the absorption spectrum of cf2

3 downloads 0 Views 1MB Size Report
1 500 I\ which have been assigned tentatively to the CI;? molecule. i\. ...... triatomic molec~~le as a function of the bending freque~lc),. Use of their graph ..... Simons' 4al orbital has been relabeled Gal, in accord with the diagram given 111. .... WALSH, .\. 1). 1933. J. Cherli. Soc. 226G. \VIIITE, J. Y. 1942. J. Opt. Soc. r1m.
Can. J. Phys. Downloaded from www.nrcresearchpress.com by Guangzhou Jinan University on 06/03/13 For personal use only.

THE ABSORPTION SPECTRUM OF CF2'

'l'he absorption spectrum of Cl:? in the 3 500 region has been photogrnphcd a t high dispcrsion, and the rotational structure of a lumber of bands has been analyzed. 'fhe anall,sis of the ell-resolved subbands cstablishcs that tliesc are perpendicular- rather than parallel-type bands, as previously assig~led.I'~~rtlier analysis sho\vs that the upper and lower electronic states are oi lH1 and ';Il symmetries respectivell-, correspondi~lgto a trinsition moment- that is perpendicular to the plane of the molecule. In the upper electronic state, 1:0(Cl:j = 1.3.2X and L 1;CF = 1%2.3",\\.bile ill the ground state, ~o(C1:) = 1.300 A and L 1;C1;= 104.94". . i n investigation ol thc vibrational structure of the band systeln has shown that the vibratiol~alnumbering in a?' I I I L I S ~be increased by one illlit fronl earlier assignments, thus placing the 000-000 band near 2 687 A (37 320 cm-I). A search~betweer~ 1 300 and 8 500 .& showed two new band systems near 1 350 and 1 500 \I which have been assigned tentatively to the CI;? molecule. i\.

IS'I'I?ODUC 1 I O S

In 1950 Iienbates\\-arlu first ol~servedan emission spectrunl l)et\\een 2 399 .lnd 3 429 ,\ \\rhich he attributed to the C17?molec~~le. On the basis of resolved I< structure he coilcluded that the emitter \\las nonline,lr in both states and that the trailsitioil moment \ \ a s parallel to the neai-s) mmetric top axis. IIe assigned the 2 350.6 -fi band a s the 000-000 tr,li~sition and interpreted the cxteilsive vibrational structure in terills of the s).mmetric stretching freclueilcies (vll, vltl) and the bending frequencies (vL1, ~2").Shortly there,lfter Laird, .\nclre\\s, and Barro\\~(1950) observed the systein in absorption, thus establishing that the transition involved the ground state. T h e absorption spectrum \ \ a s SLIIJsequently obtained froin flash photolysis of CF2Br2(JIann and Tllr~lsll1960), froin a lo\\,-temperature matrix \\ihich contained the products of a microwave e discharge through C I F s (Bass and AIann 1962), and from a p ~ ~ l sdischarge through C,F1! (Thrush and Zuolenik 1963). These authors conclucled that all be interpreted in terms of the bencling frequencies absorption bands co~~lcl only and that the 000-000 band was a t 2 6.50 (or possibly longer \I avelengths). AIann and Thrush noted that the absence of progressions in the stretching frequencies implied that the C-F bond length is approximately the sanle in both states. The infrared spectrum of CF2 in an inert matrix \\as observed by ;\Iilligan, hIann, Jacox, and AIitsch (1964) follo\\~ingits formation by the photolysis of CFzK2. They observed bands a t 668, 1 102, and 1 222 cm-I \\.hich they attributed to 12CF2and bands a t 1 073 and 1 191 cm-I \\rhi:h the)- attributed to 13CF2,present in natural abunclance. I-Ierr and I'imentel (19G3) observed a 10 cm-I \\-it11 a rapid-scan infrared spectrometer by flash band a t 1 110 'Issued as N.R.C. Ko. 9343. ?N.R.C. Postdoctorate Fellow, 1963-67. Canadian Journal of PI~ysics.Volume '15 (1067)

a,

L"

Can. J. Phys. Downloaded from www.nrcresearchpress.com by Guangzhou Jinan University on 06/03/13 For personal use only.

photol) sis of CI:zi\;2 i l l the gas phase. The?. attributed this band to the as)mmetric stretching f r e q ~ ~ e n cofy CF2. Overend (190(i, unpublishecl) obtained the infrared spectrulll under higher resolution u i t h a steadystate absorption S) stem, I)ut his analysis is not yet available. 'The intense al,sorption spectrunl near 2 300 :. has offered a iollvenient ineans of learniilg more a l ~ o u tthe chemical behavior of CIT2,especiall). ~ v i t hregard to its role in fluorocarbon reactions. 'The CF2 radical has l~eensho11.n to lje reniarkal~lystal~leand fair11 nonreactive (see, for example, .\lastrangelo 1962; .\Ial~lerlO(j:3 ; Ilall)) 1061; llitsch 1961, 1965; Schoen and .\[:unn 1964; and EIeicl;len, Cohen, zuncl Saunders ll)(i5). Xdclitional ol,servations have demonstratecl the ease 11 ith ~ ~ h i cthe h species is fornled, have siiggested that the lo~veststal)le state is singlet, and have given nleasurements of the ioilization potential and heat of forination (see, for example, lIargrave and Mlieland 1!).j3; .\largrave 1961; Kelson and Icuebler 1962; Gozzo and Patrick 1%; I; Eel\\ arcls and Small 1961 ; Hol~rockand Iiiser 196-4; Fisher, I-Iomer, and 1,ossing 1'3(i5; llodica and LeGraff 1965; lIodica 1966; and I'ottie 1963). T h e mechanisn~of formation and recoinbination of CF, has been discussed in terms of its moleculal- orbitals by Sin~ons(1965a,b) and Sin~onsand Yarn.ood (1960, 1961). Ilespite the interest in this molecule and the fact that it is easily fornled, the molecular structure and symmetry have been estal~lishedonl). recent11 by I,relinlinary reports of the micro\~~ave spectruiu (I'o\\.ell and Lide 1966) and the rotational anal~.sisof the band a t 2 550 A ( I l a t h e ~ \ - s1'366). 'The present publication offers a more detailed anall~sis of the electronic absorption s~,ectrunlof CFs. B. ESPERli\/IESI'XL 7'11 o methods have I)een used to produce C F 2 for the present studies of its absorption spectrum: (1) a pulse-discllarge technique and ( 2 ) a coi~tii~~ioiis-flo~\technique. The pulse technique gave higher concentrations of CI;? and \\.as useel to search for ne\v electronic aljsorption spectra (Section E ) and to obtain d a t a for a viljrational analysis of the '2 300 systein (Section C). The continuousflo~\-techi~iqueuras more convenient to operate and gave C F z \\-hich could lje effectively cooled. Consequently, this teclunique 11 as most useful in obtaining high-resolution spectra of the 2 500 I% systen~for rotational analysis (Section D). In the first techniclue, the C F ? \\.as produced by a pulse discharge through C 2 F i (Columbia Organic Chemical Co.) or C3l:s (lIatheson of Canada, Ltd.) 1vhic11 \\-as contained in an absorption cell 120 till long and 5 c n ~in diaineter. This pulse tecl~nic~ue* is siillilar to the one used by 'Thrush and Z\volenili (1963) and is idelltical \\-it11 that used in flash photoI~.sis,\\-it11 the exception that the initial capacitor discharge is through the absorption cell (i.e. through the parent co~llpound)rather than through a photolysis lamp. Transient species can be observed, as \\-it11 flash photol>sis, by the use of a Lyman discharge after the initial flash. 'The half \vidth of the initial d i s k w e \\-as about '2 psec and that

"A Inore detailed dcscrip~ionof the apparat~lswill be given by IIerzberg and Shoosmith (to be published) in connection \\,ith their studies of the absorption spectra of ~llolecularions. A similar technique has been described briefly by Nelson and R a ~ u s a y(1956).

Can. J. Phys. Downloaded from www.nrcresearchpress.com by Guangzhou Jinan University on 06/03/13 For personal use only.

lvf.%TI-II.3\VS: ABSORPTION SPECTRUM O F CI;?

t

2357

ol the Lym,\n discharge was about 3 psec. The tiliie delay bet\veen the t\vo discharges could be adjusted to uithin about f l psec. As noted by Thrush and Z\\olenil; (IOG3), there \\-as an optimum input energy for tlie initial discharge tlirougli a fixed quantity of parent compound I\-hicli resulted in remarkably pure CI;:! al~sorption.Ileparture from these conditions reduced (or eliminated) thc c o i ~ c e n t r ~ ~ t of i o nC F 2 and usually ~.ieldeduildesired spectra, for example, C?, CI;, or n continuum. When tlie absorption cell was filled to a pressure of Y'I'orr 11 ith C1F1, tlie maximum al~sorptionin the 2 500 A system 1172~sobtained 113 disrli:\rging 0.025 p F a t 20 kV through the gas. Although the C F 2 absorption I\-ns quite intense for any delay time I)et\veen 0 and 2 000 psec, tlie maximum intcnsit~\\as reached after about 400 psec. The absorption cell could 1)e tittecl \\it11 n mirror systein (\Vhite 1942) I\-hich allo\\-ed up to 36 traversals through a 1-111 portion of the cell (a path length a t least one hundred times . system). The spectra \\.ere photograpl~ecl that required to detect tlie 2 500 & in the first, second, or third order of a 3-111 vacuum spectrograph. 111 the continuous-flow techniclue, the CF1 was produced in a 2 430-JIIIz discliar~cthrough a mixture of helium (or argon) and C2F3Cl("Genetron 265", (;encral Chemical 13ivisioi1 of Allied Chemical Corporation) which was pumpecl r,~picll>-from the discharge through a 1-111 absorptioii cell by a 25 l/scc mccllanical pump. This apparatus has been described by ll>ouglas and Jones (IYGG), n i t h the exceptions that, in tlie present experiments, there \\-as no visil~lcafterglo\\ and standard absorption techniques were used. A 900-W, high-pressure xenon lainp (Osram) furnished the continuum and a nlultipletraversal mirror system alloned up to 36 traversals through the cell. The cell \\.as fitted 11it11 an insulated jacket which could be cooled with liquid nitrogen (cell teinperature a t - 178 OC) or wit11 cold nitrogen generated froin the liquid (cell temperature about - 100 OC). Cooling tlie cell in this manner effectively reduced the rotational and vibrational teillperatures of the CF:! and simplified the spectrum considerably. These spectra were photographed on Icodak 103a-0 or IIa-0 plates in the third order of a 35-ft Eagle spectrograph* a t a dispersion of about 0.3 ,4/111m. 'The plates \\-ere measured on a photoelectric coinparator of the Toinkins-Fred t~ pe (1951), using iron lines froin a hollow cathode discharge as a reference spectr~~m The . relative error of measuring unblended lines \\-as about f 0.003 cm-I and the absolute error was a b o ~ ~ &0.05 t cin-l. However, the measurements used in the analysis iiiay show considerably larger errors since mangr of tlie lines are blended. Another source of errors arose from the necessity of combining measureiiiei~ts from two plates, obtained with different concentrations of CF?, in order to coinpletely analyze a single band. C . I~IBRA'I'IONALANALYSIS, 2 500 A SYSTEM

Figure 1 displays two spectrograms of the 2 505) 2% system obtained \\-it11 a 6 4 pulse discharge through C2F.1 and a single pass through the absorption cell. "'I'his vacuum spectrograph (Douglas and Potter 1062) was evacuated as a matter of convenience only when these spectra \\:ere photographed.

Can. J. Phys. Downloaded from www.nrcresearchpress.com by Guangzhou Jinan University on 06/03/13 For personal use only.

Deslandres table of the 2 500 .& s),stem of C F 2

TABLE 1

\ vzff . .-

0

1 2

. ... A

3 4

5

6 ~. --

7 ------

8

Can. J. Phys. Downloaded from www.nrcresearchpress.com by Guangzhou Jinan University on 06/03/13 For personal use only.

Can. J. Phys. Downloaded from www.nrcresearchpress.com by Guangzhou Jinan University on 06/03/13 For personal use only.

Can. J. Phys. Downloaded from www.nrcresearchpress.com by Guangzhou Jinan University on 06/03/13 For personal use only.

CU

If)

o a

-

-cd rc) If)

If)

'3

.-

2

-

-2I.

0

3 *

+ *

-

.5

F 2

c
. S . and I?AAIS.\Y, 13. A. 1956. 1. Chelii. l'hys. 25, 372.

01c.1, T. and h l o ~ < ~ 1 s' .o , 1962. J . Mol. spectry. 8, 9. P a ~ , n v sJ, . and 1