0101 05-1-F THE UNIVERSITY OF MIC-IIGA N COLLEGE OF ENGINEERING Iligch Altitude Engineeriing I;latorltory'y Departments of Aerospace Engineering and Atmospheric and Oceanic Science Final Report TRANSMISSIVITY OF CARBOQN MONOXIDE S.,., yon.:.... Talla raju. L. W.' Chaney ORA. Project 010105 under contract with: NATIONAL AERONAUTICS AND SPACE ADMINISTRATION CONTRACT No. NSR 23-005-470 WASHINGTON, D. C. administered through OFFICE OF RESEARCH ADMINISTRATION ANN ARBOR February 1973

I

TABLE OF CONTENTS List of Tables v List of Figures vii Abstract ix 1. Introduction 1 2. Experimental 2 3. Analysis 4 4. Results and Discussion 8 References 16 Appendix: Data Summary 18 111

LIST OF TABLES Table Page 1 Regions of the band studied 3 2 Band strength from individual measurements 8 3 Band Intensities of the 4. 6pvm band of CO 9 4 Measured line intensities and half-widths 1 0

LIST OF FIGURES Figure Page 1 Self-broadened half-widths of CO lines versus I m 12 2 Nitrogen broadened half-widths of CO lines versus ImI. - 13 12 16 3 Spectra of the R20 line of C 12016 selfbroadening study. 14 4 Spectra of the R10 line of C12016 nitrogenbroadening study. 15 12 16 5 Spectra of R1 line of C 206 nitrogen - broadening study. 15 vii

Abstract The line strengths and self & N2 broadened half-widths for selected lines of the 4. 6u-m fundamental band of carbon monoxide have been determined. -1 -1 The band strength determined 281 ~ 14 cm (atm. cm. ) at stp. is higher than previously reported measurements. The half widths agree well with other measurements and calculations in the literature. ix

1. Introduction A detailed knowledge of the carbon monoxide spectrum is important in such fields as Astrophysics, Chemistry, Combustion, Meteorology, Molecular Spectroscopy, Planetary Atmosphere Studies, Radiation thermometry, etc. There has been a considerable amount of research to develop a method to determine accurately the global distribution of CO from a satellite measurement. (Grenda et al. 1971) In such a measurement the radiation in the region of the electromagnetic spectrum where CO has absorption bands, (and preferably where no other constituents have absorption bands) emanating from the earth is measured. A prior detailed knowledge of these absorption bands is required for the calculation of the atmospheric transmissivity. Molecular absorption bands contain many hundreds of individual absorption lines, some strong and some weak. If the line parameters (line strength and line half-width) are known for each line in the absorption bands the atmospheric transmissivity can be accurately calculated. Simplified calculations of transmissivity would result if band models could be used. The infrared spectrum of CO has been extensively studied because of its great simplicity. Detailed theoretical calculations of the CO spectra have been made by many authors. (Chackerian 1970, Kunde 1968, Young 1968) Many experimental determinations of line strengths and widths and line shapes have been reported in the literature. Unfortunately due to the inherent difficulty of making such measurements there is a considerable variation among the results obtained by investigators. This report presents the results of -1 experimental measurements conducted at high resolution (0. 05-0. 08 cm ) of the carbon monoxide 4. 6pm absorption spectrum. The experimental results are compared with theoretical calculations.

2. Experimental The experimental measurements were conducted using a Jarrell Ash 1. 83 meter Fastie Ebert vaccuum scanning spectrometer. The detailed description of the apparatus and the measurement technique have been described by Chaney (1972). A graphite rod mounted inside a water cooled housing and heated electrically was used as the source. The sample cell was 8. 74 cm. and 4. 0 cm. in inside diameter. An indium antimonide detector mounted in an IR-15 dewar was used. The output signal from the detector after passing through pre-amplifier and a phase lock amplifier is a D. C. signal varying from 0 to ~ 5 volts and is recorded on a Leeds and Northup Analog chart recorder. The analog signal was digitized by a Hewlett-Packard DY 2401 C Integrating Voltmeter which was interfaced through an H. P. 2526 cardcoupler to an IBM 526 Summary Punch and the data was punched on IBM cards. Each data point corresponded to 0. 1 counter number or drum turn (approximately 0. 008 cml ). The data scan rate was approximately 220 seconds per wavenumber. For most scans a signal to noise ratio of approximately 200:1 was obtained, and a time constant of 0. 1 seconds was used. The gas samples used were from commercial cylinders made by Matheson Co. and had a stated purity of 99. 5% or higher. The isotopic composition of the CO samples was assumed to be close to that of atmospheric CO 12 16 13 16 12 18 12 17 of C 016 (98. 654%), C 16 (1. 106%), C0 (. 202%), C (. 037%) and C13018. 002%). The pressures in the sample cell were measured with a Baratron pressureigauge to 0. 01 mm. of Hg. accuracy. Room temperature during a scan was maintained to ~0. 5~ C. The cell temperatures ranged from 295~ to 301~K for all our measurements and during a scan the cell temperautre did not vary significantly.

The 4. 6 pm band was studied using 2 separate gratings. The first grating was blazed at 16pm and used in the 4th order. The second grating was blazed at 4pm and used in the order. The measured resolutions for the two gratings were approximately 0. 05 and 0. 08 cm respectively. Six regions of the 4. 6pm band of CO were scanned. (Table 1) TAB LE 1 Regions of the band studied (unless otherwise specified the 1st grating was used) RgoWavenumber[cml - 112 16 Region Wvnme[M1) C 0 Lines' 1i 2055-2065 P19-P20 2 2090-2095': P12 3 2135-2150 R1-P2 4 2180-2185 R10 5 2213-2217 R20 6 2224-2230 R23-R24 2nd grating These regions were selected for study because of the relative freedom from overlapping of lines. The spaces between the chopper and the cell (~ 2 cm. long) and between the cell and entrance window (' 2 cm. long) were enclosed with plastic and tape and continuously purged with the boil off from a liquid nitrogen storage tank. However for low pressure data, where the equivalent width of a line was small it was noticed that a small extraneous absorption (in the worst possible case approximately 10% of the equivalent width at 10 mm of Hg. of CO.) led to erronoeous results for line strengths and widths. This was found to be due to the presence of additional CO (produced by cooling water leaking onto the hot graphite source) in the optical path. This additional absorption could also be noticed on some background spectra. Later measurements were greatly improved by having a greater flow rate of nitrogen for the pur ge.,

Some original nitrogen broadened data was also found unuseable because the technique, (gases mixed in the manifold and valved into the sample cell) used in filling the sample cell was not satisfactory. This error was noticed in the spectra, where even for strong lines for the same path length of carbon monoxide, and increasing amounts of the broadening gas nitrogen, the equivalent width randomly fluctuated and sometimes decreased considerably. The method of inletting of the gases was improved and sufficient time for mixing of the gases was allowed (Chaney 1972). 3. Analysis The wavenumbers of the carbon monoxide lines are accurately known (Benedict et al 1962). Using these as calibration it was found that the relation between drum turn and wavelength was almost linear over wavenumber ranges as large as 100 cm-. Thus the drum turns could be converted into wavenumbers using a linear interpolation with wavelength between the CO lines. The spectra were also normalized. As many authors discuss (Korb et al 1968) the determination of the 100% and 0% transmission lines is highly susceptible to error. A 100% transmission scan was originally taken, but it was found that this line fluctuated considerably in amplitude due possibly to the variation of transmission through the order sorting interference filter because of temperature variations. The 100% transmission line was therefore taken as a straight line connecting the spectral regions on each side of absorption line where the absorption appeared small; A wing and base correction (described later) must also be applied. By blocking the beam a 0% transmission line was noted and this was found to be a constant value dependent only on the amplifier gain setting. This 0% line was also obtained from the saturation region of strong lines. The spectra were normalized by using the relation (T100 - T_) / (T100 - T0).

The method of curve of growth was used ill the deternmination line strengths and widths. It was found convenient to separate the data into individual spectral lines for the calculation of the equivalent width. Penner (1959) has shown that the equivalent width is invariant with respect to the spectral slitwidth. For each spectral line the values of equivalent width for the different CO and N2 pressures were calculated using the trapezoidal rule. Where there was weak blending by a neighboring isotopic line a correction for the blending was made by subtracting an area equal to the product of the blending line strength and the absorber path length from the total equivalent width. An iterative procedure vas then carried out for the calc-uiation of the line parameters. An initial guess of line strength S and self- and nitrogen-broadened half-widths acoandaN were made. The values of the parameters a, xD and xL were calculated. a = ('Y / /D) (ln2)i/2 x = (in 2/r) 1/2 (Su/D), XL = Su/21ryL where TD is the Doppler half-width ( 3.58 107 (T/M) 1/2) cm -1 and yL is the Lorentz half-width (a= aCOPcO + aN P cm The values of the equivalent width were corrected to include the wing and base effects. The correction has been given by Korb et. al. (1968). W- W 4 Su _L [1 2 d + 3 (L L v meas +..(. T 3 2 where d 3 -2 d = 15 -2 -2. 1- a -1' -5a +1. 2 4 andnis the distance from the line center to where the expeirimental mneasurement of the area was carried out, and u is the absorber path length. u= PCO' L. (273/T) (atm -cm)STP 5

The wing and base corrected equivalent widths were corrected for Doppler broadening of the spectral line. Corrections for a Voigt line given by Yamada (196;8) in a detailed poiper were used 1. a>0. 5 W W 3 -2 -1 45 -4 -2. 2. 0<a<0. 5 W WV / [+1a6a xL + a for WL>WD and War Wv I I x 27rf/2] for WD>WL -n-l where n is specified by the order of magnitude of a ( -10 1) The subscripts L, D and V refer to Lorentz, Doppler and Voigt respectively. We had a few measurements at extremely low pressures where WD> WL and the line strength could be directly determined from WD. The line strengths calculated were however inaccurate because of CO from the source present in the optical path. For most of the data it was found that WL was much greater than WD. The corrected equivalent widths WL were calculated. The quantity Wl/ PT was calculated and its approximate constancy for a given xL showed that the assumption of Lorentz line shape is accurate for our measurements. Benedict et. al. (1962) have discussed a modified Lorenz shape, where beyond 4 cm from the line center, the line shape is given by multiplying the Lorenz expression by a shape factor given by X = exp [-0. 015 (1 - vol - 4). Burch and Gryvnak (1967) have shown that the extreme wings of the lines are sub-lorentzian. Our measurements show general agreement with their results, in that deviations nee.r the line centers are small. The Ladenberg and Reiche function f was calculated for each data scan f= xL. eXL [J (ixL) - iJl (ixL)] The data for which f< 0. 5 was used to calculate the line strength from -1 S= Wi. x1, cm /(atm- cm) STP 11. f

The reason for taking f<(5 is that for this case x /f is close to unity and so the value of S calculated is not too highly dependent on the value of x ie. on the initial guesses of strength and half-widthl. The value of 0. 5 is arbitrary. The self- broadened data for which f>0. 5 was used to calculate the self broadened half-width f) cm a CO - WL / (27 rT f) cm atm The nitrogen broadened data for which f >0. 5 was used to- calculate the nitrogen-broadened half-width. L =WL/ (27r f) cm and N2 = L COPm atm L LI PN2 N2 Averages of the new calculated parameters S. acO and N were taken. Having obtained these new values for the line parameters S,,cOand aN the calculations were repeated using these new values for the initial guess. The calculations were repeated till a convergence criteria was met. We carried out the iterations till the difference between the calculated values and the guessed values (previously calculated) differed only by. 0002, i.e. well within the estimated experimental error. Due to our limited data, 4 spectral lines P12, R1, R10 and R20were analyzed first. The line strengths of these 4 spectral lines were used to calculate the band strength of the fundamental of C12016 using the Herman Wallis (1955) formula. Sbaidlml (v/ he S Sband I MI exp B.m (m-l)hc m GQ P akT where the Herman Wallis factor F =1 for the fundamental (Young and Eachus m 1966). From our band strength we generated the line strengths of the other 12 16 C 0 lines we have studied. Using these line strengths, and adopting a similar iterative procedure (which converged rapidly) as described above the line widths a COand a were calculated from the equivalent width measurements in self and nitrogen broadening respectively, for all f values. -1 7L = WL / (2irf) cm

WVe lhave determined the line intensities ot( thle other isotopic lines and calculated the Band intensities of thie fundamlentals of the isotopic species 13 16 12 18 12 17 CO, C12018 and C07 4. Results and Discussion The line strengths and line widths of the P12, R1, R10 and R20 were calculated and are tabulated in Table 4. The Band strength for the fundamental of C12016 calculated from these individual measurements is shown in Table 2. TABLE 2 Band strength from individual measurements Line Line strength Band strength Standard -1 -1 -1 deviation cm (atm. cm) stp cm (atm cm) stp deviation % P12 7. 09 279. 9 2. 8 R1 5.35 296. 6 2. 7 R0O 10.36 279. 2 2. 2 R20 1.11 275. 4 1. 0 We did not use the band strength obtained from the R1 line measurement in calculating the average band strength because of the blending of the line with 13 16 the isotopic C 0- R15 line. The average Band strength from the other three measurements gives a value 278. 2 cm-1 (atm. cm. )- stp. Line 12 16 strengths for the other C2 0 lines studied were calculated from this value of band strength and shown in Table 4. The line half-widths were calculated and are tabulated in Table 4. From our measurements the calculations of the 13 16 12 18 12 17 average band strengths for the fundamentals of C13016 C 0 and C 0 are 2. 7, 0. 44, 0. 1 cm1 (atm. cm. )-1 stp. The total band strength of the fundamentals is estimated as 281 4 14 cmr1 (atm-cm)-lSTP. The hot bands are less than 0.1 atrn. cm 2. in strength at this temperature. Table 3 lists other measurements of the 4. 6 /m. band intensity of CO. 8

'IXAI3-E 3 Iatid Itlltlsities or the 4. 6 lm batld of (() Year Author Band Intensity cm -1 (atm. cm)-1 stp. 1951 Penner and Weber 260 ~ 12 1953 Locke and Herzberg 245 1953 Vincent-Geiss e 262 1956 Shaw and France 238 1962 Benedict et al 236 1962 Burch and Williams 260 ~t 20 1965 Armstrong and Welsh 258 ~ 4 1966 Hochard de Molliere 247. 8 1973 Present Study 281 ~ 14 Figure 1 shows our measurements of the self-broadened line widths compared with other recent measurements. The results of this work agree very well with the direct measurements of Kostkowski and Bass (1961). The nitrogen broadened half-widths are plotted in Figure 2 and compared with the theoretical calculations of Bouanich and Haeusler (1972). Our results show fair agreement with the results of other authors. Our total band intensity is approximately 8% higher than the mean -1 -1 value of 260 cm (atm. cm) STP of some of the more recent determinations. We estimate that without the inclusion of the base correction in our analysis we would have obtained a band intensity about 8% lower. This may be responsible for some of the difference between our results and those of the authors in Table 3. The errors in this type of measurement have been very well summarized by Reichle (1969). The mathematical errors of truncation and round off were neglible. The wing and base area correction contributed 18% of the total area in the most extreme case, but was usually half or less of this value. The error in estimating this correction was much less than the error 9

TABLE 4 Measured Line Intensities and Half-Widths S cm-e aC [atm-1 cm-1] CO N Line Isotope Line Center STP atnm cm atrn (ill P20 26 2059.9 1. 0.060 P19 26 2064.4 1. 39. 063 P12 26 2094. 9 7.09.071. 064 Rll 28 2133. 3 0.016 P 2 26 2135.5 4.89.089 Rll 36 2137.6 0.0875 P 1 26 2139.4 2.54.089 R12 36 2140.8 0.085 R14 28 2143.0 0. 010 R13 36 2144.0 0.072 R 7 27 2145.1 0. 0045 R 0 26 2147.1 2. 6.08 7 R16 28 2049.3 0.006 R 1 26 2150.9'3 5.35.089.081 R10 26 2183.2 10.36.070.064 R20 26 2215.7 1.11.055.048 R23 26 2224.7.38. 048 R24 26 2227.6.26. 051 *The notation used to identify the isotope is: 26 = C 12016, 36 = C 0 28 = C12018 and 27 = C12017 10

involved in estimating the background. In addition to the small random fluctuations which were minimized by taking the average of 30 data points on either side of the spectral line, there were systematic shifts in the background level from scan to scan and presumably during each scan. The shifts between scans were systematic and thought to be associated with a slow warming in some part of the instrument. They amounted to about a 5%' change during a 4 hour period and hence would be much less than 1% during the time taken to scan an individual line. In addition to the background fluctuations discussed above, some of the spectra showed depressions or enhancements of the background level over frequency intervals of a few tenths of a wavenumber of an amplitude of about 1% although usually the amplitude was much less (for e. g. see figure5). These appear to be the largest source of error in estimating the backgrounds. It is difficult to assess the effect on the final results, but we estimate an error of 2% or less in the equivalent widths for nmost of the data. The error due to source contamination contributed at worst an error of 8% of the equivalent width for measurements of 10mm Hg of CO self-broadened data, for a few of the first measurements. This error was much smaller for later data where a greater flow rate of nitrogen purge was used. Errors in the measurement of temperature, pressure, composition, path length are less then 0. 5%0 and do not contribute to any significant error in the calculations. The error in the digitizing system was less than 0. 1%. Shift of wavelength marking, due to variation in temperature, was less than 0. 5%(see fig.5 ). We estimate that the error in the measurement of the band strength is i 5%. The error in the measurement of line widths are shown as error bars on Figures 1 & 2. The error bars are the standard deviation of the individual half-widths measurements. Figures 3, 4, 5 show sample spectra obtained. The typical signal to noise ratio obtained is depicted. 11

~ —-0-I Band 0.09 --- 0-2 Band 0-3 Band BB B 4-~~~~~~~~~~~~~~~~~~~4 0.07 E E E 0.06 0 " —We III 5 1~ ~ ~~~~0 15 20 25=mm 0.05 ~~2Ls ~1; BHMS PT 0.04 5 10 15 20 2 E. Imi Figure 1: Self-broadened half-widths of CO lines versus Im Results of this work ( (D-P branch and 4 -R branch) are shown along with results in the 0-1 band of Eaton and Thompson (1959)-ET; Benedict et. al. (1962)-BHMS; 0-2 band of Kostkowski and Bass (1961)-KB; Plyler and Thibault (1963)-PT; Hunt et. al. (1968)-HTP; Bouanich and Brodbeck (1973)-BB; 0-3 band of Burch and Gryvnak (1967)-BG; Bouanich and Haeusler (1972)-BH.

I., I I I I. 0.9 * — 0-I Band 0-2 Band ----- 0-3 Band 0.08 Theory E 0.07 *-.,..BH E o 0.06 B B 0.05 DW._ 0.04 5 10 15 20 25 Iml Figure 2: Nitrogen Broadened half-widths of CO lines versus jmj Results of this work ( 0, P branch and G, R branch) are shown with the theoretical (...... ) rind experimental (0-3 band) results of Bouanich and Haeusler (1972). - BH; aind results of Draegert and Wkilliams (1968) (0-1 band) -DW; and Bouanich anrd Brodbeck (1973) (0-2 band) -BB.

0 0 0 0 cc a0 Co cr. 0 2211.00 2211.50 2212.00 2212.50 2213.00 2213.50 2214.00 2214.50 2215.00 2215.50 2216.00 2216.50 2217.00 2217.50 2218.00 WRVENUMBER 12 16 Fig. 3 Spectra of the R20 line of C 0 self-broadening study Scans. #-519, 520, 521 and 522. 14A

z= on'179.00 2179.50 2180.00 2180.50 2181.00 2181.50 2182.00 2182.50 2183.00 2183.50 2184.00 2184.50 2185.00 2185.50 2186.00 WRVENUMBER Fig. 4. Spectra of the RiO line of C O nitrogen broadening study Scans. # 502, 504, 505 and 507. ZC ta: t2147.00 2147.50 2148.00 2148.50 2149.00 2149.50 2150.00 2150.50 2151.00 2151.50 2152.00 2152.50 2153.00 WFIVENUMBER Fig. 5. Spectra of R1 line of C 0l nitrogen broadening study. Scans. # 510, 512, 514 and 516. 15

Refe rences Armstrong, R. L. and H. L. Welsh (1965), "The Infrared Spectrum of Carbon Monoxide in CO-He Mixtures at High Pressures. " Canad. J. Phys. 43, 547. Benedict, W. S., R. Herman, G. E. Moore and,S. Silverman, (1'962) "The Strengths, Widths and Shapes of lines in the Vibration - Rotation Bands of CO." Astrophysical J. 135, 277. Bouanich, J. P. and C. Haeusler (1972) "Line Widths of Carbon Monoxide Self-Broadening and Broadened by Argon and Nitrogen." JQSRT 12, 695. Bouanich, J. P. et C. Brodbeck (1973), "Mesure des Largeurs et des Deplacements des Raies de la Bande O — 2 de CO Autoperturbe et Perturbe par N2, 02' H2, HC1, NO et CO2o" JQSRT 13, 1. Burch, D.E. and D.A. Gryvnak (1967) "Strengths, Widths and Shapes of the Lines of the 3 v CO Band." J. Chem. Phys. 47, 4930. Burch, D. E. and D. Williams (1962), "Total Absorptance of Carbon Monoxide and Methane in the Infrared. " App. Opt. 1, 587. Chackerian, C. (1970), "Calculation of High-Temperature Steradiancy for Vibration Rotation Bands of Carbon Monoxide." JQSRT. 10, 271. Chaney, L. W. (1972), "High Resolution Spectroscopic Measurements of Carbon Dioxide and Carbon Monoxide." Tech. Rep. 036350-3-T. High Altitude Engineering Laboratory, University of Michigan. Draegert, D.A. and D. Williams (1968), "Collisional Broadening of CO Absorption Lines by Foreign Gases." J. Opt. Soc. of Am. 58, 1399. Eaton, D. R. and H. W. Thompson (1959), " Pressure Broadening Studies on Vibration-Rotation Bands. " Proc. of Roy. Soc. Lond. A251,458, 475. Grenda, R. N., M. H. Bortner, PO J. LeBel, H. J. Davies and R. Dick (1971), "Carbon Monoxide Pollution Experiment-I. A Solution to the Carbon Monoxide Sink Anomaly." AIAA paper No. 71-1120. Joint Conference on Sensing of Environmental Pollutants. Palo-Alto Calif. Herman, R. and R. F. Wallis (1955) "Influence of Vibration-Rotation Interaction on Line Intensities in Vibration-Rotation Bands of Diatomic Molecules. " J. Chem. Phys. 23, 637. Hochard de Mollierre, L. (1966), " Mesure de la Dispersion dans la Bande Fondamentale de l'Oxyde de Carbone." J. Phys. Paris 27, 341. Hunt, R. Ho, R. A. Toth antl Es K. Plyler (1968), "High Resolution Deter, mination of the Widths of Self-Broadened Lines of Carbon Monoxide. " J. Chemn. Phys. 49, 3909. 16

Korb,C. L., R. H. Hunt and E. K. Plyler (1968), "Measurement of Line Strengths at Low Pressures - Application to the 2-0 Band of CO." J. Chem. Phys. 48. 4252. Kostowkski, H. J. and A. M. Bass (1961), "Direct Measurements of Line Intensities and Widths in the First Overtone Band of CO." JQSRT 1, 177. Kunde, V. G. (1968), "Theortical Molecular Line Absorption of CO in Late Type Atmospheres." Astrophys. J. 153, 435. Locke, J. L. and L. Herzberg (1953), "The Absorption due to Carbon Monoxide in the Infrared Solar Spectrum. " Canad. J. Phys. 31, 504. Penner, S.,S. (1'959), "Quantitative Molecular Spectroscopy and GCas Emissivities."''Addision-Wesley, Reading Mass. Penner, S. S. and D. Weber (1951), "Quantitative Infrared Intensity Measurements I - CO pressurized with Infrared Active Gases J. Chem. Phys. 19, 807. Plyler, E. K. and R. J. Thibault (1963), "Self Broadening of Carbon Monoxide in the 2s and 3v Bands." J. Res. NBS. 67A, 229. Reichle, H. G. (1969), "The Effect of Several Infrared Transparent Broadening Gases on the Absorption of Infrared Radiation in the 15man band of Carbon Dioxide." Tech. Rep. 05863-17-T, High Altitude Engineering Laboratory, University of Michigan. Shaw, J. H. and W. L. France (1956), "Intensities and Widths of Single Lines of the 4. 7 micron CO Fundamental." Scientific Report 4 on Proj. 587. Ohio State University R. F. Vincent-Giesse, J. (1954), "Mesures d'intensite et de largeur de raies dans les spectres infra rouges de gaz on de vapeurs. Compt. Rend. 239, 251. Yamada, H. (1968), "Total Radiances and Equivalent Widths of Isolated Lines with Combined Doppler and Collision-Broadened Profiles. " JQSRT 8. 1463. Young, L.A. (1968), "CO Infrared Spectra." JQSRT. 8, 693. Young, L. A. and W. J. Eachus (1966), "Dipole Moment Function and Vibration Rotation Matrix Elements for CO. " J. Chem. Phys. 44, 4195. 17

APPENDIX: Data Summary Wave Pressure Scan No. Number Lines mm, Hg. Remarks -1 12 16 cm C12 O CO N 311-314 150 Instrunment Test 315-324 2224-2230 R24/23 0, 10, 20, 40 Self 0, 80, 160, broadening 320, 640, 0; 325-335 2145-2148 R1/0O 0, 10, 40, 160, Self 0, 0, 5, 20, 80, broadening 320, 640; 336-342 2135-2145 P1/2 0, 5, 10, 20, Self 40, 80, 160; broadening 343-349 2058-2065 P19/20 0, 10, 20, 40, Self 80, 160, 320; broadening 447-450 2224-2230 R24/23 0, 10, 40, 80; Self broadening 451-452 2224-2230 R24/23 80 —- 80, 160 Nitrogen broadened 453-454 2224-2230 R24/23 0, 800; Self broadening 455-466 2145-2148 R1/0 0, 0, 0, 10, 40, Self 600, 0, 160, 10, broadening 10, 700, 0; 467-469 2145-2148 R1/0O 10 —- 19, 40, 161 Nitrogen broadened 470-471 2145-2148 R1/0O 700, 10 Self broadening 472-478 2145-2148 R1/O 10 —- 10, 20, 40, 81, Nitrogen 161, 323, 242; broadened 479-481 2145-2148 R1/0 0, 10, 700; Self broadening 482-487 2180-2185 R10 2.5, 5, 10, 40 Self 160, 640; broadening 488 2180-2185 R10 - 10 Nitrogen broadened 18

Wave Pre s sllr'IeC -ali No. Nunibel Litlies mnm. Ilg. 1HRelmr Iks -1 12 16 cm C 0 CO N 39-496 2180-218-5 R10 5 —- 10, 20, 40, 80, NitcL),'Cell 161, 32:3, 646, b)roadened 646 37-500 2180-2185 R10 8, 648, 648, 8; Self broadening ).1-507 2180-2185 R10 8 —- - 12, 20, 40, 80, Nitrogen 162, 322, 643; broadened 38 2180-2185 R10 650 Self broadening )9-516 2145-2150 R/0O 5 —- 0, 10, 20, 40, 81, Nitrogen 161, 321, 644 broadened L7-524 2213-2217 R20 1 2 0, 240, 80, Self 160, 320, 640, 10 broadening 25-531 2213-2217 R20 10 —- 10, 40, 81, 162, Nitrogen 32 3, 645, 2 0; broadened 32 2213-2217 R20 700; Self broadening 51-554 2058-2065 P19/P20 80, 750, 0, 750; Self broadening 55 2058-2065 P19/P20 10 —- 642 Nitrogen broadened 56 2058-2065 P19/P20 750 Self broadening 57 21 35-2150 R1/P2 4 647 Nitrogen broadened 58-561 2135 -2150 R1/P2 750 37-597 2090 2095 P12 20, 5, 10, 20 Self 40, 80, 160, broadening 320, 640, 0, 0; 98-606 2090-2095 P12 1. 5 —- j0, 3, 7, 20, Nitrogen 39, 79, 158, broadened 315, 644; J7-613 2090-2095 P12 2. 5, 5, 10, 20, Self 40( 80, 640; broadening 10

UNIVERSITY OF MICHIGAN 3 9015 02844 0454