AFCRL-70-0373 U-4829

INVESTIGATION OF THE ABSORPTION OFINFRARED RADIATION BY ATMOSPHERIC GASES

Darrell E. BurchDavid A. GryvnakJohn D. Pembrook

Philco-Ford CorporationAeronutronic Division

Ford RoadNewport Beach, California 92663

Contract No. F19628-69-C-0263Project No. 5130

Semi-Annual Technical Report No I

June 1970This research was supported by the Advanced Research

Projects Agency under ARPA Order No. 1366.

This documenc has been approved tor publicrelease and sale; its distrtbution i unlimitid.

Contract Monitor: Robert A. McClatcheyOptical Physics Laboratory D D C

UNITED~~ STTE AICOCEUAUG 13 1970

Prepared for

AIR FORCE CAMBRIDGE RESEARCH LABORATORIES " 8 __ "UNITED STATES AIR FORCE

BEDFORD, MASSACHUSETTS 01730

Reploduced by theCL E A R I G HO U SE

For Fodoial Sciontific & TechnicalInformation Sptingfield Va.. 2,2151

Qualified requestors may obtain additional copies from the DefenseDocumentation Center. All others should apply to the Clearinghousefor Frederal Scientific and Technical Information.

RIWTE SECOIPk

"OiiF SECTION

0S JU 4 .........................

, ', J Ut•'k/A'VAILABILITY CODES

II

AFCRL-70-0373 U-4829

INVESTIGATION OF THE ABSORPTION OFINFRARED RADIATION BY ATMOSPHERIC GASES

Darrell E. Burch

David A. GryvnakJohn D. Pembrook

Philco-Ford CorporationAeronutronic Division

Ford RoadNewport Beach, California 92663

Contract No. F19628-69-C-0263Project No. 5130

Semi-Annual Technical Report No. 1

June 1970This research was supported by the Advanced Research

Projects Agency under ARPA Order No. 1366.

This document has been approved for publicrelease and sale; its distribution is unlimited.

Contract Monitor: Robert A. McClatcheyOptical Physics Laboratory

Prepared forAIR FORCE CAMBRIDGE RESEARCH LABORATORIES

UNITED STATES AIR FORCEb7DFORD, MASSACHUSETTS 01730

ABSTRACT

From spectral transmittance curves of very large samples of CO2 we havedetermined coefficients for intrinsic absorption and pressure-induced ab-sorption from approximately 1130 cm" 1 to 1835 cm'1 . Most of the pressure-inducedlgbsorption arises from the forbidden v and 2v 2 bands of the01 6C120IU molecule, and the major portion of the intrinsic absorptionarises from the same bands of the 01 6C12 01 8 molecule. Curves of theabsorption coefficients for samples of CO2 pressurized to approximately10 atm, or more, with He are presented throughout most of the regionfrom 1885 cm-1 to 2132 cm"1 . These curves and a table of the integratedabsorption coefficient are adequate to determine the strengths of mostof the bands of significance in this region. Spectral transmittancecurves of CO absorption between approximately 2065 cm"I and 2100 cm" 1

are presented for a variety of samples. Curves of the pressure-inducedabsorption by N2 are presented for the 2400-2650 cm"I region.

I2

TABTE OF COUTENTS

Section Page

1 SUN4ARY AND DEFINITION OF TERMS ....... ............ 1 -i

2 ATMOSPHERIC ABSORPTION BY N BETWEEN2400 AND 2600 cmI .................. 2 2-1

Figure 2-1, Normalized Coefficients of Pressure-InducedN2 Absorption Plotted Versus Wavenumber ........... ... 2-4

Figure 2-2, Curves of -LT of N2 and the CortinuumDue to the Extreme Wings of CO 2Lines .... .. ...... ... 2-5

3 ABSORPTION BY LRESAMPLES OCOBETWEEN

I100 AND 2000 cm. ............. ................ ... 3-1

Table 3-1, Centers of CO2 Bands Between1100 and 1450 cm.I ................................... 3-2

Figure 3-1, Spectral Transmittance Curves of Two

Relatively Low-Pressure Samples of CO2 Between1200 and 1400 cm-1 . . . . ... ... . . . . . . . . .. 3-3

Figure 3-2, Spectral Transmittance Curves of Two =IHigh-Pressure Samples of CO2 Between 1100 and 1450 cm . 3-4

Figure 3-3, Spectral Transmittance Curves from

1400 cm"I to 2020 cm 1 ........ ........ 3-5

Figure 3-4, Spectral Plots of K(local) and C0

for Pure CO2 from 1100 to 1500 cm"1. . .s3p2 . ~. . . . . . . . 3-9

Figure 3-5, Spectral Plcot of K(locfl) and Cs

for Pure CO2 from 1490 to 1835 cm.- ........ 3-10

liii

TABLE OF CONTENTS (Cont.)

Section Paxe

4 ABSORPTION COEFFICIENTS AND INTEGRATED ABSORPTANCEFOR CO2 BETWEEN 1885 cm"I AND 2132 cm" 1 . . . . . . . . . . . . . . . 4-1

Table 4-1, Integrated Absorption Coefficients for CO 2 .. 4-2

Figure 4-1, Spectral Curve of (-.•,T)/u for CO2 from1884 to 1930 cm" 1 . . . . . . . . . . . . . . . . . . . 4-3

Figure 4-2, Spectral Curve of (-1,.T)/u for CO2 from

1929 to 1937 cm 1. . . . . . . . . .. .. ............ 4-4

Figure 4-3, Spectral Curve of (-,&T)/u for CO 2 from1936 to 1948 cm" 1 ....... ....... .. ........... .. 4-5

Figure 4-4, Spectral Curve of (-,&T)/u for CO2 from1940 to 2030 cm". ........... ............. .... 4-6

Figure 4-5, Spectral Curve of (-j,eT)/u for CO2 from2030 to 2076 cm 1 4-7

Figure 4-6, Spectral Curve of (-•,T)/u for CO2 from2073 to 2082 cm-'. . . . . . . . . . . . . . . . .. . . . . . . . . . . . 4-8

Figure 4-7, Spectral Curve of (-,hT)/u for CO 2from2080 to 2093 cm 1. . . . . . . . .. .. .. . .. . . . . . . . .. 4-9

Figure 4-8, Spectral Curve of (.&,T)/u for CO.2 from2090 to 2097 cm 1 .. .... ........... 4-10

Figure 4-9, Spectral Curve of (-.&T)/u for CO2 from

2096 to 2128 cm-1 . . . . . . .. .. .. .. .. . .. .. .. .. . .. .. .. 4-11

Figure 4-10, Spectral Curve of '-Y.,T)/u for CO2 from

2126 to 2132 cm'.. *....... ............ 4-12

Figure 4-11, Spectral Curve of (-oT)/u for CO 2 from2130 to 2200 cm'". ...... ......... ........... ..... 4-13

Figure 4-12, Spectral Curves of Transmittance for CO,fe'om 2065 to 2100 cm'1 for Samples 1, 2, and 3 .... ... 4-14

iv

TABLE OF CONTENTS (Cont.)

Section Page

4 Figure 4-13, Spectral Curves of Transmittance for CO2

from 2065 to 2100 cm"I for Samples 4, 5, 6, and 7. . . 4-15

Figure 4-14, Spectral Curves of Transmittance for CO2from 2065 to 2100 cm"I for Samples 8, 91 10, and 11. 4-16

Figure 4-15, Spectral Curves of Transmittance for CO2from 2065 to 2183 cm- for Samples 12 and 15 . . . . . 4-17

Table 4-2, Sample Parameters and fA(v)dvfrom 2065 to 2100 cm. .1 . . . . .. . . . . . . .. . . . . . . .. . . 4-18

5 TRANSMITTANCE DATA STORED ON MAGNETIC TAPE ...... .. 5-1

Table 5-1, Sunmnary of Data Saved on Magnetic Tape. . . 5-2

6 REFERENCES . . . . . .. . . . . . . . . . . . . . . . 6-1

V

F

SECTION 1

SUMMARY AND DEFINITION OF TERMS

The absorption in atmospheric windows frequently cannot be treated by band-model methods which are applicable to spectral regions containing strongabsorption bands. A significant portion of the attenuation in windows may

result from continuum absorption due to the extreme wings of distant linesor to pressure-induced bands. The functional dependence of the continuumabsorption on such parameters as path length, temperature, pressure, andatmospheric composition is quite different from that of line absorption.Laboratory studies of atmospheric gases under controlled conditions fre-quently are required to determine the portion of the absorption due to thecontinuum.

This report deals with a variety of situations in which continuum absorptionand weak-line absorption are important. Section 2 deals with continuumabsorption between approximately 2400 and 2650 cm 1l due to a pressure-inducedband of N2 and to the extreme wings of CO2 lines centered at lower wavenum-bers. Pressure-induced bands and weak forbidden bands of CO2 in the 1100-2000 cm" 1 region are diEcussed in Section 3. By studying samples coveringa wide range of pressures and path lengths, we have determined the contri-bution due to the continuum produced by the pressure-induced bands and thatdue to the weak bands which contain rotational line structure. We havebeen able to measure the CO absorptio , even in the regions of very weakabsorption near 1830 cm" and 1130 cm

-lSeveral CO2 bands which occur between 1890 and 2130 cm play a significantrole in the absorption by long atmospheric paths. Section 4 includesexperimental curves and tables from which the strengths of several of thesebands can be determined. Also included in Section 4 are transmittancecurves and tabulated results 'or several samples studied near 20%0 cm".

I-I

A prominent feature in the spectrum near this wavenumber has proven to beuseful in analyzing spectral curves obtained by aircraft. Section 5 con-tains a summary of experimental data which we have obtained previously andhave recently put on magnetic tape. The tapes and a user's manual will beavailable in the near future for workers who need them.

All of the data included in this report are based on samples contained inone of two multiple-pass absorption cells which have been describedPreviot"•!y. 1 The base 1cCLhs of the cells are approximately 29 m and1 meter. Sample pressures up to 2.5 atmospheres are employed in the longcell; 21 atm is the maximum prersure used in the short cell.

The absorber thickness, u, of a gas sample is given by

u(molecules/cm 2 2.69 x 1019 p*(atm) L(cm) (273/0)

= 7.34 x 1021 p* L/G. (1-1)

The temperature 9 is in degrees Kelvin, and L is the geometrical pathlength through the sample. The density-equivalent-pressure p* of theabsorbing gas approaches its pressure p at low pressures. Because of thenon-linear relationship between the density and the pressure of a gas, p*may differ significantly from p at high pressures. For the gases andpressures used in the present investigation, the following expression issufficiently accurate.

p* p(l c p) (1-2)

The pressures are in atm and, near room temperature, c - 0.005 for N 0 andC02 , and 0.001 for N2 and 0. For the largest sample of C02 , p* is 9nlyapproximately 1I1 greater thian p; in many of the samples the difference isnegligible. In the following discussion certain quantities are said to bepropcrtional to pressure; it should be borne in mind that the more correctquantity, density-equivalent pressure p*, was used in the calculationswhere the difference between p* and p was significant.

The true transmittance that would be observed with infinite resolvingpower is given by

V exp(-uK), or (-1/u)LT' 4 T, (1-3)

1-2

where K is the absorption coefficient. Because of possible variations inK with wavenumber aun Lhe finite slitwidth ot a spectrometer, the observeatransmittance T may differ from T' at the same wavenumber. The quantity Trepresents a weighted average of T' over the interval passed by the

spectrometer.

Throughout this report we deal with three different basic types of absorp-tion. In some spectral regions, an appreciable portion of the absyrptionis due to the extreme wings of lines whose centers are several cm away.This wing absorption has the nature of a continuum, i.e., there is nostructure, and because of the collision broadening of the lines, theabsorption coefficient Cw due to the wing absorption is proportional topressure. (Cw 7 Cw p*) (See Eq. 1-6 and the discussion following it.)

Anotner type of continuum absorption arises from pressure-induced bands;the absorption coefficient, C , for pressure-induced absorption is alsoproportional to pressure. Thg absorption coefficient due to local lineswhose centers occur within a few cm- of the point of observation isdenoted by K(local). The latter quantity may vary rapidly with wavenumberand depends on pressure because of collision-broadening of the absorptionlines.

At a given wavenumber, there may be appreciable absorption due to locallines, the wing continuum, the pressure-inducld continuum, or to any ofthese three. Therefore, the total absorption coefficient ý. in Eq. (1-3)

may be given by

K i(local) + C + C . (1-4)w p

For a sample of absorbing gas only, such as several of the CO 2 samplej

discussed below, we can rewrite Eq. (1-4) as

K K(local) (Co C°0 . ) p*, (absorbing gas only) (-4a)o P

where the normalized coefficients C 0 and CS, are the values of C andasW rpC . respectively, when p* 1. The iubscript's denotes self-broadewning

oXthe lines or self-induced absnrption. Since u is propoitional to pL,

(-.4) due to wing continuum• or pressure-induced continulan is proportionalto p L.

We frequently are concerned with samples containing a non-absorbing gas,along with the absorbing gas. This gas broadens the absorptinn lines andmay induce additional absorption by the pressure-induced bands of theabsorbing gas. For a mixture of an absorbing gas and a broadening gas,

1-3

Eq. (1-4) be'omes

K K(local) + (Cs0 + C0) p* + Cw C P0) b (l-4b)

C 0 and C-0 are the n.,rmalized coefficients for wing absorption andCb,w bhprecsure-influced absorption by the alsor' ng gas due to the presence ofthe broadening gas denoted "-- subscript b.

The intrinsic absorption coefficient due to a single collision-broadenedabsorption line at a point within a few cm"I of the line center, vo, isgiven by the Lorentz shape:

k S 2" (1-5)T (VVo) +

The line strength S = fkdv i5 essentially independent of pressure for theconditions of the present study. It has been shown 2 ) 3 that for 1v-vgreater than a few cmi!, thc Lorentz equation may require modification bya factor of , which is a function of (v-v ). Therefore, Eq. (1-5) becomes

0

-1 k Yo (1-6)

where kL denotes the value given by the Lorentz equation. The value of N<is approx 4 mately equal to unity for small Iv-vol butmay be as small asO.01 for CO2 .ines when (v-vo) is as large as 100 cm"1 .

The 'ialf-width u is proportional to pressure so that k is, in turn,proportional to pressure in the extreme wings where V-V0 1 >> or. Thus,the continuum absorntion coefficient due to the extreme wings of manylines is proportional to pressure, as stated above.

1-4

SECTION 2

-iATMCSPHERIC ABSORPTION BY N BETWEEN 2400 AND 2600 cm

2

The fundamental vibration band of N2 is centered near 2330 cm andproduces measurable absorption from less than 2200 cm" 1 to more than2600 cm" 1 . We have investigated the absorption by this band betweenapproximately 2390 cm" 1 and 2650 cm"I because of its important role inatmospheric transmission in this region. Below 2390 cm" 1 the CO2 ab-sorption in the atmosphere is mucl. greater than the N absorption.The N2 band is pressure induced so that the probabiliiy of a transitionin a given molecule is proportional to pressure, and the absorptance bya qhort path iq proportional to p2 .

Shapiro and Gush4 have published spectral data on N2 absorption forsamples at approximately 20 atm preLsure, Farmer et a15 have alsomeasured the absorption between 2400 cm"1 and 2525 cm" 1 by approxi-I mately 1 atm of pure N2 with path lengths of a few hundred meters.The results of Farmer et al are consistent with the high-pressuredata of Shapiro and Gush if (--.T) is assumed to be proportional top2 L, according to the discussion following Eq. (l-4a). The work re-ported here was undertaken to check the previous work and to determine

the appropriate coefficient CO , for N2 absorption induced by O2.From the value of this coefficient and the corresponding coefficient,C0 ,• for self-induced absorption by N2, we have calculated the trans-mltance by.N2 in the earth's atmosphere from approximately 2390 cm"ito 2650 cm"i.

Vaiues of the absorber thickness of N2 were calculated by use ofEqs. (1-1) and (1-2). Since there is no absorption by N2 other thanthat which is pressure induced, Eqs. (1-3) and (l-4a) combine to give

2-1

I ! t

Sp* C0 (2-1)

u S~p

for a pure N2 sample, and

0 0p* C + p* C (2-1a)"4s,p 02 Op

for a sample of N + 0 Since the true transmittance T' changes slowlywi.Lh wavenumber in continuum absorption, the observed transmittance T isessentially equal to T1. Note that C 0 corresponds to absorption by N2that is induced by 02; the 02 does not'ibsorb. We detected no absorptionby a sample of 8 atm of 02 with a 32.9 meter path length. After valuesof Cs had been found from pure N2 samTles, we determined St from dataon 0 " N mixtures by use of Eq. (2-1a). The results for pu N and for02 + N2 mixtures are summarized in Fig. (2-I). The good agreemeni among fthe pure N2 data at different pressures provides evidence of the pressuredependence assumed in deriving Eq. (2-1). Below approximately 2500 cm- 1 ,

our results for pure N2 differ by less than t 5% from those by Shapiroand Gush. 4 At larger wavenumbers, the agreement is within the accuracy towhich the curves of Shapiro and Gush can be read. Data obtained from sam-ples of N2 with 5 atm of 02 compared well with data from samples of N2with 8 atm of 02, indicating the validity of the term involving Pb inEq. (2-la).

Air contains approximately 78 percent N2, 21 percent 02, and 1 percentother gases. Therefore at 296°K and 1 atm pressure, there are 0.78 x2.48 x 1024 1.93 x i04 molecules of N2 /cmZ in a 1 kilometer path. FromFig. (2-1) we see that Co -= 0.95 Cs 0 therefore, the "effective' pres-sure of I atm air is apprbKimately 0.4 atm. It follows that (-,4,T) fora 1 km air path at 1 atm is 0.99 x 1.9 x 1024 C0 k. We have used thisrelationship along with the curve of C. p from F! (2-1) to calculate(-,,T) versus wavenumber for 1 km of air. The results are plotted inFig. (2-2).

Also wn for comparison in Fig. (2-2) is a curve corresponding to thecontin um due to the extreme wfngs of the vely strong CO2 lines centeredbetween approximately 2250 cm- and 2390 cm- . This curve is based ondata obtained in'our laboratory and reported previously. 2' 6

-lNear 2400 cm 1 N and CO2 are seen to absorb approximately the same, butabove 2420 cm"4 t~e N2 absorption is much greater. The upper curve repre-sents both the N2 absorption and the CO2 wing absorption. The quantity(-,,&T) can be calculated for other pressures by making use of the know-ledge that it is proportional to the square of the total pressure P. Someadditional CO, absorption not accounted for in Fig. (2-2) occurs because

2-2

12 16 18 12 16 17of weak bands of the C 0 0 and C 0 0 isotopes and a pressure-induced band of the common isotope C1 2 01 6 . (See Reference 6.) N2 0 andH 0 also absorb in the spectral region covered in the figure with thesirongest absor'tion by these two gases occurring above 2500 cm 1 . Thus,the N2 absorption plays its most important role between 2400 cm" 1 and2500 cm"1.

2-3

-265xiO5x' 0 ' * ' ' I

4x10. 2 6

_ \ S,pS x10. 26 C;,P

E

• 2x10"2

ooI x 10"2

0 __j

2400 2500 2600WAVENUMBER (cm-')

I I 'I

4.2 4.1 4.0WAVELENGTH (pom)

FIG. 2-1. Normalized Coefficients of Pressure-Induced N2 AbsorptionPlotted Versus Wavenumber. The upper curve representsCo and is based on pure N2 . The various geometrical

fjrv s correspond to the following samples of pure N,:7, L 36.9 m., p 14.6 atm; x, L = 32.9 m, p - 13.6 atm;0, L 32.9 a, p 21.8 at=. The temperature is 2960K.The lower curve represents C0 and is based on mixturesof approximately 14 atm N2 +o' 8 atm 02 with L = 32.9 m.

2-4

' ' 5 ' I ' I ' '

9

0.10 C02 +N2

N2,

0.05-

C02

2400 2500 2600WAVENUMBER (cm-')

1 ... . . I I

4.2 4.0 3.8WAVELENGTH (pm)

FIG. 2-2. Curves of -. 'ZT of N2 and the Continuum Due to theExtreme Wings of 00 Lines. The curves correspondto a I km path of air at 296°K and 1 atm pressure.

2-5

SECTION 3

ABSORPTION BY LARGE SAMPLES OF COBETWEEN 1100 AND 2000 cm1 2

-1Absorption by CO2 is relatively weak throughout the 1100 - 2000 cm

(9.1 - 5 4m) region and has not been studied as thoroughly as most of theremainder of the spectrum. Figures 3-1, 3-2, and 3-3 show spectral trans-mittance curves for a variety of large samples at room temperature, 296*K.Some of the absorption is due to intrinsic bands, some to pressure-inducedbands, and a small amount to wing absorption. From the transmittancemaximum near 1130 cm" 1 to approximately 900 cm"I, most of the absorptionis due to the bands involved in the well-known laser transitions. Severalmedium and strong bands occur between 2000 and 2400 cm'l; some of them arediscussed in Section 4. Samples B and C represented in Fig. 3-3 areessentially opaque from 2020 cm"I to beyond 2400 cm-1. Data for theportions of the curves in Fig. 3-3 on either side of approximately 1590cm"I were obtained at different times vith different gratings in thespectrometer.

Host of the absorption between 1130 and 1530 cm' is due to the v and 2,2bands. The positions of the centers of these bands have been calculatedfor the four most common isotopes from the energy levels tabulated byStull, Wyatt, and PlasF7 and are listed in Table 3-1. Also listed arebands which have the same change in quantum numbers but arise from thelowest excited vibrational energy level. New notations for the identi-fication of CO2 energy levels and isotopes are dtscribed as footnotes ofTable 3-I; the new notations are used hereafter. The bands listed areforbidden for the symmetric 626 and 636 molecules at low pressures;however, as the pressure increases, the interaction of the ...olccules withtheir neighbors disturbs the symmetry so that the transitions may occur.As discussed in Section 1, the absorption coefficient for these pressure-

3-1

TABLE 3-1

CENTERS OF CO2 BANDS .iaBETWEEN 1100 AND 1450 cm

b (100 0 )C (1000)II (110,l)I-(O0,1)I (ll0)1) 1 1-(010l)I 1

Isotope rlo001 r02 00] (io 0-01 0] [031 0-01 10]

626 1388.20 1285.43 1409.45 1265.11636 1370.11 1265.84 1388.65 1248.03628 1365.89 1259.49 1386.93 1239.29627 1376.03 1272.35 1397.28 1251.98

-1aBand centers are in cm and are based on energy levels tabulatedby Stull, Wyatt, and Plass.7

bThe number in the fi st column refers ju ghe isotopic sp ii.626 corresponds •o 0 C 01, 636 to 0 C 0 , 628 to 0 CL0and 627 to 06C1 01 .

CThe lower energy level is the ground state when only one level is

given. Two different notations are given for the energy levelsinvolved in the transitions. The notation in brackets is widellused. The other notation has certain advantages and is gainingacceptance. S ~nce I 2v 2 for CO2 , the two egergy levels 1000

and 0200 are approximately the same, as are 1001 and 0201, etc.In order to illustrate the relatienship between the old and newnotations, we considor the thrae energy levels given in the oldnotation by 2003, 1 2 03 , and 04 3. Note that quantum numbers v3and tare both fixed and that 2v1 + v, Is constant. In the newnotation the three 1, vels become (203,0)i, (203,0)ig, and (203,0)111.The Roman numeral subscript denotes the order in which the bandsoccur without regard to the assignment of v, or v2 ; I denotes thehighest wavenumber, II the second highest, etc. The first digitin the parenthesis Is the largest value of v, in a group; thesecond is the smallest value of v . The third digit is v3 andthe one following the comma is A.

3-2

I0

0 -4(44'

O'J 3 C

Wx 41

1.U L~-4 0 Aj 4) 0 0

4) 0 U1C C4 .O0Uw

CO. 4I J 0w

*"-4 021. .

-{ W~0. '41s~w ~ 41

a 0

N - -8 '*5 m ('

O E E6 4 wC L

V~ I 0 .4'c

04 04Q

U *G -4

1w S5.- a4 4 ILW 21 0

4 '): tj

u. 00 c is> v C '~0 w a-

> c a Ld 1.

I*.e41 C IChL c

v o

di U i

I0

0

-w

w l C.,o 0 I

SII

" ~~I'dl Ki

I ku,•

0 .- dS

C> 4

0 a . 0%

001 ZI

SI-i

/ .4

S0 eO 4SiI Io 0 O--'

0001 X I

10 0 L --------- I

500D

A50

1400 1500 1600 1700 (cm-") 1800I • I I I I I I I I I I 1 1

2 7.0 6.5 6.0 (microns) 5.5

I-.

50050

1800 1900 2000

WAVENUMBER (cm-')

5.5 WAVELENGTH (g rm) 5.0

FIG. 3-3. Spectral Tranamittance Curves from 1400 cm to 2020 cm-.The spectral slitvidth varies from approx.imiely 0.55 cm'lat 1500 cm"I to 1.0 cm-1 at 2000 cm L. The samples arepure CO2 and the parameters are:

p L

(ml i-atm imeters)

A 14.6 32.95 21.3 32.9C 2.3-5 933D 1.8' 1067

3-5

I2induced ,ands is proportional to p, and (-4,T) is proportional to p 2L.Because of *he asymmetry of the 628 and 627 molecules, they can absorbat low pressures, and the dependence of their absorption on pressureand absorber thickness is similar to that of other intrinsic absorptionbands. U'e expect the 628 bands to be approximately 10 times as strongas the corresponding 627 bands since the asymnmetry of the former moleculeis twice as much and the abundance is 5 times as great aF that of thelatter. Any absorption by one of these bands of the 636 molecule wouldbe pressure-indu ced, and the ratio of its strength to that of the corres-p03ding 69 band is probably about 1.1:99, the ratio of the abundances ofCl andC

The strength of each difference band with (010,1)I as the lower state is

expected to be approximately 4 percent as great as the associated funda-mental or overtone band which arises from the same change in quantumnumbers. This relationship is based on the assumption that the strengthsare proportional to the populations of the lower energy levels. At roomtemperature The population of the (010,I)I level is approximately 4% ofthe (000,0) level.

On the basis of the above discussion, we expect most of the pressure-

induced absorption to result from the (100,0) 1 .nd (100,0) bands of the626 molecule and the major portion of the intrinsic absorp ion to arisefrom the same bands of the 628 molecule. Furthermore, we expect thefraction of the absorption due to the pressure-induced bands to increasewith inLreasing pressure. Figures 3-1 and 3-2 confirm these expectations.Sample 1, which has a relatively low pressurI and long path, shows twodistinct band centers near 1260 and 1366 cm- , the calculated positionsof the 628 band centers given in Table 3-1. Sample 3 which is at 3.70atm shows transmittance minima near 1285 and 1388 cro"i, the calculatedband centers for the 62C molecule. The 628 bands have been oLserved withrelatively coarse resolution by Eggers and Arends from samples enrichedin 0 8. The Positions of the band centers determined by these workersagree with ours to less than the experimental uncertainties. The pressure-

induced 626 bands have been observed by Welsh, Crawford, and Locke. 1 0

The major errors result trom errors in placing the zero-absorptance curves

on the sample curves and from absorption by H2 0 which appeared as animpurity in the CO9 in spite of efforts to keep it dry. Much of the H2 0abscrption was accounted for before the curves were replotted by comparingthe original sample curves with spectral curves of H O with N2 as abroadening gas. Spectral regions between 1100 and 1400 cm"I where H20absorption was partially accounted for are indicated by arrowf in Fig. 3-1.The H120 absorption was most serious between 1500 and 1800 cm" becausethe H 0 lines are stronger and the CO2 absorption is weaker. The largeamouni of gas required for the long cell could not be purified as

effectively as the smaller amount required to fill the short cell.

3.-6

Furthermore, the absorption by H 0 is approximately proportional to thepressure of a mixture of constant composition while the pressure-inducedCO2 absorption is proportional to the square of the pressure. Therefore,the H 0 absorption was less important for the high-pressure samples inthe short cell than for the lower-pressure samples in the long cell.

In regions where there is intrinsic absorption due to local lines andpressure-induced absorption, but no wing absorption, Eq. (1-4a) reduces to

K = K(local) + C K(lOcal) + C° p*.s,p ssp .

Since CS P is proportional to p*, and K(local) is not, we have been ableto determine each of these quantities from spectral curves of two or moresamples at quite different pressures. At pressures less than a few atm,K(local) varies rapidly with wavenumber because of the line structure.Since the spectral slitwidth of the spectrometer was greater than theline widths at low pressures, K(local), and thus K, could not be measureddirectly at a given wavenumber. We were not interested in retaining allof the structure in the spectra, but in the values of the constantsaveraged over intervals approximately 5-10 cm" 1 wide. The strength,S = fkdv, of ar intrinsic absorption line is independent of pressure;therefore, the average of K(local), which may result from several neigh-boring lines, is essentially independent of pressure if the interval is

several line widths wide and contains the centers of all the lines con-

tributing to the absorption.

If we consider only the true transmittance of the local lines,

(-I/u)0--TI(local) is equivalent to K(local). (The bar represents anaverage over wavenumber, and the prime denotes true transmittance thatwould be observed with infinite resolving power.) However, we do not

read -T-T' from a spectral curve, but a slightly different quantity,since we average the observed transmittance curve over a 5-10 cm" 1

interval. The difference between these two quantities decreases withwider lines and thus with increasing pressure; it also decreases with

i creasins average transmittance by the lines. We first assumed thatT•T '(local)+ &Y'(continuum) and that (-i/u),T(local) =

0K(local) in order to calculate CS P and an approximate value of 7(local)from data for two samples at different pressures. The approximate widthsand spacings of the lines are known, so that by applying simple, band-model theory we were able to calculate a corrected 7(local) and to adjustthe values of T, which were then used to determine a corrected value ofco,,. In most cases, there was less than 10% difference between(- u)iwT(local) and (-l/u))T(local), so that the remaining errordue to the averaging was probably not more than a few percent afterthe correction was made.

3-7

L

Average values of the quantities were calculated throughout the 1100 -

1500 cm"I region for intervals 10 cm- 1 wide, except near the points ofinflection where narrower intervals were used. The results are presentedin Fig. 3-4. The uncertainty in the plotted values of K(local) and C0

is greater in regions where one of the quantities is changing rapidly.s'For example, there is extra uncertainty in the heights and positions ofthe peaks of the upper curve near 1275 cm- and 1385 cm- where the lowercurve is steep.

The general shape of the upper curve in Fig. 3-4 is similar to that ex-pected for the (100,0)I and (100,0)11 intrinsic bands of the 628 molecule.The P- and R-branches are evident, and the band centers agree well withthe calculated values given in Table 3-1. Because of the relativelycoarse spectral resolution used in the calculations, the line structureapparent in Fig. 3-I has been smoothed. Any obvious contribution by thesame bands of the 627 isotope or by the difference bands of the 626 isotopeis obscured by the coarse resolution and the scatter in the experimentalpoints. However, these bands undoubtedly contribute to the absorption inaccordance with the discussion near the beginning of this section. Theintegrated absorption coefficient for the infrinsic absorption by thelocal lines expressed in molecules 1 cm2 cm is 43.9 x 10-24 from 1200to 1310 cm-1 , and 64.5 x 10-24 from 1310 to 1420 cm" 1 . The estimateduncertainty is ± 10 percent.

The lower panel of Fig. 3-4 shows a spectral curve of the normalizedabsorptioa coefficient for pressure-induced absorption. The absorption 1peaks agree well with the calculated band centers 1285.43 and 1388.20 cmgiven in Table 3-1 for the pressure-induced bands of the 626 isotope. Theabsorption by the 627 and 628 molecules probably have a pressure-inducedcomponent, but it is apparently much smaller than that by the 626 molecule.The two bands are seen to overlap and to produce measurable absorptionover a wider spectral interval than that covered by the intrinsic bandsillustrated in the upper panel. Except near the peak at 1388 cm'1, thedata points corresponding to different samples agree within the expectedexperimental error. Below approximately 1130 cm 1,. the continuum isinfluencid by the wings of the lines centered from about 1000 cm" 1 to1100 cm-

-1 2 -1 -1The integrals of the normalized coefficients in molecules cm atm cmare 67.3 x 10-24 from 1130 to 1330 cm" 1 and 101 x 10-24 from 1330 to1500 cm' 1 ; the estimated uncertainties are t 8 percent. We note that theratio of the integrals is approximately 3 to 2, the same as that for thecorresponding intrinsic bands.

0

Curves of (local) and Cs are shown in Fig. 3-5 for the region between1490 and i35 cm" . Since not all the continuum is due to pressure-Inducedabsorption, we have omitted the subscript p from C0 in Fig. 3-5. Thesymbol Cs is written without a p or w, which denote pressure-induced or

S

3-8

Ln0

00

.4 0.

C--

A. 0 .

00'

ui - o4 -

z OD- -4 0

-~0z 0

0. I

0 OA

to

00 0 A

00W iS0l3IW 0 00~ 0jj .Of~IW 0LO ty 0 CA

3-9

000

u m~

rrx x

x 0*0. CC

C~C 0c -

jý $4 t 0

0t C04I-

'-4 0 - 4 to0

I.' w j C:

4Nýd~ LA

ca0Uz4

ula~~C. 94nU0'..itWDI94nol

(I Pa )o>

3-10I

. dkf

wing absorption, respectively, when the type of continuum is now known,or, in some cases, when it is clear whic• is being considered. The curvein the lower panel from 1490 to 1590 cm- is base, primarily on Samples Aand D whose transmittance curves are shown in Fig. 3-3. Samples B and Cform the bases for the curves in Fig. 3-5 above 1585 cm" 1 . The mostserious errors between 1490 and 1850 cm- 1 arise from absorption by theH2 0 impurity. Near the H20 band center at approximately 1585 cm- , theH20 absorption is less than on either side. At this point the estimatederror in C0 is between 0.1 and 0.2 x 10-26 molecules"< cm2 atm- 1 . Atother points between 1490 and 1585 cm-1, the errors may be twice this large.

-INear 1788 cm the H 0 absorption is still buthersume, but it is much lessthan at most places getween there and 1555 zm-1 . The estimated absorptanceby the H20 impurity in Sample B was approximnately 2 percent, while that bythe CO 2 was 8 percent. We also atLempted to account for the H20 absorp-tion in Samples B and C at threE other points between 1788 and 1590 cW'l.These points were chosen between H20 lin-ý. where the H20 absorption wasless than at any other point within several zm-l. The estimated correctionfor H20 absorptance was less than 5 jercent at these points for B, the21.3 atm sample containeJ in -he shortei at.sorption cell. The correctionfor H20 absorptance was as mich as 10 percent to .5 percent at these samepoints for Sample C. At 1783 cm" 1 the estimated uncertainty in the plotof Co is 0.05 x 10-26 moleciles" 1 cm2 atm'1 ; at other wavenumbers between1788 and 1585 cm- 1 it may 4e double this amount.

The source of the intrinsic absorotion coefficient, apparently due tolocal lijes since it is independrnt of pressure, is not understood. At1585 cm- we observed approximately 78.5 percent transmittance for SampleC and 67.8 percent transmittance for Sample B after accounting for theH 0 absorption. If we had obserý'ed the same transmittances for B and8X.3 percent for Sample C, we would have concluded that there was nointrinsic absorption. The difference 78.5 percent versus 89.3 percentis greater than our estimated error; therefore, we conclude that thereis some intrinsic absorption. Spectral curves in this region show nostructure other than that due to HO lines. If a very weak C09 bandoccurred in this region, we would expect to see some structure aue to thevibration-rotation lines, although the amount of intrinsic absorption issmall enough that we cannot rule out this possibility. The apparentintrinsic absorption may result from an impurity, other than H20, whichhas no line structure that would be observed with the spectral resolutionused in obtaining the data. It is also possible, although difficult tobelieve, that errors in correcting for the H20 absorption account forthe apparent irtrinsic absorption.

We feel that the upper curve in Fig. 3-5 represents maxi.•um values forK(local) due to intrinsic absorption by pure CO2 . The correct valuesmay be much less if errors were introduced by impurities. Above

3-11

-1approximately 1780 cm , the increase in K(local) is due to very weaklines which are a part of the band seen near 1900 cm'l in Fig. 3-3.Between 1585 and 1490 cm" 1 the pressure-induced absorption is much greaterthan between 1585 and 1800 cm-1; it is also enough greater than any intrin-sic absorption that none of the latter was detected.

3-12

SECTION 4

ABSORPTION COEFFICIENTS AND INTEGRATED ABSORPTANCEFOR CO2 BETWEEN 1885 cm"I AND 2132 cm- 1

The spectral region from 1885 to 2132 cm" 1 contains several CO2 bands whosestrengths have not been determined, although several of the band centersand line positions have been measured, or can be calculated accurately, fromavailable data. (See Ref. 6 for spectral curves and a table of band centers.)The first part of this section deals with experimental data obtained primarilyto determine the strengths of the bands and of single branches of some of thebands. Particular attention was given to the Q-branches. The samples containCO2 with He added in order to broaden the lines and remove much of the struc-ture so that the observed transmittance T was approximately equal to the truetransmittance T'. Helium, rather then N2, was used as a broadening gas sinceit is difficult to obtain N which does not contain enough CO impurity toabsorb in this region. Furihermore, the extreme wings of He-broadened linesare weaker and thus produce less continuum absorption. Since the absorptioncoefficient varies over a wide range of values, different absorber thick-nesses were used in the different narrow intervals in order to produce trans-mittances which could be measured accurately.

Throughout most of the region, (-l/u)j.T changes slowly with increasingpressure in the range of pressures used, and (-l/u)f4Tdv is essentiallyindependent of pressure if the integration is performed over an intervalincluding several lines. This result implies that most of the absorptionwithin the narrow interval is due tc lines centered within the interval.In a few regions, (-1/u)f4Tdv over a narrow interval increases significantlywith increasing pressure. These regions occur where the absorption is veryweak and a large fraction of it apparently is due to continuum absorptionby the wings of stronger lines which are centered several line widths awayfrom the points of observation. In these cases, (-1/u)f1..T(v)dv does nottruly represent the integral of the absorption coefficient of the linescentered in the interval. Examples of such regions are at the absorptionminima near Q-branches, and particularly near 1945 cm" 1 , 1990 cm"' and from2135 cm" 1 to 2200 cm"I.

-lThe strong absorption feature near 2077 cm can be recorded easily in atmos-pheric spectra obtained from an aircraft with the sun as a source. Thespectral curves shown in Figs. 4-12 to 4-15 provide laboratory data on knownsamples at 297*K for comparison with field data. Table 4-2 provides theparameters of the samples and values of fA(v)dv over several narrow spectralintervals. In order to better simulate air, we added a mixture of 80% N2and 20% 02 as a broadening gas.

2

.J0

C144

I C 0 0 0 0(n El C1 CNN. C.) 0:r 0:: or0 0 0:

1-4 - 4 .- 4

CA WJ EE Z)-4W-4

Co :3 CN J!.N C4N C4 C'J4 CN -

w NN C'C to 4co) Q) N nL n LI 00c 0 00 '0ao ýo r

-4 '-, r-r 1 N , D% oz

0 E)

0 . . . . .. . .'

ý4 -4 0

o- . . . . . . 41 0*L"00 4-4P4 -4 ~ O 1 00C 0 a ;C ;C

EL.~'a ~ 0 O - - - vd~ 1~094 3t~-N CN N ~ 'N N N *

0

.61

- ~ ~ ~ ~ ~ ~ ~ ~ ~ 4 S a I a a a a a

0, t- .1 C1 eq .

c) '1fl'ý V)O- -'r.4Jf %nC1 I0-~~~e 2Z''O OO. O 0 .. 'N U

I.~& _ _ _ _ _ _ _ _ _ _ _

u -2

r-4

0

0 '0

N>

P-0

i o 1 0

CJJ

I ww

4.34

I -x -23 r---T_,

, 23

_•I..

E 500

00

o I , l...1930 1932 1934 1936

WAVENUMBER Icm-')! I

5.18 5.17WAVELENGTH (mm)

FIM. 4-2. Spectral Curve of (--.,,T)/u for CO2 from 1929 to1937 cm"1 .

4-4

0.4 10 -23

-... 0.3xlO2

EU

U4 0.2xlO" 2 3

-! -E

0.1 x i0 -2 3

0 r 1 , I

1940 1945WAVENUMBER (cm-')I ' !

5.16 5.14WAVELENGTH (pam)

FIG. 4-3. Spectral Curve of (-4,iT) 1u for cO from1936 to 1948 cm '1. 2

4-5

oy EC1

00N-

0

00

o 44'>~ 0

-Iii4

0>

4-64

-44oi

0

L CJ

ict 0

oo00

0 u

0 .0 0

4-74

S I ' I ; I '

100 10- 2 3

EQ

-23o 50x10

E

0 I , I , I • i j

2074 2076 2078 2080 2082WAVENUMBER (cm-')

I I

4.82 4.81WAVELENGTH (,im)

FIG. 4-6. Spectral Curve of (-,4,-T)/u for CO2 from2073 to 2082 cm'".

4-8

00

0*i0

NE

E"4

a: M

00 LJ

LO z

C ~ ~ C. (N 0 u

0

0 '-0

(awo iSIfslnoolow) -

4-9

20x10-23

N

E

" 0 IOx10"23

E

0 I , , I I

2090 2095WAVENUMBER (cm-')

IF-4.78 4.77

WAVELENGTH (AM)

FIG. 4-8. Spectral Curve of (-/ 4vT)/u for CO2 from 2090to 2097 cm'.2

4-10

00

0"

02 z

>_ 0

0 -4

0 ý4 0

U.6'CLJ Q 0

o 0

x 9~

4-11

1 XlO-231 1

Et

-u 5x10- 2 30

E

00

0 I_ . _.I

2126 2128 2130 2132

WAVENUMBER (cm-')

4.70 4.69WAVELENGTH (Pm)

FIG. 4-10 Spectral Curve of (-/AT)/u for CO2 from2126 to 2132 cm-

4-12

OL

E

0 E- CO 0&%tN (D M

mlz

(OZ_ww

> Ln

00rt) 0 0

o 0 0-. -- 6

( awo I-salnOalOW) n _iki

FIG. 4-11. Spectral Curve of (-4..T)/u for CO(2 from 2130 to2200 cm'-.

4-13

100

IOO

0O

- 50× 3

0

o_ 50 i

2065 2070 2075 2080 2085 2090 2095 2100WAVENUMBER (cm-')

I * I * I

4.84 4.R2 4.80 4.78WAVELENGTH (irm)

FI(,. 4-12. Spectral Curves of Transmittance for CO2 from 2065 to

2100 cm" 1 for Samples 1, 2, and 3. See Table 4-2 forsample parameters awid fA(v)dv.

4-14

100

0 50x

0

100 K5/7

0

-0 /,I

0 * . .. * t I ± .. * J * I . . .A , . . . .1 . .I • * ,

2065 '"070 2075 2080 2085 2090 2095 2100

WAVENUMBER (cm-

I I I I - -' -

4.84 4.82 4.80 4.78

WAVELENGTH (pm)

FIG. 4-13. Spectral Curves of Transmittance tor 002 from 2065 to2100 cm"1 for Samples 4, 5, 6, and 7.

4-15

S

100;Uj

0

0•

_ 5 0HJ

-4

0 , ' i I I a , i t a a , I .i I a 4 ,

100,

I7" -10 *H 4

4

.4

0 I A-1 I A A A A A I A * ~ A A AA I A A -

2065 2070 2075 2080 2085 2090 2095 2100WAVENUMBER (ckm,)

4.84 4.82 4.80 4.78WAVELENGTH (jim)

FIG. 4-.4. Spectral Curves of Transmittance for 00 fron 2065 to2100 C"01 for Samples 8, 9, 10, and I.

4"1t

VX 100

ý12

50

0/15

I-

0 J * * * i * a a I a a a a I I I

100 - 4KN, 1-YW •..2---•-

oI

2065 2070 2075 2080WAVENUMBER (cm-1)

4.84 4.82 4.80WAVELENGTH (Pml

FIG. 4•-15. Spectral Curves of Transmittancefor CO 2 from 2065 to .2183 :--' *

for 54s•ples 1-1 and '15.

,4-17

- -. - - - - - - -

*~ . . . . . . ...

4-18

SECTION 5

TRANSMITTANCE DATA STORED ON MAGNETIC TAPE

Under a previous contract, NOnr 3560(00), our laboratory published aseries of reports dealing with the absorption of infrared radiation byCO2 and H2 0. Each report is limited to a given spectral interval andcontains detailed information on the absorption by several samplescovering a wide range of pressures and path lengths. Tables of theintegrated absorptance, fA(v)dv, are included in all of this series ofreports; some also include extensive tables of transmittance. As aresult of requests from several workers in the field, we have compiledthe transmittance tables on magnetic tapes which will be availablethrough Air Force Cambridge Research Laboratories. The transmittanceis tabulated versus wavenumber, and enough values are given so that theoriginal spectral curves can be reconstructed without loss of structureby plotting the tabulated values and joining them with straight lines.

Table 5-1 summarizes the data saved on magnetic tape. The report numberis the publication number assigned to the report by Aeronutronic. Theabsorbing gas and the spectral interval covered by the report are includedin the table. A given file on the magnetic tape is devoted to the datafrom a single report.

5-1

TABLE 5-1

SUMMARY OF DATA SAVED ON MAGNETIC TAPE

Report Absorbing Spectral Interval

Number Gas -1cm 4m

U-3704 H 20 5,045 - 14,485 0.69 - 1.98U-3202 H120 2,800 - 4,500 2.22 - 3.57

U-3200 CO2 81000 - 10,000 1.00 - 1.25

U-3930 CO 2 7,125 - 8,000 1.25 - 1.40U-3127 CO2 6,600 - 7,125 1.40 - 1.52U-3201 CO2 5,400 - 6,600 1.52 - 1.85

U-2955 CO2 4,500 - 5,400 1.85 - 2.22U-4132 CO11 3100 - 4,100 2.44 - 3.22

U-3857 CO2 1,800 - 2,850 3.5 - 5.6

5-2

SECTION 6

REFERENCES

1. D. E. Burch, D. A. Gryvnak, and R. R. Patty, "Absorption of InfraredRadiation by CO, and H20. 1. Experimental Techniques0" j. Opt. Soc.Am. 57. 885 (196)).

2. D. E. Burch, D. A. Gryvnak, R. R. Patty, and Charlotte Bartky, TheShapes of Collision-Broadened CO Absorption Lines, AeronutronicReport U-3203, Contract NOnr 3560(00), 31 August 1968.

3. B. H. Winters, S. Silverman, and W. S. Benedict, J. Quant. SpectryRadiative Transfer 4, 527 (1964).

4. M. M. Shapiro and H. P. Gush, Canad. J. Phys. 44, 949 (1966).

5. C. B. Farmer and .. T. Houghton, Nature 209, pp 1341 and 5030 (1966).

6. D. A. Gryvnak, R. R. Patty, D. E. Burch, and E. E. Miller, Absorption

by C Between 1800 and 2850 cmI, Aeronutronic Repoit U-3857,Contract NOn4 3560(00), 15 December 1966.

7. V. R. Stull, P. J. Wyatt, and C. N. Plass, The Infrared Absorption ofCarbon Dioxide, Space Systems Div., Air Force Systems Coimnand, ReportSSD-TDR-62-127, Vol. III, Contract AF04(695)-96, 31 January 1963.

8. D. A. Gryvnak, 1). E. Burch, Experimental Investigation of the Absorp-tion by Various Gases of Planetary Interest, Aeronutronic ReportU-4367, Contract 951712, JPL, 29 April 1968.

9. D. F. Eggers, Jr. and C. B. Arends, J. Chem. Phys. 27, 1405 (1957).

10. H. L. Welshl M. F. Crawford, and J. L. Locke, Phys. Rev. 76, 580(1949).

6-I

UNCLASSIFIED

DOCUMENT CONTROL DATA.- R & DSe-r.1y teassItirat!or of tMWe hodv of abltra.-t and Indexing an~notation niu.I he etrd hrpn ther -,al;I r~,!..,

Philo-Fod CoporaionUnclassifiedAeroutrnic iviion2b GROUP

Newport Beach, California -)2663

INVESTIGATION OF THE ABSORPTION OF INFRARED RADIATION BY ATMOSPHERIC GASES

4 [Sk5C RI rT IVYE NO0T ES (TVp@ of report and Inelusive delta)

Scientific. Interim.5A U ,O R S I ( Firs? name. middle In~iaj, Idol name)

Darrell E. BurchDavid A. GryvnakJoh D. Pembrook

6 REPORT ::ATIE 70. TOTAL NO OF PAGES b6N ~June 1970 50 10

@a C SN IACI OR GRANT No 9a. ORIGINATOR'S REPORT NuMBERISI

F19628-69-C-0263 ARPA Order No. 1366 U42h, .- 'OJECT NO

5130 Semi-Annual Technical Report No.1

96ý OTHER REPORT NO(Ie (Any other numbers rhal -.

DoD El ement 62301lD this report)

!DoD Subelement n/a7 5STRIR.JTION STATEMAENT

This document has been approved for public release and sale; its distribution is

unl imited.

*S,,PPLEmENrAR, NOTES 12 SPONSORING t.41LITAR, ACTIVI,T

This research was supported by the Air Force Cambridge Research LaboratoriL!

Advanced Research Projects Agency. L. G. Hanscom Field (~0Bedford, Massachusetts 01730

7, -A S'N

From spectral transmittance curves of very large samples of CO2 we have determined

coefficients for intrinsic absorption and pressure-induced absotrption from approxi-

mately 1130 cm-1 to 1835 cm-1. Most of the pressure-induced absorption arises iror,

the forbidden vi and 2v,9 bands of the 016C12O18 molecule. Curves of the absorption

coefficients for samples of CO2 pressurized to approximately 10 atm, or more, with

He are presented throughout most of the region from 1885 cm-1 to 2132 cm-1. Theseý

curves and a table of the integrated absorption coefficient are adequate to detLT~Ih~nt

the strengths of most of the bands of significance in this region. Spectral trans-.

mittance curves of CO 2 absorption between approximately 2065 cm-1 and 2100 cni' are

presented for a variety of samples. Curves of the pressure-induced absorption by N.,are presented for the 2400-2650 cm-1 region.

D ,NO. 6,47 UNLASF1 El)Secrt

UNCLASSIFIEDSecurity Classification

Absorption

Atmospheric Transmission

Pressure- Induced Absorption

4I

UNCLASSI F IEP

KE OD