used in the study of Reference 14. e-.Ilu\: U‘Hh p‘ ‘~ ‘4 (a -\-l-‘- .fu. utN .10 ~ Q .~.\ €53 o=.uauuou goat .A.oe\.a..e .a..na\.__.~e Aa.na\.a.~. EgEAEeASeJ .e.m.\.e.~.\.e._. .e.~.\.e._. .N...\._... .n.n.\A..~. .o.~e\.o._e .~.~e\.o._e ...~.\.A._. .m.~.\.e... .A.A.\.e.~. .evna\.evua\.a._e ...A.\.e.~.\.e... .n..a\...ne .N..A\An.na A~.A.\.n.~. 39.35.? suoiztsueul Hmm .me .me .m. .A_ .epu mooeoeoeoa seat me_=.o¢ wee to seeee=m (seen) uotzeunu uotatsuedl ueau n.— mm 00 0 ‘u01zr4na '3lnd [9101 ulna-hm.“ @Q‘O‘ OQNHO. e M I I 'e-NM. Q—N ONO" outnu— pune Jae sanii “90443840 saute '5353' cm o—m so-aamuuao O SREEEE (I [WING 3044;“ 5’I1I5I33'IJ'U (IR) 119's 40441H :0 m— n a.c x — cum. ms m.c¢ m.nv ~.o m— .- N p h deseleta Katheg 'M150'1 4114.3 (In) Butaeds 3034;“ en m.~ _._.m~.o o. c—— _un"~"m on —"~"ouoa m.mc mn.o o.——uc—“—uo pupuonao ma. 3. Mm m m mm... mm .aM .. “” swunu .. mm Armm _.~ o_eee m_oo.o ~.o t ~.c eucagua.o u.cuuo.u an OMA o..»_oaoee g...» o..s_oooee ewe—e .e. a. ..ea_o.oee go... mo.n mm n.m>—ouogm gun—u P Kilgfi::/£6Jau3 “(:eSn) uotzeJnu pouzau “OII'IIIUI ~— nv ~v @— 43030" eoueuegau I'\) 0;; e-i .56 ‘..r ‘\ .. $5.1. ‘ e':n' ‘ll. ‘3;‘ 64 Borisov, et a1 [43] report TRS for a 114 torr mixture of HezOZ: F2:H2 = 0:2:3:l contained in a 75 cm long 2.4 cm diameter cavity. The optical cavity consisted of two flat mirrors of 96% and 6% re- flectivity. The flash photolytically initiated pulse had a total dura- tion of 4 nsec with individual transitions having a mean duration of 1.7. nsec. The previous four authors all used flash photolysis for laser initiation. The remaining work used a pulsed electric discharge for initiation. As mentioned in the introduction, electric discharge initiation produces charged Species which can complicate the medium chemistry. This difference in chemistry could be responsible for a difference in results between electrically and photolytically initiated systems. Parker and Stevens [52] present results for a 36 torr mixture of He:02:F2:H2 = 10:0.25:l:1 flowing throngh a l x 0.8 x 15 cm long channel. A 300 cm radium "high reflector" and 80% reflectivity flat mirror comprised the optical cavity. Electric discharge pulses of 0.2 + 0.7 nsec produced laser pulses of 6 pSEC total length with a mean transition duration of approximately 2.3 nsec. The authors claim an initial F2 dissociation of about 2.5%. Our 36 torr, mixture Hezozzesz = 20.8:l.O:4.6:1.2, TRS results are consistent with the results of Greiner [l9] and Parker and Stevens [17]. These studies report durations less than that presented here: 8 usec and 4 nsec vs. 200 usec, respectively. This is to be expected as: “rear .. wt. . I‘ .I 9;“ "7"...“ p. "'4. 65 (1) Both studies report fluorine dissociation values over two orders of magnitude less than this study. Pulse duration is known to decrease strongly with initiation strength [70]. (2) Both studies report cavity lengths less than this work. Pulse duration has also been shown to decrease with decreased cavity length. (3) Greiner [19] has a much less dilute mixture, also leading to a decrease in pulse duration relative to this work. Our 36 torr, mixture He:0 :F = 22.0:l.0:2.7:1.0, TRS results 2 2‘H2 are consistent with the results of Suchard and coworkers [14, 42]. The durations reported there are shorter due to the higher level of initiation reported (1% vs the 0.0025% reported here) and their slightly higher cavity total pressures. The mixtures employed by Suchard and coworkers are more dilute than this case, which could lead to some lengthening of their pulse duration compared with this. However, it is expected that this will be a much smaller effect than the over two order of magnitude difference in initiation strength. The results of Borisov et al. are consistent with our 102 torr results with mixture composition He:02:F2:H2 = 20.8:l.O:4.6:l.2. They report shorter pulse durations because of higher threshold gain and a much less dilute mixture. The former is due to the low (6%) outcoupler reflectivity employed and leads to a decrease in pulse duration [70]. The latter has also been shown to lead to a decrease in pulse duration. 66 2.4 Results of Small Signal Gain Studies 2.4.1 Introduction Small signal gain was measured on this laser at three pressures and two mixtures. Results from the two mixtures will be discussed separately. For the case with nominal mixture He:02:F2:H2 = 22.0:l.0:2.7:l.0 small signal gain could not be measured at 36 torr. The small signal gain was too low to cause a noticable perturbation in the amplitude of the signal beam. All attempts showed neither positive nor negative gain. Small signal gain was not measured for P](6), P2(3) or P2(4) at 102 torr and for P1(3), P2(3) or P2(4) at 331 torr. This was due to an old and erratic probe laser which refused to lase on these transitions. Several different probe laser gas mixtures and discharge currents/voltages were tried in an effort to get the probe laser to oscillate on these transi- tions. All failed. For the mixture He202: 2:H2 = 20.8:l.O:4.6:1.2 small signal gain was measured only at 102 torr. In addition, small signal gain was not measured for P2(6), P2(7), P2(8), P1(3) or P](8) for the reason stated in the paragraph above. It was originally proposed to probe small signal gain (SSG) on the v = 2-1 and v = 1-0 transitions with the Helios SF6/H2 probe laser and to probe SSG on all other bands with a Laser Analytics Tunable Diode Laser (TDL). Upon the advice of Butler of Laser Analytics [92], the latter was dropped. It was felt that there would be two insurmountable problems. The first was attenuation of the weak (<1 microwatt) TDL probe signal. This would have been caused by scattering off optical elements and atmospheric dust particles and absorption by optical 67 reflecting and transmitting elements and by the atomsphere. It was felt the signal level would be too low to register on the detectors. The second, and more serious problem, was electromagnetic interference (EMI) of the TDL power supply. High transient currents in the flashlamp dis- charge circuitry would produce time varying magnetic and electric fields which would in turn perturb circuit elements within the sensitive TDL power supply. This would result in mode hopping of the TDL causing a shift in probe laser signal frequency. This could not be tolerated. For these reasons, the TDL probe was not used. One alternative to the TDL as a gain probe was to use SF5 + HI in the existing SF6/H2 Helios probe laser. This was suggested by Jeffers [93]. Due to previous results using the SF5 + HI mixture [50], Jeffers felt this mixture might produce lasing on transitions up to v = 6-5 using the probe laser described in Section 2.2.2. There were three reasons why no measurements were attempted using this technique. First, iodine atoms remaining as products from the reaction F + HI = HF + I could combine with fluorine atoms to form IF, and subsequently, IF5 and IF7 [94]. These compounds are potentially very damaging to vacuum pumps. It is believed that the IFS and IF7 rapidly decompose vacuum pump oil and seals often leading to vacuum pump failure [93, 94]. Second, Jeffers believed any HI not consumed in the probe laser reaction zone would have an effect on the vacuum pumps similar to that of IFS and IF7 [94]. Third, there was a lack of suffi- cient facility time to complete the TRS and SSG measurements at the two mixtures presented here and undertake SSG measurements for the higher vibrational bands. It was felt that the TRS and SSG measurements 68 presented here were more important than the SSG measurements for the higher vibrational bands. Thus, facility time was utilized in a manner con- sistent with these goals. A second alternative to the TDL as a gain probe was the F-center laser marketed by Burleigh, Inc. This system consists of an argon-ion or krypton-ion pump laser and frequency shifting crystal which produce tunable light in the 2.1 to 3.3 micron region of the spedtrum. Several transitions of interest in the v = 3-2, v = 4-3 and v = 5-4 bands would be accessible using this device. However, because the combined cost of the ion pump laser and frequency shifting crystal was high, this system was not tried due to lack of funds. There were no further substitute sources available which would os- cillate in the required region of the spectrum. Hence, gain was not measured for bands above v = 2-1. The data taking procedure for the SSG runs was identical to that of the TRS study. This procedure was described in Section 2.3.1. For the SSG runs, it was necessary to synchronize the firing of the flashlamps with the presence of the probe laser beam in the laser cavity. This was accomplished by gating the lamp signal to the probe laser signal via a HP 214A pulse generator. The lamps would fire only when they simul- taneous1y received a ready signal from the probe laser and the operator. All small signal data was taken in the form of oscilloscope trace pictures. A sample is shown in Figure 2.12b.. For each run, both the signal intensity and reference intensity traces were digitized at ap- proximately twenty discrete points. The results of digitization were used to calculate the small signal gain at each point. The gain points 69 were plotted and a smooth curve was drawn through the result. Gains for all transitions measured in each band were then graphed on a single figure. There are two main sources of uncertainity in the gain measurements. One of these is associated with the uncertainty in the frequency of the probe laser. This uncertainty arises from the possibility of the probe laser Hmode-hopping"; shifting from one cavity longitudinal mode to a cavity longitudinal adjacent mode. This would cause a shift in probe laser frequency. Since medium gain is highly frequency dependent, through the lineshape profile term, this could lead to measurement of a significantly different gain. Mode h0pping should not have been caused by thermal effects as the cavity was constructed of material (INVAR) with an extremely low thermal expansion coefficient. However, it could have been caused by vibrations from the probe laser vacuum pump. If these vibrations were transmitted to the probe laser resonator optical elements, mode hopping could occur. A second cause of uncertainty would be due to lack of measurement precision when reducing data. Measurement precision is limited by the ability to resolve oscillosc0pe traces from the data photographs. For this work, the minimum resolvable intensity quotient was approximately 1.04. This, coupled with a gain measurement length of 53 cm, yielded an uncertainty of 0.00074 per cm. Using the data collection and reduction techniques discussed re- sulted in shot-to-shot repeatability of approximately 15%. Day-to-day repeatablilty was nearly 25%. It is expected that probe laser mode hopping affects gain magnitudes by less than 25%. See Appendix 0. 70 Much of the day-to-day variation can be attributed to a lack of repeatability in cavity mixture and pressure conditions and to a varia- tion in capacitor charging voltage. The major effect would be to vary initiation strength by varying F2 concentration and by varying F/F2 due to initiation. As will be seen in the following sections, varying F2 concentration can alter the gain behavior significantly. One possible source of error that was eliminated was saturation of the medium by the probe signal. Saturation was shown to be unimportant by the following experiment. Two sets of gain measurements were made at 102 torr. The probe laser intensity was varied by a factor of ten be- tween the two cases. In both instances, the SSG time histories were in good agreement. Since the SSG time histories were independent of probe intensity, the transitions under investigation should not be saturated. In addition, since the 102 torr case should saturate easier than the 331 torr case, saturation should not be important in either case. 2.4.2 Results of Small Signal Gain Studies for the Mixture He:02:F2:H2 = 20.8:l.O:4.6:1.2 For this case, gain was measured at 102 torr only. There are several points of interest. First, gain initiation time increases with increasing rotational level. In addition, time to peak gain and gain termination time also increase with increasing rotational level. Gain duration increases with increasing rotational level while peak gain magnitude decreases with increasing rotational level. 71 0031 2H Mind Figure 2.22 Small signal gain time history: Hezozzesz = 20.8:l.O:4.6:1.2, a) v=1-0 band, 102 torr b) v=2~l band, 102 torr 72 Prereaction of the initial cavity mixture is observed. Evidence is seen in the initial gains for the v = 1-0 band being nega- tive at time zero. This implies more population in v = 0 than in v = 1. Since the rate of pumping into v = 015 believed to be negligible [85], this must be population deposited into levels by pumping and subsequent- ly relaxing to v = 0. This must be occuring before initiation and hence, is prereaction. Analysis of the trends observed implies strong rotational relaxation of the nascent pumping distribution. This is evident from the behavior of the v = 1-0 band transitions. This sequential transition history is probably caused by a near Boltzmann distribution of rotational levels in a system whose temperature is monotonically increasing. This behavior is analogous to the rigid sequencing of transitions in a given band for rotational equilibrium based computer simulations [70]. This trend is not as pronounced for the v = 2-1 band. Conclusions drawn from this section are: Prereaction exists, even with the addition of large amounts of 02 as an inhibitor. The R-R,T mechanism is important for both levels, possibly rapid enough to make nonlasing mixtures appear rotationally equilibrated in the v = 1-0 band. 2.4.3 Results of Small Signal Gain Studies for the Mixture He:02: 2:H2 = 22.0:l.0:2.7:l.0 For this case, gain was measured at pressures of 102 torr and 331 torr. For the 102 torr case, there exists a general increase 73 eeou .mm .oeee _-~1> .ggoa New: “meow .o. p. N. N. o. F. o. NNu _1Nu1> 1.2:! 3N ON a .896 {lam cunt . 80.9 .3»; . cad Acv Eu. . 2 a. .5... In» at. ale 2 . 8o... 5 Er .2... . .3 . “Our A 5.0 r E s I- “I" I- “I IV. .eeee o- .1) av .eeoe _mm .eeee o- _1> av .eeoe «a. o. m: .xgoumpg as.» spam pm=m_m p—esm mN.N mc=m_m a 1.... I.» Amy .3: 3.: g6 . 906 (II-001m we all um 74 in gain initiation time with increasing rotational level. This is not observed for the 331 torr case. There is an increase in peak gain time with increasing rotational level and an increase in gain termination time with an increase in rotational level for the 102 torr pressure case. In addition, gain duration increases with increasing rotational level for the 102 torr case, but not for the 331 torr case. Gain on the v = 1-0 band is negative at time zero for both cases. As befbre, this implies initial absorption and hence prereaction of the H2, F2, He, 02 mixture. The technique used in this work to measure small signal gain is sensitive enough to detect HF concentrations above 0.5 mtorr. See Appendix B for. details. Measurements of prereaction show a concentration of 1.5 mtorr of HF per 0.5 torr of F2 initially in the cavity. This is approximately 0.016 molar percent HF in the F2 supply and is consistent with the F2 supply batch analysis presented in Section 2.2.1. Prereaction is more important at 331 torr than at 102 torr. Analysis of the trends for this case show strong rotational nonequilibrium effects. Since it was the observed regular shifting of gain initiation, peak and termination times that was given as evidence of rotational thermalization, and hence, strong rotational .relaxation, their absence implies a less important rotational relaxation mechanism for this case. 75 For this case, it can be concluded that prereaction still exists and that the R—R,T mechanism is less important than for this pressure at the mixture He:02:F :H 2 2 = 20.8:l.O:4.6:l.2. 2.4.4 Discussion The effects of increasing the percentage of F2 in the mixture and the effect of increasing mixture pressure can be determined by comparing the results of the three cases presented. This will be done below. The two 102 torr cases will be presented first to determine the effects of an increase in the percentage of mixture F2. The two cases for He:02:F2:H2 = 22.0:l.0:2.7:l.0 will be compared next to determine the impact of increasing pressure. Upon comparing the gains of the two 102 torr cases, several points arise: (1) Increasing the percentage of F2 in the mixture in- creases the peak gain. As the percentage of F2 climbs from 10.1% to 16.7% the peak gain rises by about a factor of 7. Thus, in this range, peak . gain appears to be very sensitive to the percentage of F2 in the mixture. Note that this is not due to a concurrent increase in F/F2 ratio with the increase in mixture F2 percentage. As shown in Appendix A.2, F/F2 ratio is independent of initial F2 concen- tration for the conditions of this work. Increasing the percentage of F2 in the mixture decreases gain duration and decreases the time (3) 76 to peak gain and the gain initiation time. Gain duration is quite sensitive to the percentage of F2 in the mixture in this range. Increasing the percentage of F2 from 10.1% to 16.7% decreases the pulse duration by a factor of 7 for the v = 1-0 band and by a factor of 15 for the v = 2-1 band. ‘Both of these observations can be attributed to the increased rate of chemical pumping which is due to the increased F2 concen- tration. This effect has been documented by many authors (cf. Reference 70). An increase in F2 concentration causes an increase in fluorine atom production rate, as in Equation (1.4). The increase in F production rate causes a subse- quent increase in the rates of Equation (1.2), and hence, Equation (1.3). This accelerates the rate of production of HF, speeds the formation of and increases the magnitude of the population inver- sions required for lasing, and hence, decreases the gain initiation time, increases the peak gain, and decreases the gain duration. The latter occurs because the fuel and oxidizer are consumed at a faster rate. Perhaps the most interesting observation is the effect of increasing the percentage of F2 in the mixture on the shapes of the individual gain pulses. For the case using the lower concentra- 77 tion of initial mixture F2 (10.1%), the individual pulse shapes appear erratic. Upon increasing the initial mixture concentration of F2 to 16.7%, the pulse shapes become more regular, approaching a common, somewhat parabolic shape. The observed patterns for increasing peak gain, decreasing gain duration, decreasing gain initiation time and decreasing time to peak gain with increasing rotational level are more evident as the percentage of F2 in the mixture increases. It is possible that this phenomenon is due to the larger temperature rise for the 16.7% mixture F2. The larger percentage of mixture F2 would increase the chemical pumping, as mentioned above, which would in-_ crease the enthalpy production, hence increasing the temperature. Directly related to the temperature rise is a shift in the relative importance of the various relaxation mechanisms. Careful inspection of the rate coefficients for the V-R,T and R-R,T channels shows a decrease in the HF V-R,T self-relaxation rate with a temperature increase up to 1040 K and an increase in the HF R-R,T total relaxation rate with a temperature increase up to 1370 K. These temperatures are considerably higher than results of model calculations at these conditions. The model predicts final pulse temperatures ranging from 360° K to 400° K. To a good approximation, the HF V-R,T self-relaxation rate is equivalent to the V-R,T total relaxation rate. This is due to: (l) The rate constants for H, F, and HF V-R,T deactivation of HF all being of the same order of magnitude and significantly higher than the rate co- efficients due to F2, H2 and chaperone gases [85]. (2) Prereaction 78 causing the concentration of HF to be consistently two orders of magni- tude higher than either the H or F concentrations. Thus, the HF self- relaxation rate for V-R,T should dominate the total V-R,T rate and be a good approximation to the total rate for purposes of comparison. It is clear from these arguments that the relaxation contribution of R-R,T is increasing relative to the contribution due to V-R,T. This is consistent with the increase in the strength of trends in gain initiation, peak and termination times with an increase in F2 concentra- tion. This increase in R-R,T contribution to relaxation leads to the nearly thermalized distribution which leads in turn to the rigid J-shifting behavior of the gain. Since R-R,T relaxation decreases with J, this leads to an increased gain duration with an increase in rotational level. It is assumed that R-R,T relaxationtscales as exp-(AE/RT) with AE being the ro- tational level energy gap. The value of AE increases with increasing rotational level decreasing the relaxation rate. This decreases the total relaxation rate with increasing rotational level and leads to an increase in gain duration. Upon comparing SSG for the two cases with composition HezozzF : 2 H2 = 22.0:1.0:2.7:l.0, the following arise: (l) Gain duration on each transition decreases with an increase in mixture pressure. (2) Peak gain on each transition increases with an increase in mixture pressure. These observations are consistent with our understanding. The in- crease in pressure leads to increases in the pumping rate and the rates of relaxation. These increased rates cause a quicker consumption of fuel 79 and oxidizer and also a sharper termination rate. The combination of the two reduces gain duration. Since all dominant chemical and relaxa- tion kinetic reactions are binary, this is simply binary scaling. This decrease in gain duration is probably due to binary scaling. The peak gain increase with pressure is probably another manifesta- tion of binary scaling: As pressure increases, the rate of chemical pumping increases as the pressure squared. This would cause formation of larger population inversions due to an increase in total population (even if population ratios remain constant). These larger inversions would lead to larger gains. 80 2.5 Results of Total Pulse Energy Studies The experimental configuration utilized for the total pulse energy (TPE) measurements is described in Section 2.2.2 and displayed in Figure 2.11. Total pulse energy measurements were made only for the 331 torr case with mixture composition He:02:F2:H2 = 20.8:l.O:4.6:1.2. For these conditions, 175 mJ of laser energy was measured. This can be converted to an energy density by dividing by the resonator mode volume and by the mixture pressure in atmospheres. .The resonator mode volume can be approximated as the volume of a truncated right circular cone whose radius is equal to the distance at which the beam intensity is 1% of its centerline value. This is a radius equivalent to five times the inten- sity l/e point radius. .The latter can be determined using formulae found in Gross and Bott [1]. For the case considered here, flat output coupler and 5 m radius of curvature mirror with an active medium length of 53 cm, the mode volume is 75 cm3. This gives a value.of 5.4 J/l-atm as the measured energy density. The value reported here lies between the reported values of 80 J/l- atm of Chen, et al [41] and 2.9 J/l atm of Hess [11]. This is to be expected as energy density increases with the fraction of mixture H2 and F2. Hess [11] reports results for a very dilute mixture of Heze: H2 = 40:1:1 while Chen, et a1 [41] show results for a mixture of He:F2:H2 = 8:1:1. The mixture used here lies between the two, and closer to that of Hess [11]: (He + 02‘):F2:H2 = 18.2:3.7:l.0. This is as expected. CHAPTER 3 COMPUTER SIMULATION OF AN HF LASER AND COMPARISION WITH EXPERIMENT 3.1. Introduction The computer model used to simulate the HF laser is described in de- tail in References 23 and 68. A brief description is given in Appendix A. This model will be referred to as the VR model. ‘The VR model was modified initially from that of References 23 and 68 by including wavelength dependence of the output mirror reflectivity and by modifying the flash photolysis temporal profile. The wavelength dependence of the output mirror reflectivity caused a wavelength depen- dent variation of the threshold gain. The flash photolysis temporal profile was changed from a sinusoidal distribution to one resembling the measured flashlamp intensity time history. See Figure 2.16 and Appendix A, Figure A.1, for further details. In addition to the modifications stated in the preceding paragraph, an update of the model kinetic rate package was performed. The most significant changes from that of Reference 23 were: (1) Removal of the multiquanta V-R,T energy transfer channels for HF self-deactivation. This was recommended by the work of Jursich and Crim [24] and Foster and Crim [25]. The remaining single quantum V-R,T energy transfer rates for HF self-deactivation were determined by using the HF V-V self-relaxation rates of Wilkins [96] and the total HF V-V,R,T self-relaxation rate of 81 82 Foster and Crim [25]. This reflects previously accepted kinetics where V-V deactivation was asssumed to be a separate mechanism, independent of V-V,R,T. In addition to this, (2) revised hot and cold pumping rates were used reflecting the work of Heidner, et a1 [26] and Wurzberg and Houston [27]. Some additional minor changes were made including: (3) slightly different rates for recombination of F and H atoms and, (4) new rates for V-V transfer between HzanuiHF [75]. The remainder of the rate coefficients are those of Cohen [95] , excluding the endothermic cold pumping back reactions which are those of Bartoszek, et a1 [97]. A table of the rate coefficients used is given in Appendix C, Table C.l. - This is the basic rate package and excursions were made from it. Model runs were attempted using the VR model modified as stated above. It was discovered that in order for a run to integrate to com- pletion, an inordinately small step size had to be chosen. This step size would have caused excessive CPU time usage had the case been run to completion. A modification was performed in an attempt to solve this problem. The most likely cause of the stepsize problem was "stiffness" in the system of differential equations, due in this case to the choice of input conditions necessary to simulate the experiments [98]. More- over, the model appeared to be spending excessive time calling and using the derivative computation subroutines. For this reason, an ap- proximation was inserted to allow the derivative computation subroutines to be called only 10% as often. It was hoped that this would speed up execution, reducing the amount of CPU time required for a complete run. The result was that the model refused to integrate. 83 At this point, a simplified version of the VR model, denoted the VT model, was run in an attempt to simulate the experiments. The VT model is similar to the VR model except it neglects the V-R portion of the V-R,T energy transfer mechanism approximating it as a completely V-T mechanism. The V-T model also approximates all rotational relaxation as R-T relaxation, instead of R-R,T [99]. The rates used in the VT model are also given in Appendix C, Table 0.2.. Two final modifications were implemented on the VR model. First, the subroutine computing the population derivatives due to V-R,T energy transfer was rewritten to take advantage of the removal of the multi- quanta V-R,T mechanism. This had no noticeable effect on the required CPU time. Second, the number of rotational levels considered per vibraé tional band was reduced from 30 to 20. This was justified by the prior removal of the multiquanta V-R,T relaxation mechanism since this me- chanism is the only one that populates rotational levels above J = 20. The model with no multiquanta V-R,T deactivation and only twenty rotational levels was denoted VR20J. The inclusion of these last two . modifications reduced CPU time necessary for a complete run by about a factor of two. Hence, in this study, two separate models were used, denoted as the VT and the VR20J models. In all computer modeling studies, it is necessary to determine the model input conditions. For this work, that was done in the following manner. An area of uncertainty in most modeling studies is an uncertainty in the rate of fluorine atom production due to photolysis. Several researchers have undertaken studies to determine the efficiency of various photolysis sources [100] and the importance of various lamp 84 characteristics in photolyzing molecular fluorine [36]. Various types of sources have been compared in detail by Berry [101]. All have reached the conclusion that it is important to be able to quantify the fluorine atom production rate. Unfortunately, there appears to be no accurate, nonintrusive experimental technique available to determine fluorine atom concentrations. Furthermore, the techniques used to infer fluorine atom production rates, from fluorine molecule disappearance rates, are inac- curate. Thus, there is no viable experimental method to determine this important laser characteristic. In this work, the fluorine atom production rate due to photolysis was estimated using the VT computer model with the original rate package in the following way. The parameters 111p and la in Equation (A.26) were varied, systematically changing the rate of fluorine atom production. Values of 11p and la were chosen which provided best fit agreement between model predictions of SSG and TRS and those recorded experimentally. This method is in contrast to other workers who have employed a variety of techniques to measure fluorine atom production rates. Perhaps the most common experimental method is actinometric mea- surements on F2. Greiner [102] has used thermal actinometry and Suchard and Sutton [103] have used laser actinometry to attempt to measure the rate of disappearance of F2. These techniques suffer in accuracy be- cause of the small amount of F produced, less than 0.5%. One is thus attempting to determine a small number by differencing two large numbers, where errors in the large numbers are comparable to the result desired. The laser actinometry technqiue also suffers because the F2 absorption coefficient is strongly temperature dependent. Under normal laser 85 conditions, this temperature dependence will have a stronger effect on the absorption signal than the disappearance of 0.5% of'F2 [104]. Other workers have attempted to use multidimensional radiative transfer computer codes [76,1(H[]to estimate the fluorine atom produc- tion rate. Adjustable parameters in these codes are based on experi- mental measurements similar to those of[103] and hence experience the same problems. It is because of the uncertainties and inherent errors in the above methods that the fluorine atom production rate was estimated in the manner stated herein. The model required several input parameters in addition to know- ledge of the F/F2 ratio produced through photolysis. A description of how these were determined is presented below. The gas inlet composition was determined from measurements of the cavity partial pressures of each of the constituents prior to a run. This gave pressures for He, H2, F2 and 02. The value for SF6 was de- termined from a knowledge of the F2 supply batch analysis, provided by the gas supply company (Matheson), and the cavity initial F2 pressure. For example, the inlet F2 pressure for the 36 torr case of composition He:02:F2:H2 = 20.8:l.O:4.6:1.2 was measured to be 6.0 torr. From the batch analysis, presented in Section 2.2.1, the SF6 concentration is 0.001 molar percent yielding 0.00006 torr SF6, initially. The initial concentrations of air, CF4 and CO2 were determined in the same manner. These constituents were assumed to be inert and were, hence, added to the initial He pressure. This was necessary as the model does not include kinetics for the species CF4, C02 and air. In addition, the 86 partial pressure of 02 was also added to the He pressure. (Again, as for CF4, C02 and air, there are no kinetics for O2 in the model so it was accounted for by treating it as an inert species. The cavity active medium length was determined to be the length of active medium illuminated by the flashlamps. This was the lamp length (56 cm) minus the length blocked by the aluminum brackets securing the quartz windows to the cavity (3 cm). See Figure 2.2. The result was an active medium length of 53 cm. The mirror spacing was measured to be 121 cm, and the initial mixture temperature was assumed to be 300° K. The output coupler reflectivity was measured at the AFWL metrology lab. It was nominally 81%. A curve of reflectivity vs wavelength is presented in Figure A.3 The copper mirror reflectivity was stated by the manufacturer to be 99%. This value was multiplied by the measured Brewster window trans- mission values of 99% each, yielding an effective mirror reflectivity of 95%. The effective reflectivity was used as an input to the model. .The results of the measurements of flashlamp emission properties were used to determine the flash photolysis temporal profile. This procedure is described in detail in Appendix A. The resulting fit to the flashlamp intensity time history was used as a model input. After having determined the input parameters, the VT model was run for the conditions He:02:F2:H2 = 20.8:l.O:4.6:1.2 at 36 torr, 102 torr, and 331 torr simulating the three TRS cases presented. The results of this set of model runs is presented in Figure 3.1. 87 The model rate package was then modified by changing the vibrational pumping diStribution suggested by Cohen [95] to an updated distribution, also suggested by Cohen [105]. See Table c.2, Appendix c. The original hot reaction vibrational pumping distribution [95] increased monotonically from v=3 to v=6. Pumping to all other vibrational levels was assumed to be zero. The updated vibrational pumping distribution [105] also increased monotonically form v = 3 to v = 6. In the updated case though, the pumping distribution was assumed to be monotonically de- creasing from its peak at v = 6 to v = 8. Pumping was again assumed to be zero for all other levels. Cohen [105] also suggested leaving the total pumping rate summed over all vibrational levels unchanged from Reference.85. The model was again run at 36 torr, 102 torr and 331 torr for the mixture composition He:02:F2:H2 = 20.8:l.O:4.6:l.2. The results of this set of model runs are presented in Figure 3.2 and compared with experiment. The model rate package was modified one final time. Here, the vibrational deactivation mechanism was changed to reflect the current belief that the total vibrational deactivation rate is a sum of the V-R,T rate and the V-V rate. The V-R,T rate was then determined by subtracting the V-V contribution, reported by Wilkins [96], from the total vibrational deactivation rate of Foster and Crim [25]. This, in effect, reduced the V-R,T rate from 1.1 x 1010 e103o/RT TO‘5 to 3.3 x 109 e1030/RT T0°5. For this rate package, the model was run only at 102 torr. The results are presented in Figure 3.3 The rate package mentioned directly above was then used with model VRZOJ to simulate the 102 torr case for mixture composition He:02:F2: H2 = 20.8:l.O:4.6:l.2. The results are presented in Figure 3.3 88 All model results will be discussed in the following two sections. 3.2 Computer Modeling Results of Time Resolved Spectrosc0py and Comparison with Experiment 3.2.1 Introduction The results of the VT model TRS computer simulations using the initial rate package are presented in Figure 3.1 in Section 3.2.2. The results of the VT model TRS computer simulations using the modified vibrational pumping distribution are presented in Figure 3.2 in Section 3.2.3. The results of the VT model TRS computer simulations using the modified vibrational pumping distribution and the modified V-T deactivation rates are presented in Figure 3.3 in Section 3.2.4. The results of the VRZOJ model TRS computer simulations using the modified pumping distribution and the modified vibrational deactivation rates are also presented in Figure 3.3 and are discussed in Section 3.2.5. All results are plotted identically to those of Sections 2.3.2 and 2.3.3. Experimental results are repeated with the presentation of model results to facilitate comparison. They will be discussed individually below. 3.2.2 Comparison of VT Modeling Results of Time Resolved Spectroscopy with Experiment: Initial Rate Package A comparison of the model predictions for this composition with experiment (see Figure 3.1) shows good agreement for transitions pumped strongly by the cold pumping reaction, Equation (1.2) and only fair agreement for transitions pumped by the hot pumping reaction, O-JNw-D< U'IU'IU‘ILNU'IC.‘ O—le-hU'IU‘I< mmmmmmc. UTU'IU'IUTU‘IU'IL “I‘ll 89 1,..— F 1'” 0 so 100 160 F...- ’1‘ 'fl: 0 25 50 75 160 (C)' 10 40 C 20 30 Time (nsec) Figure 3.1 50 oamwem< o—ern-m< o~mw4>m< J: 51. 51.. (d) 5'.r 5 _=.—_ 5..—~ 5 ,..-:-—-" go 56 150 130 200 5 up 5 dI-—-- (e) 5 :i" 5 pF--—' 5.p---- SifF-F'-’__- Jo 20 4o 60 30 100 5 ul' 5 el""" (f) s eF" 5 u-- 5‘"? 5...!— 0 1o 20 30 40 5b Time (usec) Time resolved spectral output: comparison of VT model results and experiment using rate package VT (standard rates) (a) experimental results, 36 torr (b) experimental results, 102 torr (c) experimental results, 331 torr ( ( ( e) model results, 102 torr d) model results, 36 torr f) model results, 331 torr 90 Equation (1.3). It is apparent that the model results agree well with the general trends observed experimentally regarding pulse initiation, termination and peak intensity times, peak intensities and pulse durations. Further- more, agreement between model and experimental magnitudes is generally good for the v = 1-0 band and very good for the v = 2-1 band. For these two bands, the model underpredicts pulse durations, giving in- tiation times which are late and termination times which are early. The area of least agreement for the v = 1-0 band is in the behavior of P1(5). Experimentally, P](5) is a short duration transition with weak in- tensity. It does not obey the pulse initiation time, pulse duration or peak intensity trends. In Section 2.3.2, it was stated that this be- havior was due to water vapor absorption of the P](5) signal. The model results support this contention. According to the model, P1(5) should obey the observed trends for pulse initiation time, pulse duration and peak intensity. That this is not observed experimentally leads to the conclusion that the anamolous P](5) behavior is due to some effect other than laser cavity chemistry. This effect is probably due to water vapor absorption and could be investigated by including oxygen kinetics in the model. As noted in the preceding paragraphs, agreement between the model and experimental results is generally very good for the lower two bands. This is not true for the higher bands. It is interesting to note that the VT model overpredicts the number of lasing transitions observed. Some of these predicted transitions are weak and may well have intensities below the experimental detector sensitivity limit. As a further note, the model agrees with experiment by predicting the 91 absence of lasing on pure rotational lasing transitions. It is noteworthy that the entire VT model results for the v = 1-0 band at 102 torr and all bands at 331 torr appear to be shifted upward one rotational level: i.e. the observed behavior of P](4) in the model corresponds to the experimental P](3), etc. There is no apparent explanation for this behavior. 3.2.3 Comparison of VT Modeling Results for Time Resolved Spectroscopy and Comparison with Experiment: Modified Vibrational Pumping Distribution A comparison of the model predictions with experiment in Figure 3.2 shows excellent agreement for the transitions in the v = 2-1 and v = 5-4 bands. In general, one observed the same qualitative features for the remaining bands. However, the model does not quantitatively agree with the measured spectra. This is more evident for the v = 3-2, v = 1-0 and v = 6-5 bands. It is apparent that the model results agree very well with the general trends observed eXperimentally regarding pulse initiation, ter- mination and peak intensity times and pulse durations. This holds for both hot band and cold band lasing. Furthermore, agreement between model and experimental magnitudes agree very well for all bands except v = 3-2 and v = 6-5 in general. The model accurately predicts initiation and termination times, as well as pulse durations for the v = 2-1, v = 4-3 and v = 5-4 bands. In addition, the model predicts lasing on all the transitions that are observed to lase experimentally O-—‘NL~>-§U‘l4.39 (A31) 1(A) = -395.92 + 0.48671A - 2.2172 x 10‘4A2 + 4.4511 x 10'3A3 - 3.3262 x 10‘12A4 (A32) NORMAUZED LAW INTENSITY (orbitrwy ml”) 134 1.0 - 5— FITTED A - - MEASURED 0.5-J \ ng ' o o 2 o 4.0 e o 30 ‘nME Una» Figure A.1 Total flashlamp intensity vs time 135 A.3 Threshold Gain The model of References 23 and 69 was modified to include a wave- length dependent threshold gain. Variations in output coupler reflec- tivity with wavelength made this modification necessary for accurate simulation of experimental results. The reflectivity of the 5m radius of curvature copper mirror was assumed to be 99% for the region 2.5 microns to 4.0 microns. An analy- tical expression for output coupler reflectivity vs wavelength was least squares fit to the measured reflectivity vs wavelength curves of Figures (A2) and (A3). The form of the reflectivity was found to be R97( A) = -22.794 + 0.020169A - 5.7187 x 10%2 + 5.4066 x 10'10A3 (A33) for the 97% maximum reflectivity output coupler and R81”) = -0.98008 + 0.0012030A - 2.4912 x 10‘7A2 + 1.4827 x 10"11 A3 ' (A34) for the 81% maximum reflectivity output coupler. Threshold gain as a function of wave number was computed using Equation (A17) with Equation (A30) substituted. athr is evaluated for each lasing transition. Renecnvnv [x] 136 100:) ”WEED 0 FITTED 25.‘ O. r r . 2500. 3000. 3500- 4000. waveuuuaen [6.55"] Figure A.2 Reflectivity vs wavenumber for the output coupler with nominal 97% reflectivity REFLECTIVITY (%) 137 ‘00." 78.“ I MEASURED O FITTED '— “-1 25.‘ o. r I 2500. 3000. 3500 waveuuuun (”I") Figure A.3 Reflectivity vs wavenumber for the output coupler with nominal 81% reflectivity APPENDIX B DETERMINATION OF INITIAL HF CONCENTRATION DUE TO PREREACTION APPENDIX B DETERMINATION OF INITIAL HF CONCENTRATION DUE TO PREREACTION The concentration of HF due to prereaction was determined in the following manner. The laser was set up in the small signal gain diag- nostic configuration. See Section 2.2.2 and Figure 2.10 for further details. The cavity was first evacuated, then filled with a 36 torr mixture of He:02:F2:H2:H2 = 22.0:l.0:2.7:l.0. The P](3) absorption, or negative gain, was measured as 0.0035 per cm. A computer simulation was run using the above mixture composition with l mtorr HF added, all at 300 K. The P1(3) absorption was computed by the model to be 0.00404 per cm. A second computer simulation was run producing a value-of 0.00364 per cm for the above mixture with 0.9 mtorr of HF added, again all at 300 K. The HF prereaction concentration was thus estimated at 0.9 mtorr for this case. 'Prereaction for the other cases was then scaled linearly with mixture F2 pressure, using the value obtained for the 36 torr case as a baseline. This gain technique has the sensitivity to measure even lower con- centrations of HF. To see this, we write for the P1(3) transition: dI/dx = 01. (Bl) Assuming constant gain over the entire absorption path, ln(I1/12) = 0L (82) 12 138 139 Solving for a, a = [ln(12/I])/L12]. (83) In these tests, we were limited to 12/11<0.9. IZ/Il = 0.9 was the largest ratio of intensities that could be accurately resolved from the photographs. For this work, L12 was 100 cm. This yielded a minimum detectable gain of approximately 0.001 per cm. Using the model again, as above, it was determined that this corresponded to an HF concentra- tion of about 0.3 mtorr. This would be the minimum detectable concen- tration of HF. APPENDIX C MODEL RATE COEFFICIENTS APPENDIX C MODEL RATE COEFFICIENTS The rate coefficients used in the VR computer model are listed in Table C.1. The rate coefficients used in the VT model.are listed in Table 0.2. These rate coefficients are listed by rate package: VT, VT2 and VT3. 140 141 Table 6.1 Current Rate Coefficients in H2 0 F2 Systems. Reaction Rate coefficient Number Units of cm. mole. sec, cal H, v. A, 9(v) la 02(0) 4 H2 - 2H + H2 k-‘. . 6.2 x 1017FT'02557 H2 - all species except H + H2 _ 16 -0.61 lb "2(0) + H2 - 2H + H2 K_‘b 9.4 x 10 1 lc H2(0) + H - 23 4 H (.1: - 1.2 x 10" 1°°5 _ . . 13 A-10. -2.7. -2. 2 52+“: 215+"3 1:; 5.01110 5 ‘52 ‘11- exp(-35, 100/111) x IO.(EO-EV)AI 113 - 1, .11 other: 3 HF(v)+H-H+F+ ()-.‘°2~x10‘91“ 11- - -5 -1 11 3 H3 K3 ' n + 1 F AH “HF ‘ Ab ' a exp(-135, TOO/RT) others: v t 0.....n 4 F 4 112(0) - 11m) + 11 11, - 9(v) x 4.9 1:10‘1 To‘8 v -1, 2, 3; 9(1) - 0.17, exp(-600/RT) 9(2) - 0.55. 9(3) - 0.28: 9(') ' oe V ’3 . II. 0.5 40-1 HF(4) + H - "2" ) + F K.‘ - 6.0 x 10 1 (v - 4) 46.2 111515) '+ 11 - 1121M + r x_‘( 5 - 8.8 x 10" 1°-5 . v . 46-3 HF(S) + 11 - 112w) . s 114 - 1.6 x 10'2 7°" Iv - 6) _ 09 1.3 5 H + F2 - HF(v) e F Kg - 9(v) x 3.0 x I T g(v) - 0, v I 0, I. 2. exol-QSOIRT) 9(3) - 0.08. 9(4) - 0.13. 9(5) ' 0.35. 9(6) - 0.44; 9") ' 09 V ’ 6 6 HF(v) + HF - HF(v') 6 HF K6. - 1.1 x 1o‘° v3°5 1°-5 v' - v.1 exp(lO30/RT) 6b HF(v) + 11 - 116011) + 11 K - 6.0 x107 1! 1 v' - Vol 2 2 va 6c HF(V) + H - HF(v'l + H K5 ( ) - QIV.V') x 1.5 x 1012 9(I.0) - I. 9(Z.I) ' 9(2.01 - 1.8. c v,v' exp(-700/RT) 9(v.v-I) - 360. v - 3. .... 6. 9(V.v-nl - 1.8. v - 3. .... 6. n - 2...! 6d 11m) . r - 116w) + 15 116‘ - 1.6 x 1013' (21007111 w - v-l V 6e HF(v) + n - HF(V’) + n K6 - 7.7 x 10‘7 v 15 A" AF - AA, - 1.0. A". - 2.0 ev 2 1112 Table C.l (Continued) Reaction Rate coefficient Number Units of on. mole. sec. cal H. v, A. 9(0) 7 HF(v) I HF(V') I HF(VII) K7(v, I; v+l.0) I 3.6 x 10‘5 v I l, v' I l I hF(v‘-l) y.‘ 7‘1-0 K7(v, v'. v+l, v'ol) I 3 x 1015 v I 2...... 7. v' I 1....6 v-I T"'° 8a 112M + "10 - 11261-1) + x3. - 2.5 x 10" v 1‘-3 A!” A": - 4. ‘10 - 1 .11 others H10 , v I l. 2 8b 112m . 11 - 112(1)-” . 11 ‘86 - z a 10‘3 exp(-2720.RT) v - 1, 2 9 HF(v-n) I H2(v'+n) I HF(v) K9 8 x IOn v v I l...6: n I l..... v 10. 11610.10) + HF(v,J) - km - 1.023 x 10“ 74-8“ ur(o.10-AJ) o HF(v,J) 6'2569’“T 106 ”0.10) . 11m.» - “11:1 - 3.36 x 10“ 1"“‘93 HF(l,lO-aJ) + HF(v,J) .o2436/RT *The rate coefficients are taken primarily iron Cohen [95] except where noted in the test. +£quation (6a) represents the total V-R-l rate for level v for collisions with HF. The sun of this rate and that of reaction (9) equals the rate suggested by Crie for tool deactivation of 11F(v). 143 Table C.2 VT Model Rate Coefficients for the H2 + F2 Chemical Laser rate package VT identical to VR rate package with the following exception: Equation (6a) and Equation (9) are summed to give one V-T rate: a = 1.1 x 10 T exp(1030/RT) 6 rate package V12 identical to VR rate package with the following exceptions: Equation (6a) and Equation (9) are summed to give one V-T rate: (68') K6a = 1.1 x 1010 v3°5 T°°5 exp(1030/RT) Equation (5) has a revised pumping distribution: 9 1.3 (5') K = g(v) x 3.0 x 10 T g(v) = 0, v = 0, l, 2; 5 9(3) = 0.07, g(4) = 0.13, g(5) = 0.22, g(6) = 0.32, 9(7) = 0.14, 9(8) = 0.12 rate package VT3 identical to VR rate package with the following exceptions: Equations (6a) and Equation (9) are summed to give one V-T rate: (68") K6a = 3.3 x 109 v3'5 10:5 exp(1030/RT) Equation (5) has a revised pumping distribution: (5') K5 = g(v) x 3.0 x l09 T1'3 g(v = 0, v = 0, l, 2; g(3 = 0.07, 9(4) = 0.13, g(5) = 0.22, g(6 = 0.32 g(7) = 0.14, 9(8) = 0.12 APPENDIX D ERROR ANALYSIS FOR PROBING SMALL SIGNAL GAIN OFF LINE CENTER APPENDIX D ERROR ANALYSIS FOR PROBING SMALL SIGNAL GAIN OFF LINE CENTER In Section 3.3.2, it was stated that up to 26% error could be introduced by probing SSG off-line center with the system utilized here. This will be demonstrated below. Yariv [116] gives the following expression for a pressure (Lorentz) broadened line: 1(6) = (AvL/n)/[(v-v.)2 + (A.L/2)21 (01) where AvL is the Lorentz full width at half maximum (FWHM) for the transition in question. AvL is related to the binary collision frequency for the emitting molecule by: AvL = Z/n (02) _ 2 Z - 26 NA/2nkBTE1/MHF + l/Mi] (03) where 02 is the collision cross section between the emitting molecule and the ith collision partner, NA is Avogadro's number, k8 is Boltzmann's constant, T is the temperature and MHF and ”1 are the molecular masses th of HF and the i collision partner respectively. For the 102 torr mixture 2 = 3.35 x 108 Hz. To determine the maximum error due to probing SSG off-line center, it is first necessary to determine the maximum probe laser frequency deviation from line center and then determine the effect of this devia- tion on the measured SSG. 144 145 The maximum probe laser frequency deviation from line center must be equal to one half of the probe laser cavity longitudinal mode spacing. Yariv [116] gives an expression for the cavity longitudinal mode spacing: Avc = C/ZL (D4) with c being the speed of light and L being the cavity mirror spacing. For the Helios probe lase used here, Avc = 1 x 108 Hz. The effect of this frequency shift on the gain is directly related to its effect on the line shape, 0(0). To determine the maximum error, it is thus sufficient to determine the difference between the line center value of the line shape function and that at the maximum fre- quency deviation. Substituting from Equation (01) yields: max error = l - (Av /2)2/[(v-v )2 + (Av /2)2] (05) L ° L Inserting the values for AvL and Avc calculated above gives a value for the maximum error of 26.6%. Here, (l/2)Avc = v-vo. This is the number quoted in Section 3.2.2. APPENDIX E RAN DATA FOR TIME RESOLVED SPECTROSCOPY PLOTS USED IN FIGURES 2.17 AND 2.21 APPENDIX E RAN DATA FOR TIME RESOLVED SPECTROSCOPY PLOTS USED IN FIGURES 2.17 AND 2.21 The raw data used to plot Figures 2.17 and 2.21 are presented in Tables E.l and E.2 respectively. 146 Table E.l transition tl 36‘Sotr gseeen:e 1 p p c P1(3) 5.6 5.6 3200 6.6 P1(6) 6.0 11.0 900 29.0 21(5) 26.0 30.0 60 37.0 P1(6) 25.0 36.0 330 73.0 P1(7) 76.0 86.0 160 96.0 61(8) ... -- ... -- 22(3) 2.6 2.5 1600 6.5 22(6) 5.0 11.0 200 16.0 22(5) 6.0 15.0 610 55.0 22(6) 16.0 69.0 320 135.0 P2(7) 60.0 125.0 160 160.0 Pz(8) ... ... ... ... 93(3) 3.5 3.7 50 6.9 P3(6) 3.0 3.2 860 26.0 23(5) 6.0 27.0 610 66.0 P3i6) 25.0 60.0 360 170.0 P,(7) 165.0 170.0 60 210.0 2‘13) ... .. ... .. r‘(6) 9.0 26.0 1000 35.0 P‘(5) 10.0 106.0 320 116.0 P‘(6) 112.0 116.0 6 116.0 95(3) 6.0 7.0 520 7.0 P,(6) 6.0 65.0 260 65.0 65(5) 16.0 50.0 70 50.0 95(6) ... ... ... ... 96(3) '-—- -- -- -- P6(6) -- -- ... ... P6(5) -- -—- -- -- 1 16121.6166 :1.— (ueec) 2 Peak Intenetty tine (ueec) 3 Peak Intensity (relative unite) “ ternination t1nn (none) 5 tranettton duration (unec) Data for :3 t1 1.2 2.2 25.0 2.6 9.0 2.0 68.0 6.0 20.0 15.0 6.1 1.7 11.0 2.1 51.0 2.2 127.0 6.0 120.0 17.0 -—— 68.0 1.6 2.3 23.0 2.0 78.0 2.5 165.0 8.0 65.0 55.0 26.0' 6.0 1066 L0 6.0 69.0 1.0 3.6 3mo 10 36.0 6.0 -— zno —- L2 -—- 7.0 147 TRS presented in Figure 2.17 102 tort preneure t P 2.6 2.5 6.0 6.5 17.0 1.8 2.9 6.0 11.0 65.0 76.0 2.6 2.2 3.5 62.0 68.0 6.2 10.0 52.0 3.6 6.0 16.0 32.0 7.9 9.0 I P 5000 6000 80 1250 900 3200 1600 1700 1800 1300 560 70 2600 $1355 1200 §§§ a 1 110 200 18 t c 2.9 6.5 28.0 18.0 25.0 3.3 10.6 18.0 67.0 81.0 90.0 3.2 7.5 26.5 73.0 95.0 9.0 51.0 60.0 3.9 19.0 38.0 38.0 9.6 12.0 0.7 6.1 26.0 12.0 10.0 1.6 8.3 15.8 61.0 66.0 22.0 0.9 5.5 26.0 65.0 60.0 5.0 67.0 11.0 0.5 16.0 36.0 10.0 2.6 5.0 6.0 2.5 2.1 2.6 5.7 11.0 1.5 1.7 1.6 2.7 5.0 1.6 2.0 2.1 3.0 12.0 2.0 2.5 3.0 6.0 2.3 3.6 6.2 8.0 6.0 6.1 .3.0 331 tee: peenenre t P 6.1 2.8 2.3 3.9 6.0 12.3 1.6 2.1 2.2 6.8 11.0 1.8 3.0 3.6 9.0 20.5 2.5 2.8 7.0 15.0 2.6 6.2 6.6 12.0 6.1 6.5 6.0 1 P 150 1000 3600 3800 10000 3800 2600 12100 10600 11000 6100 9000 11000 5700 25 3355335 2100 1100 E C 6.5 7.0 3.7 7.2 16.3 18.8 2.2 6.0 12.5 20.0 28.0 2.5 6.7 18.8 25.0 60.0 3.5 5.5 16.0 27.0 3.0 8.0 19.0 21.5 8.0 7.2 10.0 d 0.5 6.5 1.6 6.8 8.6 7.8 0.7 2.3 10.9 17.3 23.0 0.9 2.7 16.7 22.0 28.0 1.5 3.0 13.0 21.0 0.7 6.6 16.8 13.5 6.0 3.1 7.0 1418 0.«« «.0« u.« 0.«0 «.0« «.0 0.00 «.«0 0.0« «.0 n.n 0.«0 «.0n «.0« n.« 0.a 0.0 0.n~ 0.0 n.~n 0.«« «.« 0.00 «.0u 0.0 0.««« «.«« «.0« «.0 n.« «.n« 0.00 0.0« 0.0 n.0 0.0« «.n« 0.0 0 u o«« 000A 00 000A °o«« oo~« 000‘ o«« 00« 000u 00« OO«« :00 000 °0«« & u «.«« 0.0 «.0 0.«u 0.0 0.0 0.«« 0.0 o.n 0.0 c.«n 0.0a 0.0 «.0 0.« 0.0« 0.0« 0.0 m u nun-nun: when «an n.0u 0.0 «.0 0.0 0.0 «.0 0.00 n.0u «.0 0.0 0.0 «.00 0.0 0.0 0.0 «.« «.0u n.0u 0.0 «.n« n.«n 0.« 0.«0 0.« «.~ 0.«0 0.00 n.n« 0.0 0.0 0.00 0.0« n.« 0.« 0.« 0.«~ 0.0a n.n 0 0.00 °.«0 0.0 o.«« 0.00 0.0 °.n«~ o.«0~ n.«« 0.« «.0 °.no« 0.00 n.«« 0.0 0.0 0.00 «.a« 0.- U u 00« 000« on. coon ocun 00¢ OOnN 0°0« 000 cod 000« 0004 000« 0«0 oo«« o«« oo« coca a u n.«« 0.0 0.0 0.0 0.0 0.0 n.«« 0.0 «.0 0.0 0.«n 0.«~ «.0 0.0 0.0 0.0« 0.0 «.0 5 u enanneus hue» «an n.«« 0.0 .0.0 0.0 «.n «.0 0.00 0.n~ 0.0 0.0 0.0 o.n« 0.0 .0.0 0.n «.n n.0« 0.0 «.0 «a 0.00 o.«n 0.« 0.0« «.AA n.« 0.00u n.00« 0.00 0.0 «.q n.nn« 0.00m o.«« 0.« 0.« n.0m 0.0« «.0 0.« v 0.00 n.«0 O.« o.«0 n.«« 0.« o.n«« 0.0«« 0.00 0.0 0.0 0.00u 0.n«« 0.n« 0.0 0.0 n.«n 0.«« n.- 0.0 U u and Gama °n«« 0«~ 00« 0004 0000 can« o«« 000« 00«« 000« 0000 oo«« 000 000 000« « a u o.«n 0.« 0.0 O.«« 0.0 0.0 0.000 «.00 0.0« 0.0 0.0 n.«o« 0.«« «.0 0.0 «.n 0.0« «.o 0.0 «.0 a u ens-noun anon «0“ 0.0a .0.0 «.n 0.0 0.0 0.0 0.0«u «.n« 0.0 0.0 0.0 0.00 «.0 0.0 0.0 «.« 0.~« 0.0 0.0 0.n «a FN.N mczawu cw nmpcwmmcq wm« com mama N.“ mpnmh Aoeeav noun-90 nouuuennua m Ace-:0 enuu nauuenwluuh 5 «noun: e>uun~euv enuu «aw-nounu seen m .6..;v .... 5.3.6.663 anon N Ace-:0 enau nouueuuucm a 0.00 o.«0~ 00« 0.0 n.«an 0.00 0.000 «.«00 0.00 0.« 0.«00 n.n0~ n.«« «.« 0.«~ 0.00 an 0.«« n.«« and «.0n 0.000 and 0.00 «.00 00« 0.0a 0.0«0 n0 0.0«n 0.000 OON 0.00 0.00 111 0.0« n.« «0 0.0 0.000 and 0.00 0.n«« n0« n.«« n.~n med «.0 n.0n 111 0.0 0.«0 111 0.00 0.«0 004 0.0« on n an uuaeeemu uuom on o.«« 0.0 «.«« n.0u --' 0.00 0.0 0.0 0.0 0.00 0.0 0.0 0.0 Amen. 1.3.. 1.... ....a .6031 A”... .136. onnm Anvnn A.3". Anvn. A~v~a .ov- ...u. ..v- Ame“. A1031 .ovun Aevua A”... nouuuecnuh APPENDIX F TRS AND SSG PHOTOGRAPHES OF OSCILLOSCOPE DATA 149 Table F.l (Continued) r -1 '1'~1'k1' '1 . 1, 50011111920):- 4129:. 1 1 9:02:60} 1 11.5 . 3.6%.. £5. . . at. 17, . 1.1111 1.. 151 I"), 500111: .1 50067: P26264129 ‘1 Table F.l (Continued) 1.2, 36 torr :1.0:4.6: 20.8 02 Table E.2 TRS photographes, He ~....-_..‘.4 —. ...... .. -0... ..-- q. -... -:-u -. 153 154 S 4..ch 17.65.11! .1. «II. {116161.111 Am 6 so .85. A umscwucouv 155 Table F.2 (Continued) 6.5. f... .15....» .7, 4,0007 if 3V- 4/20” x. 1 w 13. ”“59; 4720.. 20,133 ' r , ...» h 1.1 '; 1 . 1 ‘v’j’ISCflil/j; I 3 V m m. . 3% LLOH New N F 0 v o._”m om No w: .mmsaocmouocn mm« m m m—ao« 158 159 A 3:52.33 mg m 32. 160 30:50:08 ”2.1023 161 x , m I. u. - 16%)},6-‘6 ....,.§. ..1.» . ...llIlii €255.88 m. “_ m3”... Table F.3 (Continued) 7 I 1, ., 4...... I. I @6214!” 164 . .3 . T L r .. . «. ... ,6 . Table F.3 (Continued) Lnou «mm .N _ 0 w o F m.omnN=" N . N one: .mocamgmouoza mm« w 111.11%).1 a. a wpnm« 3 1.1. 1.. . 165 166 .‘1 ...»... .¢—.«w A W H . . New? , gm .‘wwVv‘wNvf, 167 .2 am 168 A 333:8 . ah 169 i L; 1 1 . _ 1L“... 1 4. 1 :3 H 170 .m « if- .... -..—— .r a 6 fl. W «1.x, A. A 171 3 ..u F. 1.1:.7klt tun. Table F.5 SSG photographes, He:02:F2:H2=22.0:1.0:2.7:1.0, 102 torr‘ 'x i‘ 19/19. 91$? ;. 20., 20,3 Table F.5 (Continued) 1' we". ¢>2¢§V.-;_.9./1Qf_1<4.> ms - A”? -.W: .‘ vzmv ’ ”20.1!~ L9/gqp1gow - zaps magi/m Plioggae: . . l 3 . . i s s . 3 . 5 ' l- ; - + g . f ‘+4—-——H~o—‘—J—+-+-o—o~——v—o—v—~vfi‘p o 1‘ E - ' I‘ r. l . ? .‘é _ ; E 0611 H ,. +++ 174 Table F.5 (Continued) .. 051W“ 9/109162'20125 +i.9/1_°;51:6_;.s+:2o ' moo-v ”meant/“9119.935? . 50px 175 .L . .. ...H...._..m... 3% 3.. >82» ,, .v . ........... 38 9? .3. .33 :A. * _. 33 : em::_ucouv,w.u mFth 176 .heu". . ...‘I‘I ‘ . 30m 3qu OS :3 2.87%qu SQ.” mwlmfi>8~Q¥ Avugficoi m .m. m 33 Table F.5 (Continued) Md. 122(7):? 1001;33ng Table F.5 (Continued) 178 ‘vzo-v +>20dv___._ A f 02m.- »2oay _ 3719. 92¢; _ f'IOOpS . . 3mzmwmfimmfi-.im ...- 353 .9, N LLOH me .o.—K.Nuo._uo.NNnN:uNn_H cum: .mmcamgmouosn wmm m.m m—nmp ‘ . in 711611- . a. e 3.. 31mm: llrr 180 v« . To r171; . L l.‘ 1 9 335208 9% 2E: 181 _ , ....r.:........»~ » . . :2 m..- Sosa—HEB o. .,_ wES 182 own \5 5 ...: 5883 Who. car. C .. Anuscnmcoov o. ... wEE 183 Hflfifigfififili Hgfinwfimméahlg .u . .. . .. nu; b .. . . . ..a . , .... r 582588 a; £93 Table F.6 (Continued) l84 ._.. . '—“‘j_l— 9297‘. ”go-v 19/09 92(6); 10115. i . t F 1; ) Table F.6 (Continued) 185 _ -..... ‘ mm” Table F.6 (Continued) 5, ._.... _....._...-__.__._,. ._ -... n'rflKDN: Juflmmw”... "28'."- 020-11 9m 92a» . zaps} +4 11:1 '11 Lgou om .o._n~.Nuo ...5 5.5%.. 5 A . . 5 m 5 .N 06: . mmnqemouo ea map 5 u mpnmh 187 188 . 17.1.3.3... 1.533.1f1. Table F.7 (Continued) Table F.7 (Continued) 189 1 NH -HH HH 1 l .114”— C. ...- 1 5W .1 5V. ”21’3“”. .10/’55?" v - --.\—-r~.; r" w 5 ,3. m: i x a :n' Mums maker?- ”Tu? TM, ox .'.-‘;‘ L}....nl3'.1.9 . ._e. . 3:35.: . . 1...... L. . ‘ ska-tut ...L... V (be: .. . 3. . L. a. fifiwflnfifiufiufiu , maxi. ...s..f.o. 5.1111 )1 ... .L Table F.7 (Continued) \m JD" “ . 24131.52? 1 . .2115. " '1 1 z. -. 1 511%"? ‘7‘. 5V1“ T 2r Ii . . r L ‘ I , 1 ‘ ----' g . . ‘ ‘ 192 Table F.7 (Continued) __ 5 . L5.5. 5 555;:555 55.: Table F.7 (Continued) 194 5.555.555 555. ......thlrmiri . .. . . .. . . 5.55m”... _. .. ......555 $2 5555. .11Lr 1111. 1. 5825.55.85. ..5. 5 555.5555 195 Table F.7 (Continued) 96 1 5.2.4.5957: c.5884. . .. 5% .. i . .l 51} r . 555.555.5555. .5 rutLJIT 197 . v1.11 1.11111 v 55m9a55 1 . 5 5...? >14». 5.. m . $51.45.. 4.5m 5 558m 5 5.1.1.7.14... ......55555 5555. H ..555 5 5555 5 ‘19. a. 1. ‘ 2*. 5 vmzcwpcouv 5.5 m—nmh 98 1 I"‘Y' “w «r . --2. vv+7~P1 , . .- “'1‘!" ' .ma-ewmywrwv . I 1 ' 1 . 1 z 1 1 4. 4. V r 5.5% . 3.3, 55.. .5 555 a 5555. « '13....rlz- Avwsc5ucouv 5.5 '10 'l:'1"r9‘1 YF v‘ _. 5 m >1 (1 .1lo1v11l111l.1 .5555 ...; V 1 .. 3.5 3.5.5 5555.... ...-..5355 5.5. .23 NS .o._"5.Nno.To...uNNnN= N. ..No. m: mmfifmouofi mg m. n. 5532. Table F.8 (Continued) 4. 2115‘ .. v V I ' ' *9/24" P1<4>jj 9125121“) 12p: ’ Table F.8 (Continued) 1.. 5 i9/24 p1(¢>.~f 5p; D . _ . . v 7 5 .. w . . .5 5x5 .W m I. ... . 5. 5 {we .Vr...) . ... ... r4... "1;, 5M1, H {rm { ,19/247-p1‘c» - ‘59"- + we? TabIe F.8 (Continued) 203 a. .-.)..wrtllll ))))) 5 hi or“... v LLll l; 5.55}; I .. .5 \ . e 55.55. -_.. .... .53 5. '4‘? .5 z . ‘,IVI|)I|.E 582.583 52 2%. 204 5 . ...55.. 5.5.5555. Hwy... 5 .2. 5w. 5.... 53.8 .5. ..3. .... .5. 5.50....” 5555:5588 .5... 5 2.5.55 205 __ 55.5.5... Afim::~wco.uv. m. ... .U—QMH Table F.8 (Continued) 207 Tab1e F.8 (Continued) Tab1e F.8 (anfinued) 02:; "...-.71" - "2}U‘W "“ ’w rig-:4- 52,3; ,4 ri- . »~5 ‘5au‘wirii - ' 9/2 P2< 209 5 w .5 5. 1 we .553- » 655.5555 5.5.. 5 5.85 5... 5.555.555.5285 .m. 5. 35:55 210 . 5.5.5 ...... ...355 .5555. . t§ull.n|(|L||'lllr! |L||IIDLHI ...: 15.".- 5 .5 ‘ /l+—.. 5 33555568 m. 5 53555.55 2H s 311.511-! m 55.55....mm.\5.. . 5 5 I III inn-mun; «A». 0.1., .. , 1‘1. :1? .‘ll. 3553:555533 55.5 m 355 . 1431.5 ’5'ELYJfi-l . :-$/ . ...v v rw.s,.fIM.LwLx.5 ..l. .ix.u.....kfin.vr 151...». .. ..., 5 .. fi3<7>1§£250p - y t 0 . 9/25. 93%» f 212 - i /2 f f: v5 £5 5v.’ 5 A m 5 5 i _ 9 55. 5 5. _ ...... . 5 ) {2'me Table F.8 (Continued 213 .r. 6.r.9V.v»I.l 5 m .. 5 5 214 u 554Ak'o-5én u‘u/ I! EAfli ‘uf’vi Acmsc5ucouv w. , K5“. , 059m5 215 2555:3553 w.5 m 52.5 216 Tab1e F.8 (Continued) 5 _ _ 5:9/25‘ p5<4>> 45“. ‘. 5ygi'. .. ,5 r ‘SZOOnV’i . "3F \ 5 ~ ‘ . v .n", :i .. . v :I.1xt!5§'£§} ., .--. -5- n _ 9 I 5.... 57.5 ... H E 0;;3" i 0133 I 1 I I P5. '9‘ Q ¢ :5/25 p56) Aw , . . ‘ '1' . . _. 9/25; ..+..5v? "“1‘. , + 5w. .550» 5mm .o._”m .N o. Todmumzumuumonmz .mmznemouofi m.u m—nch 217 218 Table F.9 (Continued) 35W Y 59534595 .Y..§.-;.f9./1 6 ZOOIV‘ +5 219 Table F.9 (Continued) 5 3... 5w“ W D. W. wzqo-v; :sv' H . 5 , .. #1.“. 5 5+ 3 -51- i 220 Asa: ...: . .......«.w L. :08 m5 .5 2an 221 5.55.5 55.5.55i15555. . .z1lliyll' l.l 5 53:55:03 m. 5 «53¢ 222 5 . . 51.1.. 5 +1 5. -I.Tr 5555.5 5.55 .555- 55 .5 @555. Table F.9 (Continued) 223 ‘* ‘Hjfl 7'. - . , . g. ‘ I L‘ .' .- -~ ...- u. 5-... “._.-V ... u“ . c? : ‘ 31mm“ 10 S ~lml§nw II III] 224 I 5"!»‘11 58:55:08 mm. 553.2. 225 « L. . . , 582.5585 5.5 .5355 226 5 ... .. .55.. .....E . 555 . . ...]. 5.. H r + 5 .5 5.... - NEEHHSS 5.5 535.5 227 Table F.9 (Continued) 1 -_ ~:_§ I 7 , - I 1 I I +9/1 P3 7) zaps, V . ...-.. I I o i 1:1 - P . . . .“ 20M; i Tab1e F.9 (Continued) 1.57?— . I f ,. , i 320m?" ’5V :9/1" 94 0* ' ... I 1 ‘ " " ‘ . > _ > a: k 229 Table F.9 (Continued) ?. 3,... W. ,1 p ..M. W- . "Vs 505M; £51.!- . 230 b... . E. ... miter. €23.83 2 2% Table F.9 (Continued) W m .- Ill:- , b ,j.:-._. I ‘ _., . 7" “1.3-(qr; Iggy “72': 232 31‘“ $13 32% Eduom ......3 gm .o._§.mno.7o.mmnwzum ”No5: .mozafmouofi m”: of“. 32¢ 233 234 NNNN H. 3N... NNNNH .....NNN :NNNNN . ‘ l r. . u. . '58“ .... 5.. Q L. I ...r .-.¢.¢4“:¢¢ ...NNVNN NNNNM ..NN N MEN N; A 82588 E. N 2%; 235 Table F.l0 (Continued) 235 . \:+ E? fie“ .N ‘AVL EL: go Howie, _ ’ 1 '9/18 P163} '2 T101151; 7- ‘ 237 Table F. l0 (Continued) --.-.4: - g 1 smv ~ 1 5v_ 9/18 P1<6> 10ps:‘ N 9/1 91 D 5,15 1 , T “.31 .‘d 5 , 238 6 €255.53 S. N mEmN. 239 <1 .4; ' J’fql ...... ... N ..u .w a ‘ ... NBNEESV S. N NZNN Table F.l0 (Continued) 240 1 saw 1? svl 5g m9: 5p: 3 l l 241 tinued) .9- - --- --.. ’7“ t , ; I ~ .=> ;9/18 92w 2 1955'. f ‘ . Table F.l0 (Con 5 , ................. 1'; .4 :11 u ”T v'.V ‘I J M N , -l 4' .2” 243 Table F.l0 (Continued) 1t» 1 0 /13 9315) i 5113.1 :9 N N. rill—£41,. CIT.-. l. 244 ) Table F.l0 (Continued 3 10m; .0 K. D. M? W. 39/113931» é 2pp5:;2_ l . . ......r.. I... IN.» firmuu ...-o.- .1 .‘a 245 .«N 1.1. (at. 9:17..- ..w. quwww N... . 32558 EN as: Table F.l0 (Continued) 246 Table F.lO (Continued) 247 ‘1 wow 1 5v? 39/13 95w» 1 1115:; III ":1 Illin- Em. Table F.10 (Continued) 1 I l 1 7‘. ? 39/1} 95(4) 7 ‘ g LIST OF REFERENCES .[l]. [2]. [3]. [4]. [5]. [6]. [7]. [8]. [9]. [10]. LIST OF REFERENCES R. n. F. Gross and J. F. Bott, Handbook of Chemical Lasers, (New York: John Wiley and Sons, 1976). J. V. V. Kasper and G. C. Pimentel, "HCl Chemical Laser," Phys. Rev. Lett. lfi_(10), 352-354 (1965). K. L. Kompa and G. C. Pimentel, "Hydrofluoric Acid Chemical Laser," J. Chem. Phys. fiz_(2), 857-858 (1967). 0. D. Krogh and G. C. Pimentel, "Chemical Lasers from the Reactions of ClF and ClF3 with H2 and CH4: A Possible Chain-Branching Chemical Laser," J. Chem. Phys. §§ (2), 969-75 (1972). N. Jonathan, C. M. Meliar-Smith, S. Okuda, D. H. Slater, and D. Timlin, "Initial Vibrational Energy Level Distributions Determined by Infra-red Chemiluminescence II. The Reaction of Fluorine Atoms with Hydrogen Halides," Molec. Phys. §2_(4), 561-574 (1971). K. G. Anlauf, J. C. Polanyi, H. H. wong, and K. B. Woodall, "Distribution of Reaction Products 111. Cl + H1, C1 + DI," J. Chem. Phys. gg_(11), 5189-5190 (1968). J. C. Polanyi and K. B. Woodall, "Energy Distribution Among Reaction Products. VI. F + H2, 02’" J. Chem. Phys. §Z_(4), 1574-1586 (1972). J. C. Polanyi and J. J. Sloan, "Energy Distribution Amon Reaction Products. VII. H + F2,“ J. Chem. Phys. §Z_(1l?, 4988-4998 (1972). 0. M. Batovskii, G. K. Vasilijev, E. F. Makarov. and V. L. Talrose, "Chemical Laser Operating on Branched Chain Reaction of Fluorine with Hydrogen," Sov. Phys. JETP Lett. 2_(6), 200-201 (1969). N. G. Basov, L. V. Kulakov, E. P. Markin, A. I. Nikitin, and A. N. Oraevskii, "Emission Spectrum of a Chemical Laser Using an H + F Mixture," Sov. Phys. JETP Lett. 9 (11), 375-378 (1963). 2 ‘ 249. [ll]. [12]. [13]. [14]. [15]. [16]. [17]. [18]. [19]. [20]. [21]. [22]. [23]. 250 L. D. Hess, "Pulse Laser Emission Chemically Pumped by the Reaction Between H drogen and Fluorine," J. Chem. Phys. §§. (5), 2466- 73 (1971. N. R. Greiner, L. S. Blair, E. L. Patterson, and R. A. Gerber, "100 Gigawatt H + F2 Laser Initiated by an Electron Bean," IEEE J. Quant. Elec.2 QE- (9L 780 (1974). R. L. Kerber, A. Ching, M. L. Lundquist, and J. S. Whittier, "An Efficiently Initiated Pulsed H F2 Laser," IEEE J. Wuant. Elec. 59-9 (6), 607-609 (1933). J. V. Parker and R. R. Stephens, "Pulsed HF Chemical Laser with High Electrical Efficiency,“ Appl. Phys. Lett. gg_(9), 450-2 (1973). S. Marcus and R. J. Carbone, "Gain and Relaxation Studies in Transversely Excited HF Lasers," IEEE J. Quant. Elec. QE-B (7), 651-5 (1972). D. H. Stone, "An Information Theory Approach to Hydrogen Halide Reaction Product Distributions," Ph.D. Thesis (1980). S. N. Suchard, R. W. F. Gross, and J. S. Whittier, "Time-Resolved Spectroscopy of a Flash- Initiated H2 + F2 Laser," Appl. Phys. Lett. 19 (10), 411 -3 (1971). G. G. Dolgov- Savel' ev, V. A. Polyakov, and G. M. Chumak, "Laser Emission in the 2. 8 Range Involving Vibration- Rotation Transi- Eions)in the HF Molecule," Sov. Phys. JETP 31 (4), 643- 6 1970 . N. R. Greiner, "Submicrosecond Pulses from a Hydrogen- -Fluorine Laser with High Energy Density and Quantum Efficiency," IEEE J. Quant. Elec. QE- 8(12L 872- 6 (1972). N. Cohen, private communication. J. K. Hancock and W. H. Green, "Vibrational Deactivation of HF(v = l) in Pure HF and in HF-Additive Mixtures," J. Chem. Phys. §Z_(ll), 4515-4529 (1972). W. R. Watkins, R. L. Spellicy, K. 0. White, B. Z. Sojka, and L. R. Bower, "Water Absorption Coefficients at HF Laser Wave- lengths (2.64-2.93 um)," Appl. Opt. 1§_(10), 1582-9 (1979), R. L. Kerber, R. C. Brown, and K. A. Emery, "Rotational Nonequilibrium Mechanisms in Pulsed H 2 Chain Reaction Lasers. 2: Effect of VR Energy Exchange, Appl. Opt. 19 (2), 293-300 (1980). [24]. [25]. [26]. [27]. [28]. [29]. [30]. [31]. [32]. [33]. [34]. [35]. [36]. 251 G. M. Jursich and F. F. Crim, "Vibrational Relaxation of HF (v = 3,4,5)," J. Chem. Phys. Zfi_(8), 4455-4464 (1981). T. J. Foster and F. F. Crim, "Vibrational Relaxation of HF (v =3 4 ,5) Between 300 and 700 K, " J. Chem. Phys. Z§_(8), 3871-5 (1981). R. F. Heidner III, J. F. Bott, C. E. Gardner, and J. E. Melzer, "Absolute Rate Coefficients for F + H and F + D2 at 295K," J. Chem. Phys.‘ E (10), 4509-14 (19791. E. Wurzberg and P. L. Houston, "The Temperature Dependence of Absolute Rate Constants for the F + H2 and F + D2 Reactions,? J. Chem. Phys. 22 (9), 4811-4 (1980). G. G. Dolgov- Savel' ev, V. F. Zharov, Yu. S. Neganov, and G. M. Chumak, "Vibrational- Rotational Transitions in an H Chemical Laser," Sov. Phys. JETP §4_(1),34-7 (1972).F V. T. Galochkin, S. I. Zavorotnyi, V. N. Kosinov, A. A. Ovchinnikov, A. N. Oraevskii, and N. F. Starodubtsev, "Investi- gation of the Characteristics of a Chemical HF Laser Excited 11953; Laser Pulses," Sov. J. Quant. Elec. §_(l), 66-69 H. Hokazono, K. Hishihuma, K. Watanabe, M. Obara, and T. Fujioka, "A High Efficiency HF(H/ ) Chemical Laser Ini- tiated with a Surface-Spark Ultravigl eg Flash, " J. Appl. Phys. g; (3), 1359-63 (1982). R. W. F. Gross, J. Cohen, and T. A. Jacobs, "HF Chemical Laser Produced by Flash Photolysis of F202H2 Mixtures," J. Chem. Phys. 4§.(8), 3821-2 (1968). K. L. Kompa, J. H. Parker, and L C. Pimentel, "UF6 Hydrogen Fluoride Chemical Laser: Operationa and 6Chemgstry," J. Chem. Phys. gg_(10), 4257-64 (1968). K. L. Kompa, P. Gensel, and J. Wanner, "New Hydrogen Fluoride Chemical Lasers," Chem. Phys. Lett. §_(4), 210-12 (1969). V. S. Burmasov, G. G. Dolgov-Savel'ev, V. A. Polyakov, and G. M. Chumak, "Quantum Yield of Generation of an H2 + F2 Mixture," ZhETF Pis. Red. 19_(l), 42-4 (1969). A. N. Chester and L. D. Hess, "Study of the HF Chemical Laser by Pulse-Delay Measurements," IEEE J. Quant. Elec. QE-8 (1), 1-13 (1973). P. Gensel, K. L. Kompa, and J. Wanner, "Flash Characteristics as an Important Parameter in Chemical Laser Experiments," Chem. Phys. Lett. Z_(6), 583-6 (1970). [37]. [38]. [39]. [40].. [41]. [42]. [43]. [44]. [45]. [46]. [47]. [48]. [49]. 252 L. D. Hess, "Chain Reaction Chemical Laser Using Hz-FZ-He Mixtures," Appl. Phys. Lett. 12_(l), 1-3 (1971). J. Wilson and J. S. Stephenson, "Atmospheric-Pressure Pulsed Chemical Laser," Appl. Phys. Lett. 29_(2), 64-6 (1972). S. N. Suchard, R. L. Kerber, G. Emanuel, and J. S. Whittier, "Effect of H2 Pressure on Pulsed H Laser. Experiment and Theory," J. Chem. Phys. §z_(1§), 5665- 75( (1972). D. B. Nichols, K. H. Wolstad, and J. D. McClure, "Time Resolved Spectroscopy of a Pulsed H2 Laser with Well-Defined Ini- tial Conditions," J. App1. 2PhyE. 45 (12), 5360-6 (1974). H. L. Chen, R. L. Taylor, J. Wilson, P. Lewis, and W. Fyfe, "Atmospheric Pressure Pulsed HF Chemical Laser," J. Chem. Phys. 61 (l), 306- 18 (1974). S. N. Suchard, "Lasing from the Upper Vibrational Levels of a Flash-Initiated H2 + F2 Laser," Appl. Phys. Lett. 2§_(2), 68-70 (1973). - V. P. Borisov, S. D. Velikanov, S. B. Kormer, M. V. Sinitsyn, and Yu. N. Frolov, "Investigation of Spectral and Time Char- acteristics'of a HF Chemical Laser," Sov. J. Quant. Elec. Z_ (2), 187-90 (1977). ' .Dbara and T. Fujioka, "Time-Resolved Spectroscopic Studies on the TE HF Chemical Lasers Using the Mixtures of SF6/H2 and SF 64,/CH " Jap. J. Appl. Phys. 13 (4), 675- 583 (1974). N. G. Basov, V. T. Galochkin, L. V. KuIakov, E. P. Markin, A. I. Nikitin, and A. N. Oraevskii, "Chemical Lasers Utilizing Mixtures of Fluorine or Nitrogen Fluorides with Deuterium," Sov. J. Quant. Elec. l_(4), 348-54 (1972). W. H. Green and M. C. Lin, "Pulse-Discharge-Initiated Chemical Lasers. III. Complete Population Inversion in HF," J. Chem. Phys. §fl_(7), 3222-3 (1971). T. V. Jacobson, G. H. Kimbell, and D. R. Snelling, "A High- Repetition- Rate Chemical HF Laser," IEEE J. Quant. Elec. g_;__(4), 496- 7 (1973). H. Pummer, W. Breitfeld, H. Wedler, G. Klement, and K. L. Kompa, "Parameter Study of a lO-J Hydro en Fluoride Laser," Appl. Phys. Lett. 22_(7), 319-20 (1973). S. Marcus and R. J. Carbone, FPerformance of a Transversely Excited Pulsed HF Laser," IEEE J. Quant. Elec. QE-7 (10), 493-4 (1971). [50]. [51]. [52]. [53]. [54]. [55]. [56]. [57]. [58]. [59]. [60]. [6l]. [62]. [63]. 253 N. R. Greiner, "Time-Resolved Output Spectrum from a Hydrogen Fluoride Laser Using Mixtures of SF6 and HI," IEEE J. Quant. Elec. gE-ll (10), 844-5 (1975). C. J. Ultee, "Pulsed Hydrogen Fluoride Lasers," IEEE J. Quant. Elec. gE-6 (10), 647-8 (1970). N. G. Basov, V. T. Galochkin, V. I. Igoshin, L. V. Kulakov, E. P. Martin, A. I. Kikitin, and A. N. Oraevsky, "Spectra of Stimulated Emission in the Hydrogen-Fluorine Reaction Process and Energy Transfer from DF to C02," Appl. Opt. lQ_(8), 1814-20 (197l). G. G. Dolgov-Savel'ev and A. A. Podminogin, "Pulsed Laser Utilizin a Fluorine and Hydrogen Mixture," Sov. J. Quant. Elec. g_(4), 348-352 (1973). K. J. Pettipiece, "A TEM 0 Short Pulse KF 05cillator,‘ Chem. Phys. Lett. 15_(2), 261-9 (1972). ~ H. H. Green and M. C. Lin, "Pulsed Discharge Initiated Chemical Lasers-Part II: HF Laser Emission from NF and N2F4 Systems," IEEE J. Quant. Elec. 95-7-(2), 98-l0l (197i). A. A. Podminogin, "Hydrogen Fluoride (Deuterium Fluoride) Pulsed Laser," Sov. J. Quant. Elec. §_(3), 229-3l (1973). T. F. Deutsch, "Laser Emission from HF Rotational Transitions," Appl. Phys. Lett. ll_(l), l8-20 (1967). D. P. Akitt and J. T. Yardley, "Far-Infrared Laser Emission in Gas Dischar es Containing Boron Trihalides," IEEE J. Quant. Elec. QE-G (2), ll3-6 (1970). O. R. Wood, E. G. Burkhardt, M. A. Pollack, and T. J. Bridges, "High Pressure Laser Action in l3 Gases with Transverse Excitation," Appl. Phys. Lett. l§_(4), 112-5 (1971). O. R. Wood and T. Y. Chang, "Transverse-Discharge Hydrogen Halide Lasers," Appl. Phys. Lett. 39_(2), 77-9 (l972). J. H. Smith and D. w. Robinson, "Chemical Pumping of Pure Rotational HF Lasers," J. Chem. Phys. Zfl_(9), Slll-S (198l). E. R. Sirkin and G. C. Pimentel, "HF Rotational Laser Emission through Photoelimination from Vinyl Fluoride and l,l-Difluoro- ethene," J. Chem. Phys.'Z§_(2), 604-l2 (1981). T. F. Deutsch, "Gain Measurements on Uniformly Excited HF/DF TEA Lasers," IEEE J. Quant. Elec. QE-lO (1), 84-7 (l974). [64]. [65]. [66]. [671. [68]. [69]. [70]. [71]. [72]. [73]. [74]. [75]. [76]. 254 C. R. Jones, "Gain and Energy Measurements on an HF/DF Elec- trically Pulsed Chemical Laser," Appl. Phys. Lett. gg_(15), 653-5 (1973). R. A. Chodzko, D. J. Spencer, and H. Mirels, "Zero-Power Gain Measurements in a CH HF Laser Using a Pulsed-Probe Laser," IEEE J. Quant. Elect. gE—9 (5), 550-3 (1973). S. N. Suchard, "Small-Signal Gain and Time-Resolved Spectro- scopy of the Dz-F /CO2 Pulsed Chemical Transfer Laser System," a. Quant. Elec. gE-io (l), 87-9.(l974). L. J. Denes and L. A. weaver, "Laser Gain Characterization of Near-Atmosphere CO /N2/He Glows in a Planar Electrode Geome- try," J. Appl. Ph 5. fl§_(9), 4125-36 (1973). R. C. Brown, "A Theoretical Assessment of Vibrational to Rotational Energy Exchange in the Hydrogen Fluoride Chemical Laser," Ph.D. Thesis (1980). J. J. T. Hough, "A Theoretical and Experimental Investigation of the Mechanisms of the Hydrogen Fluoride Pulsed Chemical Laser," Ph.D. Thesis (1975). R. L. Kerber, G. Emanuel, and J. S. Whittier, "Computer Model- ing and Parametric Study of a Pulsed H2 + F2 Laser," Appl. Opt. ll_(5), 1112-23 (1972). N. Cohen, M. A. Kwok, R. L. Wilkins, and J. F. Bott, "The Status of Rotational Disequilibrium in HF Chemical Lasers," (to be published). M. C. Potter and J. F. Foss, Fluid Mechanics, (New York: The Ronald Press Company, 1975). D. J. Spencer, J. A. Beggs, and H. Mirels, "Small Scale HF(DF) Chemical Laser," J. Appl. Phys. §§_(3), 1206-1211 (March 1977). R. C. Brown and R. L. Kerber, "Parametric Study of Rotational Nonequilibrium Models of Pulsed HF and DF Chain Reaction Chemi- cal Lasers," Michigan State University Technical Report, MSU-ENG-BO-Ol3 (January 1980). N. K. Jaul and R. L. Kerber, "A Theoretical Assessment of the Kinetic Limitations of the Chemically Pumped HF Optical Re- sonant Transfer Laser," Michigan State University Technical Report, MSU-ENG-82-018 (September 1982). M. F. Heisbach, V. R. Buonadonna, C. J. Artura, N. B. Shepherd, J. E. Schrader, and J. D. McClure, "PHOCL-SO: A 50 Liter Photoinitiated Chemical Laser," Boeing Report 0180-24898-1, Boeing Corporation, Seattle, Washington (November 1978). [77]. [78]. [79]. [80]. [81]. [82]. [83]. [84]. [85]. [86]. [87]. [88]. [89]. [90]. 255 J. S. Whittier and R. L. Kerber, "Performance of an HF Chain-Reaction Laser with High Initiation Efficiency," IEEE J. Quant. Elec. QE-lo (11), 844-7 (1974). J. B. Levy and B. K. w. Copeland, Hydrogen-Fluoride Reaction. II. action," J. Phys. Chem. §9_(2), "The Kinetics of the The Oxygen-Inhibited Re- 408-16 (1965). R. L. Taylor, P. F. Lewis, and J. Cronin, "A Mechanism for the Photochemical Induced Combustion of H/ Mixtures Inhibited by 02," J. Chem. Phys. 73 (5) 2218- 23 (1950 R. A. McClatchey, N. S. Benedict, S. A. Clough, D. E. Burch, R. F. Calfee, K. Fox, L. S. Rothman and J. S. Garing, "Air Force Cambridge Research Laboratory Atmospheric Absorption Line Parameter Compilation," AFCRL-TR-73-0096 (1973). R. T. Lagemann and E. A. Jones, "The Infrared Spectrum of Sulfur Hexafluoride," J.C P 19(5), 534-6 (1951). H. R. Carlon, "Infrared Absorption Coefficient (3- 15 pm) for Sulfur Hexafluoride (SF6 ) and Freon (CC12F2)," A 0 18(10), 1474- 5 (1979). G. Herzberg, Molecular Spectra and Molecular Structure III. Electronic Spectra and Electronic Structure of P01yatom1cTMolecules, (Princeton: D. Van Nostrand Co., 3318). Inc. 1967). ' G. 8. King, private communication. G. Guelachvili, "Absolute Navenumber Measurements of 1-0, 2-0, HF and 2-0, H35Cl, H37C1 Absorption Bands," Opt. Comm. 12(1), 150-4 (1976). w. Lafferty, private communication. "High Resolution Infrared P. E. Frawey and K. N. Rao, Spec. 29(3), 348-64 (1969). Spectra of Water Vapor,“ J. Mol. C. J. Ultee, l'Compact Pulsed HF Lasers," Rev. Sci. Instr. 1174-6 (1971). R. V. Jacobson, G. H. Kimbell and D. R. Swelling, "A High Repetition-Rate Chemical HF Laser," IEEE J. QEL QE- 9(4), 496- 7 (1973). O. D. Krogh and G. C. Pimentel, "CIFx -H Chemical Lasers (x = 1,3,5): Vibrational- Rotational xEmgssion by HF from States with High Rotational Excitation," J. Chem. Phys. 67 (7), 2993-3001 (1977). "’ [91]. [92]. [93]. [94]. [95]. [96]. [97]. [98]. [99]. [100]. [101]. [102]. [103]. [104]. [105]. [106]. 256 N. Skribanowitz, I. P. Herman, R. M. Osgood, M. S. Feld, and A. Javan, "Anisotropic Ultrahigh Gain Emission Observed in Rotational Transitions in Optically Pumped HF Gas," Appl. Phys. Lett. §Q_(1l), 428-31 (1972). J. Butler, private communication. W. O. Jeffers, private communication. L. Kanko, private communication. N. Cohen, "A Review of Rate Coefficients in the H + F Chemical Laser System-Supplement (1977)," TR-78-31, Tie Aerospace Corporation, El Segundo, California (June 1978). R. L. Wilkins, "Mechanisms of Energy Transfer in Hydrogen Fluoride Systems," J. Chem. Phys. §Z_(12), 5838-54 (1977). F. E. Bartoszek, D. M. Manos, and J. C. Polanyi, "Effect of Changing Reagent Energy. X. Vibrational Threshold Energies for Reaction Paths HF(v)+D, F+HD and H+DE," J. Chem. Phys. §2_ (2), 933-5 (1979). ' H. A. Watts, private communication. R. L. Kerber and J. J. T. Hough, "Rotational Nonequilibrium Mechanisms in Pulsed H + F Chain Reaction Lasers. 1: Effect on Gross Laser Performgnce Parameters," Appl. Opt. lZ_(15), 2369-80 (1978). P. A. Lovoi, R. Maynard, and B. Smith, "Flashlamp Efficiency for Photoinitiation in HF/DF Lasers," Proceedings Internation- al Conference on Lasers 1979, 530, (December 1979). M. J. Berry, "A Comparison of Photolytic Fluorine-Atom Sources for Chemical Laser Studies," Chem. Phy. Lett l§_(2), 269-73 1972 . N. R. Greiner, "A Thermal Actinometer for Flash Photolysis," Chem. Phys. Lett. l§_(2), 324-7 (1972). S. N. Suchard and D. G. Sutton, "Laser Actinometry of Flash-Photolyzed H /F /He Mixtures," IEEE J. Quant. Elec. 95-10 (5), 490-3 (1973 . ‘ B. 0. Crane, Ph.D. Thesis, 1980. N. Cohen, private communication. S. J. Leone, private communication. [107]. [108]. [109]. [110]. [111]. [112]. [113]. [114]. [115]. [116]. 257 G. Emanuel, W. 0. Adams, and E. B. Turner, "RESALE-l: A Chemical Laser Computer Program," TR-0172(2776)-1, The Aerospace Corporation, El Segundo, CA (March 1972). R. L. Wilkins, "Vibration-Rotation Bands of HF and DF," TR-0077(2063)-7, The Aerospace Corporation, El Segundo, CA (September 1977). J. J. T. Hough, "Lorentz Broadening in the Modeling of the HF Chemical Laser," Appl Opt l§_(8), 2297-2307 (1977). R. E. Meredith and F. G. Smith, "Investigations of Fundamental Laser Processes Vol. II. Computations of Electric Dipole Matrix Elements for Hydrogen Fluoride and Deuterium Fluoride," 84130-39-T(II), The Environmental Research Institute of Michigan, Ann Arbor, MI (November 1971). J. M. Herbelin and G. Emanuel, "Einstein Coefficients for Diatomic Moelcules," J Chem Phys §Q_(2), 689-696 1974). L. F. Shampine, H. A. Watts, and S. M. Davenport, "Solving Nonstiff Ordinary Differential Equations-The State of the Art," SIAM Review l§_(3), 376-411 (1976). E. B. Turner, G. Emanuel, and R. L. Wilkins, "The NEST Chemistry Computer Program, Vol. I," TR-0059(6240-20)-1, The Aerospace Corporation, El Segundo, CA (July 1970). FL Daniels and R. A. Alberty, Physical Chemistr ,jgth edition, (New York: John Wiley an ons, 1975). R. K. Steunenberg and R. C. Vogel, "The Absorption Spectrum of Fluorine," J Amer Chem Soc Z§_(5), 901-902 (1956). A. Yariv, Quantum Electronics, 2nd edition (New York: John Wiley and Sons, 1975).