The Absorption Spectrum of Chlorine Fluoride in the Region λλ4650-5200A. The Electron Diffraction Investigation of Chlorine Fluoride and Chlorine Trifluoride Citation Wahrhaftig, Austin Levy (1941) The Absorption Spectrum of Chlorine Fluoride in the Region λλ4650-5200A. The Electron Diffraction Investigation of Chlorine Fluoride and Chlorine Trifluoride. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/jm8w-mj73. https://resolver.caltech.edu/CaltechTHESIS:10312023-173840321 Abstract [No Abstract] Item Type: Thesis (Dissertation (Ph.D.)) Subject Keywords: (Chemistry) Degree Grantor: California Institute of Technology Division: Chemistry and Chemical Engineering Major Option: Chemistry Thesis Availability: Public (worldwide access) Research Advisor(s): Unknown, Unknown Thesis Committee: Unknown, Unknown Defense Date: 1941 Record Number: CaltechTHESIS:10312023-173840321 Persistent URL: https://resolver.caltech.edu/CaltechTHESIS:10312023-173840321 DOI: 10.7907/jm8w-mj73 Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided. ID Code: 16220 Collection: CaltechTHESIS Deposited By: Kathy Johnson Deposited On: 31 Oct 2023 17:50 Last Modified: 31 Oct 2023 19:18 Thesis Files PDF - Final Version See Usage Policy. 37MB Repository Staff Only: item control page

The Absorption Spectrum of Chlorine Fluoride in the Region )J4650-5200; The Electron Diffraction Investi gation of Chlorine Fluoride 2.nd Chlorine Trifluoride . Th esis by Austin Levy Wahrhaftig In Parti al Fulfilment of th e Requ irements for th e De er ae of Doctor of Philosophy 8alifornia. Institute of Technology Pa s a den a , California Tabl e of Cont ents I. Introduction. 1 II. Prepa ration o:E' Ch lorin e Fluo ride and Chlorine Trifluoride. 1 III. The Chlorine Fluoride Spectrurn-2xpe rimcntal Procedure Description of Bands. IV. V. VI. VII. VIII. 6 The Chlorine Fluoride SiJectrum-Analys is of the Rotational Structure. 11 The Chlorine Fluoride Spec trum-Analysis of the Vibrational Structure :Determination of the Vi brati onal Congtants. 32 The Chlorine FluoTide Spectru:11-Determination of the aotational Constants. 40 The Chlorine Fluod.de .3pe ctru:rn-Deter:nination of the Dissoc iation 2nergy The Potential Curves. 48 The Chloi~ine Fluoride 3pectrum-- Some:; Ap:pli cations. 57 IX. The Chlorine Trifluoride S1) s ct:ru1n. 61 X. The El ectron Diffraction Inv e stigation of Chlorine Fluoride and Chlorine Trifluoride. 62 The Absorption S_pectrum of' Liquid Eydrogen Fluoride s.nd its Aque ous Solution in the Region .M BCC0-11000 . 70 XI. XII • Sumnary . XIII. Fhysical Constants. XIV. Refc r encss. ~Cl. Propositions . Illustrations f'age Figure 1. Apparatus for th e Preparation of Chlorine Fluoride and Chlortne Trifluorid e 2e. 2. Absorption Cell and Associated AFparatus 6a 3. ThG Spectruf!l of Chlorine Fluoride lla 4. A ~icrophotometer Trace of the Spectrum of Chlorine Fluoride llb Determination of th e Conve rgence Limit of the v 1 ~0 :;:)re gress ion of Chlorine Fluoride .iga 6. Potential Curves of Chlorine Fluoride 55a 7. 11 Com· '· l-oari· so·n of.>.!.. ,_ Ke 1 5. 8. and D 11 for the r e n' 0 Halogen Mo lecules 57b The Radial Distribution Function and the Reduced Theoretical Intens ity Curves for Chlorine Trifluoride 67a Ta bles Table Fage 1. :.:el tin.g Points and Boiling Points 4 2. Yi eld s Obtained in the Preparations of Chlorine Fluoride and Onlorine T;:-ifluoride 4 3. Fine 3tructure of Bands 6 ~ 0 and 7+- 0 of 18 Chlo:cine Fluor ide. '1. 0. 6. Fine Structure of i34nds Chlorine Fluoride .s ~ 0 s.ncl s~ 1 of 19 Fine Structure of Band 9 ~ 0 ::i.ne 10 Ch lorine Fluoride Fine Structur e of 3Qnd 10 ~-o and ! of 10~1 20 o:f' 21 Chlorine Fluoride 7. 8. Fine Structur e of Band s 14 Chl or ine Fluoride 0 and 15 +-0 of F 11 (:'"~ Values for th e 2 Q .) I Various aands of Chlorine Fluoride 22 Cocir·a :cison of' the d 23 Tables Tabl e s 0..,, . 10. "' n~: arison . vo 0 1,, th _,.,e A .£...J 1 E: S 11'""' II (J) VcLU 2 1 Fine Structure of Bends ( 3 ~ 0) f., and ( 10 ~o) i, of Chlorine Fluo ride l 24 25 11. 'v 'oy"" ·' a::ri of t 1!.1 P.:, ... i. iJ· .. · SO"'"! L 12. Band Origins of: the Ch lo r i ne Fluoride Band System 29 13. Diffus e Fine Structure of the 11.- 0 B2nd 31 14 . Band Hea.ds of t he Unresol vecl Chlorine Fluoride BaJ1ds 31 15. JJsslandres Ta1)l c of th e Chlor ine 1', l uoride Ban ds 33 16 . Is oto re Eff e ct in Chlo rin e Fluoride 39 17. Rot ational Con stants of Chlorine Fluoride 44 18 . Co~:::) arison o:f' the B11 A L._J 2.i.H1 0 II ( ~T) - va.1 1'-·1·"-''- S 2 1 Values of Chlorine Fluoride 19 . 20. 21. Values of Various Halo gen Mol e cul e s 26 46 ~ ., ec tr os co p ic Constants for the 57 a El e ctron Diffraction Investi gation of Chlorine Fluoride 65 :C.: l e ctron Diffr a ction Inve sti gation of· Ch lorine Tr if1 uo r ide 66 The Absorpti on Spectrum of Chlorine Fluor i de i n the Region I. \A 4650 - 5200 Introduction 1 Fluorine is the strongest oxidiz ing a.sent lmovm ,, so its pro erties are of very great interest to the chemist . Yet ,, there is very little accurate i nformation available as to its thermodynamic constants . All attempts to obtain the absorption spectrum of fluorine have yielded z only a continuum. ~ while the bands obtained in emission surely do not 3 involve the ground state • Hence ,, it we.s decided to investigate the ab- sorption spectrum of chlorine fluoride ,, itself a very interesti ng molecule , in the hope of obtaining more information about f l uori ne . As all the hal.ogens e .. cept fluorine have a relatively simple band system near the blue region of the spectrum, there seemed to be an excellent chance that resolvable structure might be obtained . In the preparation of chlorine fluoride , chlorine trifluoride is also obtained . This also was investigated spectroscopically, but nothing of i nterest was found . II . Preparation of Chlorine Fluor ide and Chlorine Trifluoride Chlorine fluoride was prepared by the me thod of Ruff 4 , direct combination of chlorine and fluorine . The method is simple--chlorine e.nd fluorine are passed i nto a reaction vessel at 250 ° C . ~ using an excess of fluorine , so that all the chlorine reacts to give chlorine fluoride and chlorine trifluoride . The products and the excess fluorine are then separated by fractional condensation. experimental difficulties . There are, however , certain The great reactivity of the fluorine and the even 6 reater re Qctivity of the chlorine fluoride and chlorine trifluoride 2. re quire tha t the appa ratus be made e ntirely fr om copper a.n.d quartz, with no vmx or gre as e i n irnr.l.ediate contact with t he products; and the :impuri tics in the fluorine,, carbon and silicon tetrafluorides ,, are difficult to remove. I n these experiments,, the fluorine vm s obta ined by the electrolysis of molten potassium hydro gen fluoride i n a generator 1Nhich was kindly made available by Professor Yost. Several different preparations of the chlorin e fluoride were made,, in betwee n which it was necessary to regenerate the charge in the generator,, replacing the hydrogen fluoride decom- posed by the electrolysis. It wa s found tha.t a method much f a ster and cheaper than putting in a new charge or regenerat i ng t h e old by the method of Simons 6 is to distil anhydrous hydroge n fluoride into the me lt. An- h yd rous hydro::;en fluoride of quite good purity c en no'i,.1r be obtained con:a0rcially ,, and to carry out the distillation it is necessary only to lead a tube fro:.n the t ank to n e e. r the bottom of the me lt, the hydroc; e:n fluoride. beins absorbed ve ry rapidly and smoothly, "':nthout e x cessi<re evolut ion of heat. In this man ner a regeneration c an be c a rried out in a fevr hours instead of a day or r.i.ore . ide c ~m A...nothe r a dvant a ge i s th.at more hydrogen fluor- be di stilled in than corresponds to th·a fonnula KHF 2 , and so a greater quant ity of fluorine c an be obtained before rege neration is again n e ces sary. After tre atT;ient in this v'm.y,, fluorine i s obtained within fif- te em r.1.i Ylutes of the start of the ele ctrolysis . -.:rn.s nl we.ys a ll01.ved to prOC i-~ ed Hovrever, the electrolysj_s for at le ast two hours before the start of the re a c t ion , i n o rder t o e li:.rn.i11Hte s ili 0011 and oxygen compou~1ds fr om the The a pparatus used to prep5.re the ch l oril1e fluoride is diagramed i n Figure 1. From the fluorine ge ne rator A, the sas flows through the traps B 1 C, and D into the furna ce E. Trap B,, of copper,, i s filled With fu. sed pot a s si u :n. fluoride to r emove hydroe;e n fluoride fro:m the fluori ne. + 15 VOLTS ® MON EL -;;iigure Apparatus for the Frep~raticn of Ctlorin F uoride 3nd Chlori~e Trifluoride Traps C and D, of copper wi-th thin-~Talled monel t ubi11c; le ?.ds a re coo led to about -75° and -130°C, respectively, trap C by a dry-ice alcohol bath, tro..p D by a liquid propane bath. The te ~'TI.perature of the latter, de termined by means of a five-junction copper-constantin t h ermocouple, vra s he ld in the ra~ge -140° to -J.30° by bloYiing liquid air through a spiral of coppe r tubing im:ne rsed in the liquid propane. The reaction vessel E, a copper block bored out, with copper tubes silver soldered in place as indicated in Figure 1, ~'-m s heated electrically by a heater wo1.mcl abou-c it. The temp- I.. 250° anc.] 27r::v 0 d uring . crac;ure, measure d b y a ..L.h t,,_ ~ermome t er, Nas ,ne ld ,oe-i,,vveen 1 J all preparations. The chlorine used was com.ne rcial tank chlorine which is believed to be Nasonably pure. It YiaS bubbled twice through concentrated C.P. sulfuric acid, pe.ssed through a column containing tightly packed glass 1Yool,, and then led into t.1-ie reaction -vessel through a copper tube. The rate of flow of the chlorine was determined by its rate of bubbling through one of the sulfuric acid traps. For ease in dis as se.rJbling,, the reaction vessel outlet was fitted with a copper standard t2. per C'1f7/25), and the next two traps., F and G, were constructed of fused quartz with simil a r tapers. Trap F v,r!:..s cooled in a dry-ice a lcohol bath to about -75°; trap G was kept at about -150° in the se.m.e marn1er as trap D. It was found necessary to use a sr.:mll am- 01mt of lubricant on the tapered joints, Nothing that was tried 'Nas very satisf a ctory,, but graphite bound together with apiezon grease and applied to the slightl:r war med taper was found to hold up for a week or so. The freezing points and boilint; points of the substA.nces wllich need to be considered in this preparation are listed in Table 1. 4. Table 1 Melting point,, oc ~ Bo:i. ling po int, oc HF -92 19.4 ClF'~ -83 13 -59 9.9 -34 -102 SiF4 ClF 731 ' -95 -154 760 \subl.) -100 CF 4 -130 -223 -188 It is seen that the impurities present in the fluorine are largely re~;ioved in traps A, B,, and c. Thus, proYided a sufficient excess of fluor- ine is present,, the only substances issuing from the reaction vessel are chlorL11e trifluoride, chlorine fluoride, and fluorine. The first of these is condensed in trap F,, the second in trap G,, and the last is not condensed. The fluorine generator current was held at about 10.5 amperes, -which at 100% efficienc:l corresponds to 4. 9 liters of fluorine per hour. The efficiency o.f the generator is unblown,, but Ruff has reported 6o% to 83%. The rate of flow of reactants,, the length of the ru::.n., and the est- ilnated amom1t of products collected are given in Table 2. Table 2 5. At tho conclus ion of a pre par ati on ,, the quartz t r aps v.re re di sass e r:1bled from th e appar atus and t heir outlet s off quartz tapers. closed~ with se aled The traps contai ning the chlorine fluoride and the chlorine trifluoride we r e kept in li quid air and dry-ice alcohol baths , respectively , Un.til completion of the experiments. Before being used,, . the products vrere distilled in aJ1 all-quartz system,, approximately the first and la.st 16~ of the distillate being disc arded. It is believed that the purity of' the products was reasonably good. The most likely impurities in the chlorine fluoride are the c arbon and silicon tetrafluor ides. Both of these are present i n the fluorine as it comes from the generator , but should be largely removed L"Yl trap D,, as previously mentioned . Also, carbon tetrafluoride could be obtained by the action of' the chlorine fluoride on the lubric ant on the tapered jo tnts, and silicon tetrafluoride could be obtained by the action of the chlorine fluoride on the qu artz . to be very slow. The rate of both these r eaction s was observed In any case, no trouble from any unknovm absorption was experienced in photographing the spectrum. Also, the electron diffraction i nvestigation (discussed in a later section of this thesis) gave no indic ation of the presence of silicon tetrafluoride, which would show up pl ai nly if present i n large araount,, more than fifteen percent. In the course of' the spectroscopic i nvestigation, chlorine dioxid0 wa s found to be present in bo th the chlorine fluoride and the chlorine trifluoride. This will be discussed later . A.n.y other impurities ·which mi ght have gotten into the chlorine trifluoride have boil ing points suffici ently differe nt from the boiling point of the chlorine trifluoride to be removed by the simple distillation . 6. III. The Chlorine Fluoride Spectrum ExperD~e nt a l Procedure Description of Bands The absorption spectrmn of chlorine fluoride was studied under both low and high dispersion. The absorption cell_, used for all photo- graphs, was a three meter len~th of brass tubing, of twelve millimeter i nternal diaLleter. The windows of fused quartz were cla~ped onto the ends of the tube using copper gaskets. The apparatus used to ~vacuate the cell and to introduce the chlorine fiuoride is shovm in Figure 2. Valve A is a special needle v alve, designe d primarily for use on the electron diffraction machine. It is so constructed that :nothing but copper comes into co nt~1 ct vd th the gas. The needle of copper with a spherical end of 3/16lt radius· is forced aga~:.flst a narrow seat about 3/16" in diameter by a stront: spring. A o.006" copper diaphra.m 1 1/2" in di ameter per:mit s a 1/16 11 motion of the needle. The pressure gauge B, also constr ucted so that nothing but copper comes in contact with the gas, has a 0,0015 11 copper diaphram 3 3/4" i n diameter on wh ich re sts a steel needle, the uppe r part of which is visible through a gl ass tube . The g au :;e vms used as a nul 1. indicator~ the a ir pressure above the diaphrrun being adjusted to balance the chlorine fluoride pressure below. The sensitivity of the gauge is 101.';r; also the dia- phrai.n showed considerable hysteresis.., but it was adequate for adjusting the pressure in the cell to ·within ha lf a centimeter of' the desired value. The valv.e C is a commercial stainless steel needle valve, placed so the packing gland is on the pu.~p side of the line. It was entirely satisfactory, save that the packing had to be replaced on ce. The 3olid sodium hydro xide in the trap D~ placed there to protect the pump and the mercury manometer F, was not very satisfactory, as the rate of re action 6a ® TO PUMP @ Figur e 2 .:'~bscrpti n Ce ll and AssocL.t e d f paratus TO MANOMETER AND PUMP 7. of the solid dry sodi'Uill hydroxide vd t h the ch lorine fluoride was s l ow. The chlorine fluoride that did get through the sodium hydroxide reacted with the ptunp oil. This was no doubt rather hard on the pump , but did keep the chlorine fluoride out of the laboratory, vrhere it could have caused far more daznage. The entire system was reas onably vacuum tight , considering the fact that most of the connections were made by standard brass couplings. When evacuate d to the l:Lmi t of the rather old mech anic~tl pllinp used, a pressure of about two millLmeters, the pressure would rise to perhaps one and one-hal f centimeters in an hour. As t he exposure times never ran much longer thai."'1. this,, and as only the moisture in the air could cause trouble, such a leak ·was of no consequence. The experimental procedure was to evacuate the system completely with the chlorine fluoride cooled in liquid ai r. Then, vabre C was closed and the liquid air removed until the pressure had reached the desired value, when valve A was closed and the liquid air age.in placed around the chlorine fluoride. The exposures were then taken as quickly as possi.ble, after which valve E was closed and valYe C opened , allowing the chlorine fluoride to come into contact with the sodium hydroxide in trap D. Then, after ten or fifteen minutes ,, valve E was opened a...11d the system pumped out~ The light source used was a five hundred watt projection lamp focused into the tube by a short focus lense. At the exit end of the cell e. lens and pla....."Yle mirror served to focus the light on the slit of the lo·w dispersion prisr:i instrwnent ,, while a lens only was ne eded to focus the light on the slit of the grating. With unfogged windows ,, exposure times of t vvo minutes and thirty minutes were needed on the prism and grating instruments ,, respectively , at A 4800. However , the quartz windo·rn were attacked fairly rapidly) the light by causing dissocj_ation of the chlor- s. ine fluoride gre atly increasing the rate of the reaction. The ·window at the exit end of the absorp tion cell remained in a usable condition for perhaps ten hours, but the window at the source end became badly fogged after only two or three hours . During several of the exposures a filter consisting of a ten centimeter layer of a ten percent solution of potassium ferrocyanide was placed before the cell to cut out radiation of wav e length s short er than A4400, but this had very little if any effect on the fogg ing of the windO'ws. The 101.ir dispersion studies were made on a Littrow quartz spectragraph with Eastman 40 and Eastman lF plates. The dispersion obtained was about 64 cm- 1 per mm. at ). 4000; the slit ·width was o.02mm. for most plates, O.Obnn . for several. An iron spectrum was placed below and above the ab- sorption s pe ct ra on each plate . and in the region ved. The region ~A. 3500-7000 was investig ated AA 3500-4830 a large number of band heads* were obser- The long wave length heads were weak with an atmosphere pressure of chlorine fluoride in the cell, but the strength of the absorption i ncreased rapidly on going to shorter wave lengths# so that in the shortest vmve length region examined a fraction of a millimeter pressure was suffic i ent. La rgely due to the decrease i n output of the source at - shorter wave lengths, the exposures time had to be i n creased from one-half minute at .,1 '6000 to fifteen minutes at A 3500. Me a surements were made on one low dispersion plate on which four spectra had been taken , the first 111ri th a chlorine fluoride pres sure of 50 cm. and an exposure of 1 mi nute ,, t he others at 50 cm. and 4 minute s ., 3 :mr.1 . and 4 minutes ,, and 0.3 Illin. and 13 minutes ,, respectively. eight band heads were me a sured. Fifty With a f ew exceptions, four measurements from at le a st t'wo of the above spectra were made on ea.ch band head. ---=· . .. . ---~· *On pp. /3-/~ of this thesis a short description of some of the spectroscopic nomenclature is given. Similar measurements were made on thirty-one selected iron lines. The wave lengths of the iron lines we re obtai:>J..ed from the M.I. T. Wavelength 6 Tables • The reciprocals of the wave lengths were calculated, the cor- rection to vacuum as given by the International Critic al Tables 7 ap- plied, a linear i nterpolation formula obtained , and a correction curve plotted in the usual way. On plotting the positions of the band heads on a wave number scale it was seen that there were at least four progressions, e.nd the general appearance was that of a polyatomic rather than a di atomic molecule . At the suggestion of Dr. Eyster, the above bands were compared with the ch_orine dioxide b ands in this region, the vibrational structure of which 8 has been thoroughly investigated by Urey and Johnston • It was found that fifty-seven of the fi~y-eight bands measured were reported by them, the mean deviation between their results and the results of this investigation being a.bout seven wave numbers. The remaining band head, at -1 21,140 cm- ( A 4729) , was l a ter found to be of chlorine fluoride. As reported by Urey and Johnston, the chlorine dioxide absorption is very strong i n the near ultra-violet but becomes wea.k as one goes into the blue region of the spectrum. They found it necessary to use pressures rangine; from 0.2 to 600 nnn. i n absorption cells 30 and 60 cm. in length. As the difference in boiling points of chlorine dioxide and chlorine fluoride is very large ( see Table 1), it seems impossible that a large concentration of the fo rme r could be present in the latter after distillation. It seems more likely that the amount of chlorine dioxide was very small, probably less than one percent,, this percentage be ing more than eno ugh to account for t he observed absorption. The presence of chlorine dioxide causes no trouble i n the spec t roscopic investigation,, as the bands due to i t are easily ide nt ified on the his h dispersion plates , btt the appe a rance of the low dispersion bands has been treated 10. in so~ne detail here for use later in re [;ard to the investigation of chlorine fluoride and chlorine trifluoride by the electron diffraction method, in which it is not usually possible to determine the presence of or the effect of Lnpurities. The high dispersion studies were made in the second order of a t vYe nty-one foot concave grating, usinr; Eastman 40 plates. cell ·was used exactly as in the lmir dispersion ·work. The abso rption A very le.rge number of b and s vYere found, most of them ·with very complicated, on ly partially resolired, fine structure, and hence due to chlorine dioxide; also several bands obviously resulting from a diatomic mo lecule of low moment AA 3990-5380, was of inertia we re observed. The region about these, then intensively studied. The pressure of gas used varied from 60 to 80 cm. It would have been advantageous to use a higher pressure or a. longer path length, but neither wa s feasible. Exposure times varied from one-ha lf to two hours, depending principally on the condition of the windows . The slit width used was o. 035 or o. 040 nun., and the dis- -1 persion obtained was approximately 5 cm per nun. .A..Yl. iron arc comparison spectrum was placed on either side of each chlorine fluoride spectrum. A beautiful progression of five bands~ de grade d to the red, we. s 0 found,, with sharp heads at 4901, 4868 , 4812, 4761 and 4728 A. struc ~~ure in t h ese b = m ds i s cle e.rl;y resolved. The fine The b ands D.t ~:901 , 4868 0 and 4761 A seem to consist of doublets whose separation incre P.ses as on e goes furtr:er e:wa;;T fr om the ba.nd he 8.ds. The bands at A 4812 e. nd ft. 4 728 consist of single li:n es h e.ving a spacing uniforr;i. ly incre G.s irig with dist nnce from the origin, but the s pacint; in the il 481 2 be.nd is t wice that in the ./l 4728 b and . Bet-ween the he ad 8.t A4728 and ) 4680 there is a region of bros.a.. diffuse lines in which there e.re faintly visible three diffuse band heacls but no corresponding fine structure. Then at about 4662, 4654,, 4649 and 11. 0 and 4647 A there e.re four more band he a ds vihich plainly co nverse to v. limit very near the last mention ed he ad. The firs t three of t hese last mentioned heods ha.ve plainly resolved fine structure, but it was possible to measure the lines of the first two only. A contact print of the above region is given i n Figure 3; e. microphotometer truce of part of the region is given in Figure 4. The region beyond )\ 4647 was carefully examined for more dint- omic bands., but none could be found. A large number of bands of chlor- i:ne dioxide were found to the violet of .A4600, but their intensity de- cre8.ses rapidl~r with i n cre e.sing wave l ength,, so that i:.o be.:nds arising from chlorine dio::;-dde could be detected to the red of ~ 4620. It can be d efint tely said that there are no bo.nds of chlor:Lr: .e fluoricle in the region M4600-4647 of intensity at all compan.ble to the bands in the region 4901-4728. On the lo nE:; wave side of A4901 the bands soon become indistinct. The intensity of absorption rapidly decrcecses and the bands begin to e:verlap badly. i\ 5250 the absorption lines fade out c01:iplete ly Beyond ·with the path length and pressure used. Two bands were anal;)rsed in this region, their heads being at .A.4907 and r:v" A4987. The Chlorine Fluoride Spectrum Analysis oi' the Rotational Structure The first step in the analysis was to make high-contrast en·1arge:r:~ents of the chlorine fluorine bands and to pick out the obviov.s branc hes . This ·wa s very easy in the bands shovm in Figure 3. easil~r be seen that the bands consist of single P and R branches , ·w-hich gi1.re the bands at in It can the bmd at )4868 and 4761 a doublet appe a rance and which overlEtp ) 4812 0 After pickinr.; out the bre.n ches .o f several bands.., the positions of all the lines on the plates in the v-icini ty of the bands so tre 2.ted vre:re Ila 17+-0 16~0 ;s-.-o 14~0 .X 4647 4649 465"4 4662 13~0 12..-.0 //+-0 10..-0 4728 9.-0 4761 84--0 48/Z 7 ....... 0 4868 Fi.gure 3 6~o 4901 II 6. 17 ~ 0 113 ~ 0 15 ~ 0 14 ~ 0 :_3 - 0 12 - 0 11 ~ 0 10 - - 0 Figure 4 12. ~-Ieasur~ments ·1dere made on two plates , the region '14654 to r:J.easured. i\ 4 900 being measured on one (the plate shovm in Fi gure 3) and th e region A4900 to A5032 on the other. Three settings were made on each line of the absorption spe ctrum and tho iron arc spectrum . The accuracy of these ;:u.ea.sure~ne nts varied greatly. On most strong sharp absorption lines ru1d on the the iron a.re comparison lines the ma.ximu.111 diff e rence between any two of the three measurements :ca.de on each line we.s 0 .002cn . corr es ponding to a pr obab le error of less than -1 0.1 cm • But very f ew of the lin es measured were both strong and sharf, so that differences between successive measurements of two and three time s the a.bov e fi .s ure were not uncorn.mon. The calculations of the frequencies from th e comparator readings were made in the way describ ed for the low dispersion plat es . The deviation curves were smooth and had , a probable error l ess than 0.05cm-J.. Hence , it is be li eved that the 1 prob ab l e error in :r.i:ea.surement of strcng shar p lines is O.lcm- , of most of the remaining lines 0.2cm- 1 , whil e for some of the weakest lines ( Principa.lly from the 11 :i.sotope 11 bands) and for overlapping line s ther e ffia.y -1 be errors of 0.5 cm . ' -1 responds to C.2cm • It might be noted here that the slit width cor- and that the theoretical resolving power of the -1 grating in this region for an infinitely narrow slit is O.llcm • The rotational analysis ·, ms now carried out in the usuG..l way . For convenience in the description to follow, the nv.rn.bering of th e bends a s fin a lly de cid ed upon will be us e d, t he r easons for this choice of nu.r:.ibering being given l ater. The connection b e t woen this nu. rnbering . and t he wave lengths pr~viously given for certain ba.nd heads may be seen by a gl ance at Figure 4. It is evident that this beii:.g the absorp tion 3 ef. o"f' b anas ' 6 ~ O, 7 ~ 0 , 8 ~ 0 , s pectrum. of a col d gas , t l1c S9Fl08 °.:1~ 0 , and 10 -E- 0 rn.ust be a progression ar/ising from a. co:rrnnon initial stste a.nd hence there must be a numbering for ea.ch of thes e bands such that the 62 F( J) s 1 are the sar:w for all bands . Thr oughout this paper the stm1dard spectroscopic nornencle..ture is &llO 9 used .,,• We will here digress to summarize the more important definitions and formulas used in the following discussion ~ T=T e (1) +G+F, where T is the total energy of the molecu.le expressed in WB.ve number units, :md Te, G, and Fare the electronic, vibre.tional, and rotational enert;ies, respectively, of the molecule. Then for s. diatomic molecule, G = G(v) is a function of a vibffational quantrun number v, and F = Fv (J) is a fm1ction of a rotational quantum number J~ and also of the value of v. and G(v) depend on the electronic state of the molecule. Both Fv(J) The two states involved in any given transition,, in this C[1.se two electronic levels, are de signated by applying pri.mes to all quantities referring to the upper or higher energy state, double primes to all quantities referring to the lower state. It is in general found that:,' 2 - G ( v ) = we (v + i) _, ii'e xe (v + ~) + '°eY e (v + 13) 3 + • • • • • • • • • • • ( 2) we 2 » we x e >)' UleYe Fv (J ) = Bv J(J + 1) - Dv J (J + 1) B v Bv 2 + (3) ••••••••••• » Dv = Be - (Xe(v -l- i)+ •••••••••••••••••••• B e >> a e (4) For the fre quencie s emitted or absorbed, ::: T t - T" = T e t - T 11 e + G' - G" + F' . (v~ (v") v'(J.') F" vn (J") • • • • ( 5) where the selection rules a re that Av = v' • v" = any integer and .d J = J' - J" ..t1 J =+ - 1,, or 0 in general, but in eerte.in speci al cases = +- 1 only . 14. Thus e.. given ban d systen1 arises fr om t he tot a lity of t ransitions between t vm electronic levels of the molecule. Each band of the band system con- sists of all the transitions involving p. single vibrational 1.eve~ in the u pper and lower electronic states and so can be denoted by two quantmn numbers,, v' and v"·,, the transition giving rise to an absorption band being Each line of a given band then involves a de- represented by v' ~ vn. finite rotational levels in the initial and final states,, but because of the selection rule on J it is more convenient to give only the quantum number of the lovrer state together with the value of .AJ. It is usual to consider the rotational structure,, resulting from F' F" ( ") as superimposed '1b011 the vibrational structure. vt (JI J v" J latter is described by The = -Ye + 111'e (v' + i) - cue1 x'e (v' + i) 2 + w'y' (v' + ~·)3 + •••••••• e e J _rw"(vu + l) - w"xn(vn + 2i)2 + • • • • • • • • • • • • • • • l11 e 2 e e 'VIJhere (6) 7/ = T' - T" is the origin of the band system and the values of zt e e e given by the equation are the band origins~ the positions of lines resulting from an imag;iri_ar;y transition (v' . _ v") in which F' (J') = F" (J") The rotational structure of a band is then given by ?/= v 0 + F' v' ( J' ) - F" v" ( J" ) • A set of lines in which ,A" J = J' = +l is called an R bre.nch,, a set in which ..AJ = 0 a Q branch, a.nd a set in which Ll J :::: -1 a. P branch. As no Q branch was observed in these experim- ents no r.i.ore consideration .will be given here to the case tl J = O<> Thus, we c an ·write:R bran ch:- J/ = l/ + F' (J + 1)- Fn (J) -=- P."(J) P branch:- ; / =l/ 0 0 + F' ( J + 1) - F" (J) = P (J) .......... (7) The quantities obtaine d by direct me a surement of the spectreg raphic plate are R(J) and P(J)• To obtain from these information about • 15. the rote.tional constants of the molecule one calculates the quantity A - F + l) - F v(J v(J - l)• F ~2 v(J) - It is at once evident from equation ( 7) that L12F"v(J) = F"v(J + 1) - F"v(J 1) = p(J + 1) - R(J - 1) ~zF'v(J) = F'v(J + l) - F'v(J - 1) = P(J) - R(J) As the values of the A ••••••••• (8) F ( ) vary quite appreciably with v, it is 2 v J ·- appare nt that if the same A F ( '../ are obte.ined by the analysis of several 2 V (.T) different bands, these bands must have one state in comm.on. Usirg equation (3) but :neglecting D , we may write more explicit v expressions for R(J) and P(J):R(J) = "l/ + 2B' + (3B' - B" )J + (B' - En )J2 0 =V0 (B' + B" )J + (B' - Bn )J 2 .......... (9) If B' - B" is appreciably different from zero,, it will be observed that there exists either a minimum in the values for p(J) or a maximum in the ve.lues for R(J)' t he former signif;ying that B' - B" is greater than zero and giving a band described as shaded or de c raded toward the violet, the l atter si gnifying that B' - B" is l ess than zero and giving a band des- cribed as shaded to-vva rd the red. This maximum or minimum is known as the band head and is the most prominent feature on a pl6.te taken vdth. low dispersion. The distance between the band origin and the band head is small if B' - Bn is greatly different from zero 1 as is the ca se in these bands , and can often be neglected i n qualitative cons iderations of vibrational structure. It is usual to arrange the band heads or band origin s into progressions ,, a. progression being a g roup of bands having a comrJ.on up per or lovfer vibrational level. J?rom equation ( 6 ) it is seen that the band or- i gi n s for a progression having a common 101Her state (v" con st.) must satisfy a:nf equation:- 16. -r/ v7,J = £.-" + w'e ( v' + 2l) - •'ex'e (v' + 2l )2 + "''eY'e ( v' 2l )3 + •••••••• +· 0 where 'J/:0 is a conste.nt for a given progression. We have so far ignored the fact that chlorine has two isotopes, Cl 35 and Cl 37, so that the spectrum observed actually arises from t1Ho molecules, Cl 35 F and Cl 37 F. The electronic terms T' e and T11 e1 depend- ing on the nuclear charges and the electron interactions are practically identical for the tvYO molecules, but the vibrational and rotational energies, depending on the reduced mas ses of' the molecules, vlill be differe nt. To distinguish between the two sets of bands and the two sets of constfrn.ts obta ined therefrom, the usual convention is follov·r ed in applying the superscript i to every symbol having to do with the less ab- unde.nt molecule. re~t:i.o of Cl 37F to Cl 35F is 1 to 3, relativ- Since the ely few be..nds of the isotopic molecule were found. To return to the ClF' spectrum, it was found that the rot&tional fin.es . of seven of the bands in the region studied could be numbered so as to give the same A 2 F", s. molecules in states having v" Since at room tempere.ture the number of > 0 decre c:_ses very rapidly as v" incres.ses, the state these seven b ands h ave in common must be the lower state. e x perimental values for R(J) and P(J) for these bands, called 8 ~ O, 9 ~ O, 10 ~ O, 14 ~ 0 and 15 .c:-- 7 ~ 01 0,, are given in Tables 3 to 7 along ·with the values of .L1 F 1 (J) = R(J) - P (J)• 2 In Table 8 1 are given the values of ~ F 11 (J) for these bands and their average. 2 The It is seen that the agreement is qui te satisfa ctory for the middle ran6e of J ve.lues. For small J 1 say J < 4 the lines involved come very near the band he ads and so are difficult to nier1.sure bece.use of the crowdi:ng together occurring there. o. Also the intensity becomes lmv as J approaches For high J the i nte n sity a.g Edn approaches ze ro, so that the lB.st lines me e.sured i n a branch are ~1ccurate. P In band 8 ~ 0 the overlapping; of a nd R branches give an additional error, which is discussed in some de- tai 1 l e. te r. 17. Overlapping the 6 .-.-- 0 band is another band, the 10 ~ l band. havin r-" the same Li F' 0 "' 2 (J) as the 10 ~ 0 band. . At a sli c:htly v v lo nser wave length another band,the 8 ~ 1 band, was found having the sar.ie .Lt 2 F• (J) as the 8 ~ 0 band. The 8 ~ 1 band is also like the 6 ....-- 0 band in having the lines of the P and R branches overlapping. On c01~maring the 4 ... 2 F" ( ) for t he bands 10 ~ J 1 and 8 found that they e.lso e.gree vvithin experimental error. -E.-- 1 it is The values for P(J)' R(J) and ~ F 1 (J) for the band S <:--- 1 are given in Table 4 2 where the A 2 F1 f\ tT ) may be cor.ipe.red with those fol." the 8 ~ 0 band. The ' 10 ~ 1 band is similarly tres.ted in Table 6 1. and in Table 9 tho~ Fn (J) 2 are compared for these two bands. Ta.ble 3 J 6~o i w, ! · . R(J) • , (t) i';' , 7-E:-O H', 14· (J) i L'.1 2' (J) i J + R(J) 1/2 - - -T- - ------ --------- -·--~----~-- - -·---- ---·---·- ·- -- -. " I ....,1 I I 3 l 1: (J) 391.191 4 .04 l 393 . 91 : 388.66 j 5 . 25 5 392 .15 . 386 .031 6.12 ' l J + -· --- ---~-- --------+----- --- ! 20618 . 65 2os20 .o s11 .oc 2.96 1.184 1.154 3 1 8 . 65 615 . 21 3 . 44 C.983 . 167 617.06 612 .37 4 . 69 1 . 042 . 1.13 615 . 21 609 .45 5 .76 . 047 .163 612.93 605 . 93 '7 .oo . 076 ~097 610 .17 6Cl . 80 8 .37 . 116 6 390 . 02 : '7 I 387 .16 :i 382.461 7.56 ! 373 . 931I 8 .23 8 384 .0l j '7 '7 ·J 4-.J 1 1:.t: 71 9 . 34 .099 606 . 66 597.44 9.22 . 085 9 380.50 1 369. n i 10.79 .114 602 . 78 592 .58 10.20 .073 i I,.' .u - -. IL] 2F ~ ! I 4 I ~ i1 .20393 .ool 20395.23 ; ~,-----r---- --:~•- . I l 2 i '"(J) - -------·--;·;-------- - - I I 10 376 . 33 ; . 364 . 621 11. 71 .11 5 598 . 49 587 . 25 11.24 .070 11 371. 78 358.871 12.91 .123 593 . 5~ 581.29 12 .23 .063 12 36-6 . 6 1 352 . 2.0 13.81 .105 5ss .1e.I 57·4:.88 13.30 13 361.25 345 . 87 lE· .3·8 .139 562 .G~ 5G7 . 81 14 . 87 .06·1: .101 14 355 . 26 338.84: 16 .36 .128 576 .4"1 560 . 25 lS.15 .114 15 ~4° ..., u . 78 !1 331.21 17 . 57 .13'1 569.55 16 ( 342 .4 7)! 544 .54 17.67 .070 4'">1 . c.. 535 . 60 18 . 82 .075 546 .101 526 .28 19.82 .071 516 . 23 20 . 97 .075 .861 505 . 92 21.94 .070 ! 562 . 21 ~523.23 I 17 33t.1: .37 i ( 314 . 4:3 V,1 ._) ;)-.i; • ! I 18 326 .28 !I 305 . 801 20 . 48 1 . 107 19 318 .15 296 .46121 .69 .112 537 . 29 20 309.39 ·-:is~ 54i "" o • · I 22.85 . 115 527 I 21 300 .24 275 .991 <J,1 0~ t=., -x • ~.v .128 513 . 09j ·1 95 .21 22 . 88 . 064 22 390. 47 265 .1 81 2.5.29 . 124 507 . 88 483.76 2LJ: e 12 .072 23 280.32 25 11 26.03 49 7.00 4 71.87 25 . 13 . 069 c:: ,.., c.;I L.h59 .4 9 26 . 26 .071 24 269 . 33 24;:~:1 27 .1 7 .10 8 25 258 . 05 ! 229.83 28 . 22 .107 26 248 .3 0 : 27 234 .3 8 : "'"' .109 48v • I vi I 473 . 75 446.81 26 . 94 .056 iI . 067 j I 203 . 85 I 1.109 30 . 53 433.28 Li:Sl.58 28 . 30 ! 448 . 80, 419 .54 ! 29 . 26 ! 435 . 68 L'"5 4· 1 30 . 27 . 061 1. f 28 I •'-' i • l. ~ . 062 : 74 I I ' L.__;?_9___________________J ______ ____________·-· -. - ·--·------·---·-·i_.!.22-~~st--------·--~L-·-·-----·~- - __ L _ __ J ' 29 • ! if 421. I ~ 19 . Tab l e 4 -,.-----...__ ..--- .. J -~-- R(J) ···- -----·-· ·- .r ( J) _ __ 8 +- 1 s ~o ---r-·--·-:---~-- · ~F ' ( J) 2082 1. 53 2.17 20821.53 2.0 8 19 . 36 ') 817 . 73• 3 . 00 820. 73 '"3 . 64 815.72 3 8 19 . 36 4 . 50 8 13 . 23 4 81 7. 73 5 815 . 7 2 210 . o9: 5 . 63 806 . 41' 6 . 82 6 813 . 23 802 .34 7 . 75 7 810 . 09 797 . 77 8 . 64 5 806 . 41 792 . 54 9 . 80 9 902 . 3..1 7 8 6 • 7 5j ll • 0 2 lC 797. 7'7 720 . 66j l l. 38 11 792 . 54 774 . 00! 12 . 75 12 786 . 75 766 . 81 13 . 85 13 780 . 66 7 59 . 04i 1£1. : . 96 14 774 . 00 750. 78: 16 . 03 15 76 6 . 31 74 2 . 04 17. 00 16 759 . 04 73 2 .. 53: 18 . 25 17 750.78 74 2.Q.1. 722 . 91 19 . 13 18 712 . 52: 20 . 03 19 . !i 732 . 53 20 ; 7 22 . 91 701. 711 21. 20 690 . 26 22 . 26 21 712 . 52 22 . 7Cl. 71 78 . 42i 23 . 29 2~ I 365 . 9 1 24 . 35 590 . 26 653 . 29 25 . 13 24 678 . 4 2 ,39 . 51 26 . 40 2b 665 . 91 625 . 45 27 . 84 26 653 . 29 611 . 0 9 28 . 42 27 639 .. 51 596 . 53 28 . 92 28 625 . 45 29 6 11.09 564 . 18 32 . 35 596 . 53 30 I 54 7 . 63, 31 1 i 530 . 53133 . 65 32 i 564 .18 33 i 512 . 82i 34.81 547 . 63 fl9·· - . 271 36 . 26 34 I 530 . 53 4 75 . 27! 3 7 .55 35 5 . 2 . 82 455 . 99!I 32.• 28 36 494 . 27 -··--·--· - . . - - ----·· -·--· ------.-- . · R(J) --- - - - ? (J) 4F'(J) Lll ' (J) 1 ~ve . ---- A l ' (J) J + 1/ 2 -·-- ·--~-- -· -· 0 .I. 1 I I 20047 . ll 045 . 99 04: 4 . 62 042 . 4:8 039 . 88 036 . 82 C33 .18 029 . 27 024 . 51 019 . 66 OJ 4.09 008 .13 001 . 48 19994 . 32 986 . 78 97 8 . 58 969 . 91 960 . 79 951 . 22 940 . 85 930 . 30 919 . 23 9C7 . 50 i 89 5 . 02 882 . 51 20045 . 9.., 20044 . 6 04 2 . 48 039 . 8 036 . 8° 033 . 18 029 . 27 024 . 51 019 . 66 0 14 . 09 008 . 13 00 1.4 19994 . 3 959 . 9 960 . 79 951 . 22 94C . 8 . . 930 . 3 919 . 23 907 . 5.__, 895 . 0 88 2 . 51 2 . 49 3 . 51 4 . 74 5 . 66 6 .70 7 • ._5 8 . 67 9 . 61 10 . 4 2 11 . 53 12 . 6 1 13 . 81 14 .70 15 .74 1 . 87 17 . 79 18 . 69 19 . 94 20 . 92 21. 6 2 22 . 80 24 . 21 24 . 99 2.17 2 .74 3.58 4 .62 . 5 . 64 6 .76 27.84 28 .42 28 . 92 1 . 4:4 7 1 . 096 1 . 022 1.027 1 . 025 1 . 040 l. 020 1.019 1 . 021 . 0 21 . 017 . 014 . 024 . 023 . 0 25 . 027 . 030 . 022 . 025 . 027 . 018 . 024 . 033 . 023 . 035 .050 . 033 . 015 32 . 35 . 061 33 . 6 .5 . 035 . 039 . C51 . 058 7 .65 I 8 . 66 9 . 70 10 . 7 2 11 . 70 12 . 68 13 . 83 14 . 83 15 . 89 16 . 94 18 . 02 18 . 91 19 .9 8 21 . 06 21. 89 23 . 04 24 . 28 25 .06 26 . 4 0 34. 81 36 . 26 37 . 55 38 . 28 . 049 - -·-· -···--.1- 20 Tabl e 5 [f --·-·-- -· -·- g ···~-o -··---- - =====----==: -- ·-·--c=-~-- j l A2F' (J) R(J) p (J) I --··--·-·-···--- ...... ·-·---------·----- ··- - - - -----··--;I 0 ! i 1 20997.53 20995.69 2 996.85 3 995 . 69 4 S93 .92 5 991. 84 6 i i A/' (9: I ·-··" - -- - · - - _______J__±__L~- I I 1. 85 1.233 20993 .9 2 2.93 1.172 991.84 3.85 1.100 4.43 0.984 986.41 5 .43 .987 988.92 932.70 6 .22 .956 7 985.67 978 .37 7.30 .973 8 981. 73 973 .67 8.06 .948 9 977.38 968 .22 9 .16 .964 10 11 972.48 966.78 962.16 10.32 . 982 955.86 10.92 .949 12 960.89 948.87 12.02 . 96 1 13 954.30 941.17 13 .13 . 972 14 947 .09 933.13 13 .. 96 . 962 15 939 .35 924 . 39 14.96 .965 16 931.09 915.16 15 .93 .965 17 922.38 905 . 46 16.92 .966 18 912 .87 89 5 .13 17.74 .958 19 902.85 884 .28 18.57 .952 · 20 892 . 45 87 2.7 9 19.66 " .959 21 88 1.41 860 . 91 2C.50 .953 22 869.77 848 .30 21.47 .954 23 857 .53 835 .22 22.31 .949 24: 844 . 71 25 83 1.51 807. 44 24.07 0.944 26 777 .13 27 28 788,05 761.06 26.99 0.947 29 772.38 744 .55 27.83 C.943 756.20 727 . 41 28.79 0.944 739.34 709.77 29.57 0.939 703 . 39 671 . 88 31. 51 0.941 31 32 I r: J ~ I ! l -r Table 6 10~0 10.._.1 --1 ~-.-~~:~ AF-,~~l---~--~-l-~1---)~~., L1 ~;F-,~r-,- ~F'(J) ,I ~2F'(J) I,: ! .. (J\ ;4 '2 (J) I H J ..: ( / • / ' ! (.J) ' v I Ave ! J + 1/2 rt( T\ l u J 1~1----~-~;-----~·--~--+· -- ---- ·- ---·i ; 21146 ;osl 2 145.23 \ l!J,3 .s5 i i I 141.92 ! I 139 .53 ! 5 --- --- - 1------~t-·- ------~---- I I i 4 1 21142 96 • 14C .69 138.07 2.27 C. 908 3 .16 3 .16 0.903 3.85 C.856 4 .93 C.896 5.75 0.885 6.67 0.889 3.85 1.3 4 .60 ! i 2:21 4.93 .20358.58 ! I 365.86 ! 781 5.75 362.62 l zo357.50 6.67 352.80 6 136 .53 ( 130. 7 132.96 126 .29 : I 5.12 I 8 128. 65 121.20 7.47 355.20 347.66 7.54 7.50 0.882 123 .92 ll5 .46 1 8 .46 351.05 342.47 8.58 8.51 0.896 118 .54 109 .3o i 9. 24 9 .24 0.880 112.57 I 106 .01 1 i 098.91 : I 094. 99 : 11.02 08 s . 7 9 ! 12 • 12 336.08 339.49 l 329.12 10.37 10.27 0.893 333 .13 i 3 21 • 86 ll • 27 11.14 0.891 314 .43 11.ss 11.99 0.888 ( 20318 .15) 20305. 80 ( 12 .35) 12.87 0.887 I 326 .2s ; 091 .• 18 ; 078.31 : 12.87 082 . 90 ! 06 9 • 0 7 : 13 • 83 310.25 : 296.46 13.79 13.81 0.891 073 . 87' 059 .19 i 14.68 301.5~ (286 .54) (1 4 .98) 14 . 68 0.890 064.32 048.82 ! 15 .50 292.10 (276.99): (15.11) 15.50 0.886 054.00 037 .s2 : is .1s 282.12 16 .18 0.874 19i O<i:3 .32 026 .30 17 .02 271.65 : ( 254 .2}3): ( 17 .3 6) 17 .02 0.873 201 031.97 013 .90 18.07 260 .38 ( 24 2 .1 6) : ( 18. 22) 18 .07 0.881 211 019 • 7 3 001 .15 18. 63 248.34 (229.83) · 18 .51 ! 22i 18.57 0.864 007 .22 20987 .59 19 .63 235.80 216.11 19.69 19.66 0.874 i i l 222.75 23 ! 24l 2098C .04 958.86 21.18 209.55 128.23 21.32 21.25 0.867 2sl 965 943 .s 21.73 195.22 173.22 22.00 21.87 0.858 26( 950 .32 927 . 68 22.64 179.94 157 .32 22 . 6 2 22.63 0.854 27 934 .3 8 910.87 23 .51 28 917.93 893.98 23.95_ 23.95 29 900 . 88 875 .68 25.20 25.20 864 .41 237 .66 26.75 26.75 l .,u 140 .93 I i I 3 31 0.855 ! o.s4o I _ :::_ ::J i 22 . Ta.ble 7 1 . :142 . 35 , 21 41C 3 4 39 . ~8 5 121435 . 9C I ' 428 . 39 1 3.23 . 0 . 497 422 . 51 I 3.98 420 . 58 416 . 15 1 4 . Ll........'Z 9 4. H . 05 10 408 . 60 ! vc:: ;l C. 53 1 •-X~J ( 0 . 574) -1C'3 • 25 L1Ql.20 : 5.05 ( 0 . 481) 11 : 398 • '1 5 39 2 . 36 : G.09 C. 530 12 : 39 8 . 3'7 383 . 03 6 . 34 0 . 507 13 I 379 . 64 372 . 72 6 . 86 0 . 508 II 3 6 9 . 44 36 1 . 83 ; 7 . 61 C. 525 350 .1 5 i 7 . 61 0 . 491 337 . 35 1 8 .co 0 . 4.85 17 1 332 . 50 323 . 79 1 S .71 0 . 4:9 8 18 \ 3C9 . 60 \ S . 6 1 0 . 5J.9 294 .62 ! 10.49 0 . 532 , It ..L. '":I: I C", I 15 1 357 . 76 16 1 2: 45 . 35 319 . 21 l S I 305 . 11 ' i.59 o . 45 11 I I . Ll:70 . 85 I l.9 8 0 ~ 440 i • I"',... c::: 7 I .2..- . :Jl - ~ ±:00 . 0 1 I ,... 51"' l.} . 4 o I 214 14 . 21 I 1 0 . 522 4 2c . 49 ·214:75 . ss '-172 . 83 L.t33 . 03 ! 2 . 87 I 7 o . 551 · 2143 7 . 10 I 2 . 33 431.0 2 I t±;~ 6a . o 8 1 1 ~rn i . 5 5 : 2 • s3 0 . 45 1 :1 58 . 86 4 55 . 74. ' 3 . 12 o.416 . 4 52 . 52 448 . 91 l 3 . 6 1 0 •'±25 1~4 5 1i-41. 38 ' 4 . 29 0 . 452 I . 57 423 . 82 i 2 _10~l·---~i -2-7-2-1._7_1~!~--~~ ____ L______~---·---- -·- - --- ----···-----· -·- -' . __ - .42 1I 1. 93 . - 262 . 19 j 23 Tab l e 8 ~-1 A ii111 1~ 0 ! s ~·;---· !-- 9--~ o ; io -4-o '. 14 --o I 15 ~o I t -- - - - - 9.20 '11. 45 10 !21 063 11 .23 .53 23 . 6 1 12 j25.91 13 :27. 77 14 '3c . 04 15 !3L97 16 : - -17 ~ -18 13 7 .91 19 l39 . 74 0 ?O !a. I _.:., • 16 21 [44 . 21 22 :45 . 95 23 ;48 . 31 25. 71 27. 93 I 13 . 22 .15 .35 :17 . 4 5 \19 . 39 -- - -~ - -- -- ~ 3 .80 5.81 7.50 9.27 11 . 32 13 . 38 15 . 46 17 . 55 19.66 21. 68 23 .7 7 25. 73 27.71 29.88 9 .. 20 11.13 13 . 4 1 15. 49 17 . 59 19 .. 4 1 21 .. 49 5 6 7 8 9 - 7 . 65 3 4 - -- 25 ! - - - 31. 86 : 31. 96 33 .95 .34 . 28 36.13 35 . 93 38.26 :38 . 10 4C .1 8 40.33 42 .. 08 : 42.27 44 .10 44.49 46 . 22 l 46 .61 48 . 39 j 48.42 50 .1 9 l 50.75 52 . 47 ! 52 .9 7 26 ;s4 . 25 5'1:r .21 27 28 56 .17 I 24 ~ 50 . 4 9 i 54 . 82 : 56 . 76 61 .27 63.46 66 . 00 29 ' 30 31 32 33 34 35 I - · - - --- -- ~ .. ..- - - - -- L ' --- ---·-- + - - - - - - -- ~-- -- ·--- I 5.69 7 .36 9.28 11 . 22 i 13 . 47 15 . 25 17 . 45 19.~7 21.52 23 . 61 25.61 27 .76 29.91 31 .93 33 . 89 35 . 96 38 .10 40 . 08 4 1.94 44. 15 Ll.:6 .19 5.36 7. 16 9 . 25 11.14 J.3 .24 15.33 17 .50 19.35 21.52 23 .. 55 25. 78 27~70 29.84 31.99 34 .08 36.05 38 . 02 40 • lQ 42 . 17 44 . 33 48 . 36 50 . 09 54 . 33 60 . 64 62 . 61 52.36 54.54 56.34 58 .70 63 . 22 7.05 I 9.32 i I 9.29 . -- ! LL2J.· ( J) Ave 1 3 .80 5 . 620 7 .3·44 9.258 11. 232 I 11.09 i 11.28 13 .. 39 ! 13 .34 113.350 15 . 47 15 . 57 115 . 4 17 17 . 89 17 .·43 17.558 19 .. 38 19 . 460 l ! 21 .69 21.69 21 . 602 I 123.5 4 8 23 .22 25. 735 ' 25 .67 27.77 4 I <21 .54) 29.91 7 i ( 29 . 49) 31.9·1 2 I (32.09) 34.050 33 .97) 36.017 ! (35 .75) 38 . 078 ! (37 .88) LJ:O . 086 I ( 40 • 50) 4 2 .1 29 : ( 42 . 92) t4 4 . 26 6 . 146 .242 ;48 . 370 150 . 330 152 . 600 5·4 .440 I !56 . 423 !58. 700 :6c .9 55 1,63 . 096 66 . 00 I I I J + 1/2 -----···-----l-.--- - ----1 I ( 2 .5333 2.2480 2 .0982 2.05 73 2 .04 21 2.0538 2.0556 2.0656 2 .04 84 2.05 73 2.0476 2 .0588 2.05 73 2 . 06 23 2.0607 2.0636 2 .C5 81 2 . C582 2.0556 2 . 0548 2 . 0588 2 .0552 2 . 0583 2.0563 2 .C627 2 . 0543 2~ 05 17 2 . 0596 2. 0662 2 . 0687 2 . 0952 69 . 91 72 . 36 169 . 91 ;72 . 35 2 . 0868 2. 0 973 I 7 4 . 54 74 . 54 2 . 0997 I -·-· *In t he (8 ~ o) band, t he ,~ an d R br an che s ov e rl ap a l :no st p erfectl y fo:r ·s~a ll J , but g i v e ban d l ines f o r J . -'l 1able ----- 9- ·- r-·r-- ·----·- -·-- Li P ' .i II J _ _Jl --8_+- 1·--'--+----1-0-~ ~__ _ _L__:,-: :_ -___!_ ~&1 I '"i ..!.. ! 0 c.. L 3 7 ..-23 4 9 . - 17 I I 7.23 2 .066 9 . 17 2 .033 f I r;:· i.J 11.44 11.08 ll.2C 2 . 047 6 13.21 13.06 13 . 14 2 . 022 7 15.37 14 .96 15.16 2 . 021 s 17 .16 17 .lC 2 . 019 9 19.09 19 .12 lS . 10 2 . 010 10 21.14 ( 21. 93) 21.1 4 2 .013 11 23 .03 23 . 03 2 . 003 12 25.34 25 . 06 25 .20 2 . OlS 13 27 .31 27 . 33 27 . 32 2 . 024: 14· 29.55 29 .82 29 . 68 2.047 15 31.57 (31. Gl) 2 . 037 16 33.53 r .•,...,. 2'°'~ ( ;)O • 0 2 . 032 17 35.50 35 .56 2 . 032 18 37.73 37.73 2 . C39 19 39 . 55 I\ ._ -::a • ._, • oc>) .::;V 2.028 20 t '. l . 56 ( 41.2.2) 2.027 21 43 .72 22 45 . 83 23 4. 7 . e5 1 .cc 2 . 056 .:5 . 83 2 . C37 ·~ 7. 71 2 . 030 1 ~ --1 47.57 2 . C22 24 49 .. 53 25 52 . 23 ~ 2 . 23 54 .29 5 .-: . 29 26 I ·- -- -- --·--~ - - ·-- - - .. 2 . 018 25. . --- . ·-· ---·-··-------------·- 1 rs ~ o"i ~ I J RC;u \j 1 2 3 20804 . lC ' 801 .61 4 5 ;20804 . 10 5.66 1.029 6 .56 ·11.009 790 . 88 7.56 l.C08 9.44 0 . 994 798.44 6 801 . 61 . 795. 05 7 798 . 44 I 8 795 .05 i 9 7 90 . 88 ! 781.44 10 775.86 i 097 . 59 11 781.44 769.85 11.59 1.008 100 • 1±9 12 775. 86 703.34 12.52 1 . 002 :2 1093 . 88 13 I 76 9 . 85 lLl.: I 7G3 . 34 15 748 . 81.l: 14: . 50 l.COO 16 i 748 . 8-4 17 74:0 . 61 I 19 i . ___ _______ _ _ 0 . 855 074 . 80 11.65 0 . 863 056 . 91 13 . 48 0 . 870 061 . 38 ' 047.1 8 14 .. 20 0.861 036 .5 1 15 .18 0 . 86 7 I 041 . 53 i o3o.9s I . : J_ _ _ _ ~' -------------------··--- -- --··'----------~' l .. 9 . 83 070.42 ! 051.69 j ----- - ---- ·--- -·-- ·-· ·'· · · - 090.66 066.24 740 . 61 I ; 086 . 45 18 !- 103.93 .I 26. Table 11 L1 -qtll i,, 24 ( J) J (s~oi (10 1 -=-o/ I Ave. LJ 2F" ( J) J .... + 1/2 Q 1 2 3 4 5 13022 13.22 6 2.034 ------ 7 8 17 .co 9 19.19 19 .39 lC 21.03 --- --- 11 --- 1"c.. 17.00 --25.69 27.02 27.64 14 29.24 29.51 --- G I 19 .29 2.030 I 21.03 2.003 I --- I -- 25.69 2.055 27.33 2.024 I I I I I 13 G5 2.000 I I ...., __ 2.026 --- II 33.91 I 33.91 2.055 29.38 I i -- 27. The above bands ac count for al 1 the stronc lines a.rid some of the ·, 1ea2rnr 1 ones in t he region stud ied, but l eave unexp lained the vast m'a j ori ty of weak lines . ed ; the Cl 37 Hov;ever, so far only the mo lecule c 135F has been consider- F mol ecu l e will also give rise to a band system, id ent ice.l in api--.;earanc e with the 01 35 F system and describable in the same manner but with slightly diff erent constants. :1 e consider here the change in the rota.tiona.l constants. value for Be is given by 11 The ·- Be • • . • • • • • • • • • • • • ( 10) h IS It is convenient to define a quantity f ~ • • • • • • • .. .. • • • • ( 11) Then, it is seen that B~ • f 2Be. A more complete tre a t ment 12 inDl dicates that ·.1e s o e obtain ; Bv l = i3 ·l ·e I n the next section of this thesis it is c al'culated that j/ = 0 . 99055 . Hen ce the second term on the right i n the above equation is very small comp ared to t1e 1n.1rs t an d we can say -'Rvl r.:i r/J 2-c")o. v' or ( 12) Using the v alu e s of A Ft ca.lculated in this 2 manner as an aid , i t wss possible to pick out the isotope bands cor resp onding to two of the bands p r e vious l y found , nar~,ely , the 10 ~ 0 and 8 ....-- 0 b a nds. The line s be long ing to these isotope bands , la.be l ed ( 10 4--- 0) l and ( 8 ...- 0) l , are listed i n Tab l e 10 a.long ~.-1ith th e v a.lues of A 2 F ' (J) 28. In Tabl e 11 the value s for A 2 F 11 ( ~) obtained fro m the two ba.nds are compared. Hext the positions of the band origins were determine d. It is seen fro m Equation 9 that i.\ J _ l) + F( J) = 2 Z,: + 2 ( B' - B J 2 ... ... ( 13) . 11 ) ~ Hence if one picks out an a r.proximate value for B' - B11 , sa.y B' - B11 , '] 2 a nd p lots the quantity R(J _ l) + P(J) - 2(B' - B 11 )J~ 8gs.i crn t J , one will obtain a straight line whose inte rcept on the ordinate a.x is is 2P(, whil e the slop e of the line will g ive the correction which must be a pp lied to t he a pp roximate value of B1 of B 1 - B 11 • - B11 to obtain the true valu0 The terrJs involving D h av e been ne g lected in the above. Th ey cause the line obtained to be s lightly curv e d and introduce a slight corre ction into the value s for B 1 - B 11 , but do not affect the extrapolated value of 2 Z/ ~ 0 . As it was only des fr e d h e re to o bta.in accurately1 the terms in D wer e i gnored. The r e sults of such calculations ar e giv en in Table 12. third colurn..r1 of the tabl e is liste d J/ extrapolation of 0 , In the the band origin, a s found by H(J _ l) + F(J) - 2(B 1 - B11 ) J 2 to J = O; in the fourth and fifth columns are listed the ap proximate va.lue of B' - B11 used in the calculation and th e a ctual va.lue of 3 1 - B 11 v,'hen corrected for the slope of the r e sulting curve. ·,'/hen measured, the value of Z/ ( band hea~ is g iv en for later refe r en ce. ~ rob able error in];/ It is e stinate d that th e in th e bands liste d is ±C.05cm- 1 , exce pt for the 0 15 + - 0 an d 10 +-- l ba nds and the isotope bands wher e the number of p oints a vail a ble for the e xtrapolation was very s ~all, so that the p rob able error is perhaps four ti21e s th e above f i. gure. ;S imilarly, the B11 is e sticated to be 0.002 C:r.l-1 save for the probab l e error in B1 - bends listed above whe r e it is several time s l arge r. l/ Band 29 . - - - - --- Ta b l e 1 2 . ba.nd h ead I I ! b ..n 2na,Yor igi -· ~·-------- -- - a) =i=z( - p n-~ \ corr _, c +._.t;:;"' d) B 1 _ . 11 (a s sume i Bt _ 0 ,._, - -- --- __ __.._ I ___ - ·-·· .. 6.- o 20396 . 45 20395 . 82 0 . 24 0 . 233 7...-0 620 . 02 619 . 32 .245 . 243 e~ o 821.53 820 . 37 • 26 . 258 g ~ o 997 .53 996.94 . 275 . 272 10 + - 0 21H:6 . 05 21145 e5 Ll .30 .• 288 445. 87 . 40 . 384 479 . 95 e -zv /I("\ . 407 20047. 04 . 25 . 259 lO c- 1 372 . 07 . 29 . 287 ( 3..-G) i, 808 . 60 . 25 . 248 \~o ..- c) ~ 134 .3 1 . 3C . 295 14 ~ 0 15 ..-- c s~ 1 20047 .1 I I I I I ,--- i ! I 1· 30. It has been me ntioned that in be tween bands 10 ..- 0 and 14 ..e-- 0 there is a re8ion of broad diffus e line s. from the band 14 ~ ori g in) Some of these 1 ines co!'.o.e 0 1 in 1:1hi ch th e 1 in e s ar e sharp near the head (or but be g in to be co:ne diffuse in both branches at about J 11 = 14 . However, it is seen from Table 7 that the perturbation causing this broadening docs not CB.use any great <lisp lacern.ents of 1 ine p osit ion. There ar c three !TI.ore band heads in this r eg ion which h2.VC not been mentioned yet; thos e designete d 11 ~ O, These can be seen in Fi gur e s 3 and 4 . 12 ~ 0, and 13 ~ 0. In two of the ba~nds, 12 40- 0 and 13 ~ O, there is not a trace of observable structure, even the The band 11 ~ 0 hea.ds being two diffuse to measure on the comparator. has a littl e structure for a s can be seen fro m Figure 3 and 4, there is a branch of very diffuse lines converging to a head. On a four- fold enlargement these lines have s. definite doublet structure, inO.icating t hey c onsist of mneared together P and R bran ch line s. of these as de t e r mined by meas urement on the Table 13. compa r~tor The c ent e rs are g iv en in It is seen that in th e bsnd 11 +- 0 the r e is some r egu - larity,, but there is consid e r able perturbation in t':·1c ~~, os itions of sev- e r al of the lines. The }'Os itions of th e ban d heads of these three diffuse bands were found :P.rom r:Jea. sur e~nents en a rni cro1:hoto;::-:. eter ;Jl r:3.te containing t rac e s of chlorine fluoride s .. e ctrum e.nd of the a djoining iron arc spectrum. The :~1.ate ~:a r o tor, rd croph otome t e red hsving previously been ~uc ·::sured on th e com- t he i n t e q : olati.on formula and correction curve then obta.ined were used t o cal cul ate the fr eciuen cies fro:'.n the :·.1 easure:1ents on th e tr a c e . 3 1.. J ! I L ____ m* _.- - - I 12 Tabl e 13 ----========--.--:..:=:-=·~~~~:::_::::::. .::__:_:::=:::::::::~ Z/ LJZ/ - --·--- - ------------ ~~-------------- ·21236.9 232.3 4.6 6.2 3 226.1 5.9 4 220 .3 7.2 213.l 8.6 l 204 .5 6 7 \ 8 187.1 9 173.8 164.7 ii L :ir:This n 'J.JD.b e rin n· is a rbitr a ry~ Tabl e 14 t-·------Z{ and- Band 1·- - -·--... 1 ----.- ~-·-----·------ - - -- h~~ ;--~I - - ------- ------ ·-··----------·-· ·· j n--- o 21255 .5 I 12.,_ 0 338.3 I 13 ...._ 0 398.4 1 l I ! I 499 . 8 17=--0 507.8 ___ ___j 32. Al s o me8 ~:ure d in th e san-:e way '!:!er e two b ands, 16 .__ 0 s.n d 17-+-- C, to the high-fr equency si de of thE; 15 ~ C band • . The p ositirJns of th e se The ~_irob3. ble error in l a st fiv e band s h eed s are g iven in Tabl e 14. these values is ± 2cm-1 , the error coining almost ent ir ely from th e p os- s i bl e e rror in th e choice of the p oint on th e trs.c e to be taY:e. n as th e hand he ad . It i.s o f course i r'.li~ ossib1e to find the b a nd o ri z ins for th e se bands, but it is seen fr o·,-:_ Ts.bl e 12 tha t the separation of bsnd he2.d and 08.n d ori gin of th e a~1d i3 oi' th e order of one wave ~ r o~ r e ssio n . Henc e , 1 u.r:1ber withi~ exp e r~me ntal for th e othe r bands error, the ba~d o ri g i n b o.::d. h es d '·nay b s t a.k -:;n r-.s coj.ncid en t for th e ba-Qds list sd i::i Tabl2:: Analysis of t he Vibrational St~ucture Det e rm i~ ati on c f the Vi brational Constant s . ).n a n a lysis of th e vibr a tional no u b e carri e d out, using the r :ce vi..::n.rn ly o bt-:dn e d. b8_nds . th e s e st!"J.ct u ~~ e ~c cur ~t e of th <3 b :=rnd. s ys t er"~ b a~ d o ri g i n s v2 lus s f or th s It is a.t once app3.r c:.1.t fr or.:1 the v a lue s of -;:iay Z/0 T1:..crc a.r e 8. -ls o hro i sot opic bsnds which b s l onc; to th 1.; f irs t p~o 2; r e ssion.s. of Ih s n th e Jc- s18ndr c s t s.bl c f or th e b3.n ds ;·.n. u st b e a s :-~i v ·,;n L t 'l:·a bl e 15, 11h e re .for the .,., r e s e nt t hs n;_rri_'o cri:'.1. g of th e be.: lds is to be te~ e n as t Gntativ e . 33. Ta bl e 15 I 223. 5 22 . 45 319.32 7 201 . 05 8 24 . 48 3 20 .. 37 170 . 37 9 996.94 10 211·1:.5 . 54 11 255 . 5 12 338 . 3 20017 . 04 ( 77'3 .33) 27. 97 1-12 . SC 38.G I 20372.07 ( 773 . 47) llC . O 27.2 2-2 •.s 22 .7 GC .l 12.3 393 . 4 17 ._J 17 •.5 13 . -'± 4'-1 5.87 3 'l . 0 8 479 . 95 1-5 10 0 · - '-1 • .J 3 .0 17 507 . 8 i)i~ ''l . . - 0 .. - ·1 __JL -...___,,._ 1 - - -- - - 1. l 9 I I 10 208C3 .. sc I I I 21134 . 31 J i L_ ____ ! --------- I II· Is o-::,o .:_::e ba.YJ.d::; 3 14.3 11 . 8 ·~: 99 . 8 16 ! I - ______ ___] 34. Since most oi' the rn.ol~cul e s ir.. a ga s ar e in their lm·r2st vi.brat- ional l ev e l, it is expe cte d that the lowest l ev e l abs e rved h 2, 3 v 11 so. This is found to be the ca.se in almost all studi e s of the s bso~ p tion s pe ctr 1 o~ other rn~ l e cules. Since a rather long v' ; rogr s ssion is ob- s0rv8d for chlorine fluori de , th e Condon pa.rabo la t;ype. 13 D.Ust be of the open This in tui~n r e quires t hat if the lo :1est vibra.tbna.l l ev·d 1 in th e 2;:r·ound stat e d oe s not giv e rise to t ho strong observ e d v' p rogressions, it must ~ive Be ca use of the that such conti:iuw~: ris e to a continum:i. e.t short e r wave l en.zths. chloj~ine dioxide b a nds, it is impossibl e to say ic: 8.bsent. H01.·rsver, it can be. said that there are no bands of chlorine fluorid e t o the s hort wav e length side of ~4650 ui th :i.nt cms i ty a s !'.:'lu ch as one-tenth th e intensity of th e 10 +- O band. -1 773.40 cm it is C3-lculat e d fror1 the 3ol tzma..Yl exp ression that th e ratio of th e mu:.1ber of ~ olec ule3 in a g iv en l e vel to the nwnber in the n ext higher l e vel is for loH vll rou[:.;hly forty to one-. The ir..t cns ity of a.n electronic 02.nd is the p roduct of two factors, the population of the initial state and th s tran si tion p robability b e tween tirn i::dtial and fi~3l states. There- fore, it is necessary that if a level e xists corresponding to v" • -1, th e transitL:m probability for th e band 10 ~ -1 must bt:: only onef ou.r hund r edth of the tr an s it ion p roba.b ility for the band 10 +- C. is 2. This v e ry l arge ratio fo:::- acl je,cent bands in the Deslandres table, es- r:e cially e..s the vibrational frequency in th:; ground state is f a irly hi gh so th a t the a:np litude of oscillation doss not chang8 much 1,1 ith vll fo~ v 11 s rnal 1. tlence it t1ay b e concluded that al though the nu::::i.bering in t rrn ground stat0 :might poss ibly be wramg, the ch ance tha t such is the case is ~1 eg ligibly s".:12.ll. A.dditions..l evidence in f a vor of ths conclusion 35. that the s e l ected numbe ring in th e lowe:r sta.te is corr e ct is obtained from t he pot ential curv e s for the chlorine fluorid e mol e cul e which a r e discuss e d in a later section of this th e sis. The bands v' .......- 1 a.re pe rhap s one-fifth as strong a s the bands v'+- O. This i ndic at e s that the t r ansition prob abilities for the fon ne r bands are eight times as l a rge a s the tra.nsi tion prob a.b ili ties for the l a tter bands, a not unreasonable f a ctor. o:is erv e d in U1e re g ion in '/llhich bands having v!I Ve ry weak lines we re :i;; 2 would a _?p C8.r, but th e se lines inight e qually we ll aris e from transitions v' ~ 1 with v' equal to or l e ss than aix. ·1. i i th the numberin g of t he lo wer state fix e d, th e isoto pic shift can n m,1 b e used to de termine the absolute numbe ring of levels in the uppe r state. It is first necessary to fit an equation of the form of e quation 7 to the i:: ro gression v 11 = O. It is cl ear from th e first and s e cond differences giv en in Ta bl e 15 that no simpl e quadr atic or cubic in ( v' + 1/2) will fit all th e bands of this progr e ss ion. be ex ~ e ct e d, ur e for U-1.3 This is to fo r the perturbation wh i ch destroys the rotational structbands 114'-- O, vib rn.ti onal l eve ls. 12 +- 0 and 13 ""f-- 0 surely also affects th e Such a pe rturb ation usually r e quires that th e lev e ls a bov e t h e perturbation r egion be fitt ed by an equation ap pr e ci a bly diff er ant frrnn th e e quation fittin g t he leve ls be lo w the p erturb ed r eg ion. It is evid ent fro m Tabl e 15 that only th e band's :'3 ..._ 0 to 10 .._ 0 ca.n be fit te d by a s i mp l e equ ation. It is n::c es s a ry h ere to writ e t he equations for the vibr ational isoto pe eff e ct. _rr ,·,'":II ( l~ v II Je start with Zquation 7:- + 1If.2) ' /I' ~ ti ,. xe If/\ v , + 1/2) 2 + ••••••••••••] • ( 14) 36. It is we ll knovm. that a; _L J(k; = e where k e 2rt c r-::zt ...................... ( 15) is the bond force constant. isotopic molecule, (,// t 8 :ro:: Hence it is seen that for the (Jt!Je' where/ is as defined before1 it can a lso be sho'tm that the d e p endanca of the other vibr a tional con.- f stants on , 2 we t x e t ::./] r to e x e C' • .J,. . , • and 0 f; .e y i,, e '7 =.0' ._; Ql y • r e e The :?Ositi ons of the band origins for the isoto r ic · r~10 l e c ules will then be givcm by : t/1.1 =- l/~ +(W8 '(v 1 + 1/2) ~f 2 (,() 6 1 x 8 ' : · '(-:·v- 1 + l/2f + f 3W 6 'ye'(v: 1 + 1/2) 3 + -{f{l)en( V 11 + 1/2) -,t? 2we"xe 11 (\/"+ i/2) 2 + •••••••••••](16) The isotopic displaceuent is p..pf.t • ( 1 -1) l(/ e ( I V I + 1/2) - ( 1 -1 ~ UJe t Xe t ( + (1 - As Vf + l/2f f 3) (,{) e 'ye' ( v + 1/2) 3 + • • • • • • • • t f is usually v e ry nearly equa l to one , t t e only excflptions being the ·rr:;olecules cont a ining hy·dro gen or deuterium., the a pp roximation may be made t hat -r 1 1 2 = ( 1 -f ) (1 +f) z 2( 1 -1) -fv - <1 -f )( l + f + f 1 = 3 ( 1 r ) ............ <18) '"Z i::quation 17 can t ben be simplifi ed:- i -f/' j, = ( 1 -r {w Ie(V I + l /2) - 2 u;, + l /2) 2 .. I Xe I (VI + 3 l{) 1 y I e e (Vt 3 + 1/2) + • • • • • • 37. As bands were fou..nd i':vol ving only two vibration levels in the lovrer electronic state, the values of W onlv G ,j 11 ( ~ 11 - 1) V , (OJ = {).) e n 11 e x 11 e cannot be obta.ined, but .:e wish to exa.:nine the iso- 2 W 11 x 11 • e e tope shift in th e i; ro g ression V 11 = O, so that the quantity desired is 11 as seen from eoua.tion 19, 112 I' a; e 1r~ ~ {l) 11 e xn e • - L one pe rcent of W 1 /2 ui' I' e e , the a;;proxima.tion can evid ently be ma.de that - 1 ~ W 11 e x 11 e = 1'-/1 2( G 11 ( 1) I!: 1/2[G r 11 ( 6" ( 0 } 1) - As (,()e x e is only about - /':_ 11 l.7 ( o) ) 11 7 · ui' e x e + i/2 a 3 86 • 7 Ocm- l • . v~/ - v-r/~ ' arising . . f rorn t,ni.s ~ a.ssump t ion . . a v mos t 0 • 1 cm- l • Th e e rror in is .l.. ·i1e might now calcul a te,?. From a su.'!'unary by Hahn the atomic. weights:- F19 • 19.0045, (A s ton scale, 0 16 = 16.0000). ;t{ A and (J • " c1 35 01 = 34.9803, 14 37 we have for = 36.9779, Thus, in at omic weight units, = (34.9803)(19.0045) = 12.3140, .A//'= 12.5529 53.9848 = 0.99045. Substituting the nu_merical value s just ob- t ained in equation 19, He have "!/- 't' t = 0 .00955{w e ( v I I + i/2) - 2 [,(/ I ex I e ( 3 (P I ey I e ( v I It was found that no quadratic in ( V 1 v + y2) .+ J/2) I 3 + 0) 2 + - 386 • 10) . would fit the p rogression of bands 6 ..,_.. 0 to 10 .c- 0 in a satisfactory manner, but that with the addit ion of a cubic term the ex;:) erimenta.l d at a could be fitted fairly we ll over a consid•:; rabl e r ~mge of w' x 1 a.nd ~ e e UJ' e y• • e As the isotope shift c a lculate d might de p end on the choice made, thr ee differ- ent equations were fitt ed to the band origins. The isotope shifts were t hen calculated, using each squation with seve ral choices for the V 1 nu.:aber ing . The thre e equations; the equations for the isotopic shift 38. obta in e d t h ere from, and a comparison of the values calculated from th e seve ral equations with the exp erimental values are given in Table 16. ( The isotopic shifts are given not only for the V nur.abering finally chos en but also for the cases wh e re the nur.1b e r assigned a given b.and is increased or decreased by one. variable It is to be noted that the c!lange of required for such a new choice of Y1 ( V 1 = V1 - l in or- der to increase the number on each ba'.1d by one) must be ma.de in equation 14, the equation for the band origins, and then a new equa tion 20 for the isotop e shift calcul a ted; th e change of variable cannot be made directly in equation 20. Equation 21 is th e best quadratic in ( V 1 + ij2) through the experb1enta l points, and 8quation 2la is the corr e s p onding equation for the isotope shift. It i·s seen that there are deviations from the ex- perimenta.l values of over a wave number, many times any possible exp erimental e rror. -L!;quation 22, _with a large negative value for W 1 y 1 8 fits the data very well, but has a negative value for tl.) 1 e x' e last means that the spacing of the vibrational l e vels for low would be con~rary ~o • , 8 This V' the spacing generally observed. As one purpose of this equation is to g ive reasonable ex~rapolation. to v'a. -0, a third equation, equation 23, equations 21 and 22. ~as arbitrarily selected as a compr6mise between This equation fits the data in a satisfactory man- ner and also has a value for UJ 1 x 1 of the same order of ma.gni tude as e e is usually found. It is seen by an ins p ection. of the isotope shifts as given in Table 16 that all three equations g ive essentially the same result. The nurn.bering chosen yi e lds very satisfactory agreement with experiment for v 1 = 8,, but p oor a g reement for v I = 10. If the number assigned to ea.ch ba.nd is increas e d by one, then satisfactory agr e ement is obtained 39. Table 16 V ~ 18336 911 + 396 . so( v 1 + i/2) - 12 .3oo( v ' + V2) 2 p'= 18618 . 48 + 292.604( V 1 + y'z) + 0 .304( V 1 + J/2) 2 •••••••••• ( 21) - 0.500( V I+ 1/2) 3 (22) Z/= 18561.65 + 313 . 4 84 ( V ' + 1/2) - 2.217( 1/ 1 + y'2) 2 ... 0.400( V' + 1/2) 3 ••• (23) v --g --r i 7 I I 20820.37 820 . 24 820.51 820 . 44 9 20996 .94 995.64 996.96 996 .7 1 le 21145 .54 21146 .44 8 21145 .52 . -~-· - - - --·-·· - 21145 . 76 ... -~ -- --- -- . 11. ( 21255 .5) 272. 64 263 .19 265.17 12 ( 21338 .3) 374.24 346.96 352.54 13 ( 21398 . 4) 4 51.24 393 . 85 405.49 14 . 21445 . 87 503.64 400 .84 421.59 15 214 79.95 5.3 1 . 44 364.94 398.47 16 ( 21499 .8) 17 ( 21507 . 8) -·- -- --+------ 224 . 91 I - -- I y-pi A~ca c . 00955 - -- - 2 396 .so( V' + lj2) - 24.600( V 1 + 1/2) - 386. 70 •••••• ( 2la) 3 2 1 L::>l/= 0.00955 292.604 ( + + 0 . 608( v' + - o . 5oo( v + V2) 0) v' 1;/2) - 386.70 .L1V'= 0 . 00955 313.484( V' + y2) - 4 . 434( v I + i/2) 8 ' ll~;~t. r 7 I ~I is decr ya sed by I i 11.77 ••••••• (2 2a) 3 + 1/ 2) -386. 70 • • • • ( 23a) -~ ll:~ ~{22 _ ca](;jll-.~~:=-1 -=~;'.~9 cs_:~~~~~ :~~i;=~~all ~9__:_1-1:_.23 __J_~~20 - ,- ----~·03 _ - __J___~~-7_?__ +__1_'.~~ If 1 -;.;~ Ef:f ·;:" -:. ;tizr i 6v.Jf~,j}/'- ~w--: -~43Cl : v' :AV'. l - 0 . 400( V -- -====--==---===--=-=::-_==-=--r-:::.--=-=---=-=-=· :_--::_ ---- -- --- --- F ·--- 1- 2 . 9.75 ore I I 2.02 9.87 I II 1.90 ~ '. 8 1 ~ - 1.4 2 9.84 1.93 1 11.23 2.35 8 . 43 : 2.80 8 .51 2.72 r------9 ·-+-- -·-·i -- 8·--.·-88· --· ---1I------· --- - ----+ ·-- -------·-- - -- -------·· - ·-- -- -. \ If v' is in er ea s ed by one I j t ! 9 11. 77 13.34 -1.57 13.49 1 -1. 72 !13.46 : -1.69 11.23 11..52 -0.29 ; 10 .98 +0.25 Ijll.09 I 11 ' +O .14 ! - - ---1 40. for v 1 = 11 but very poor agreement for v 1 = 9. If v 1 is dec r eased by one, the 8. ~re ement is entirely unsatisfactory; similarly if v' is increased by two the agr e ement is not satisf actory. There are two reasons for choosint:; the v' given r ather th an v 1 incree.sed by one. Fir st, the nur2bering chosen does give sli[htly bet- ter agreement tha.i.vi the a.l terna.tive g iven. Secon d, and more i mportant, t he data on the ( 8 .-... 0) t band ve more comp lete and tire believed to be more reliable than the d2, t2. 011 -L.ie ( 10+- 0) l Jene~, t:ie ~1.1.,Dlher- band. ing giving best agreement with the data. on the ( 8 ~ 0) i, band was s el- lected. tJ1/fll1Jvfh fh, It might be pointed out that~v' numbering is uncertain by 1, the quantiti es u s eful in tte: r modynarn.ic calculations and hence of i nt erest to t he chemist, the dissociation enercnr Do 11 '• (A) e 11 , o.; (,,() 11 11 -· e ~e ' Be 11 , all re- l atE: to the lower state and are inde;;sndent of the nu:m.be ring i n the upper state. VI. The Chlorine Fluoride Sp ectrum J c ter~ination of t he Rotational Constants. The rctational conste.nts of chlorine fluoride cr. . n now be obtained. Subst ituting :2:quation 3 in Equation 8 it is found that Fv~I ..T - .1.,) =- ( Lrnv - SD) (J + :yl2) - 8Dv(J + Hence , i.f the values cf Alv(J) are lmown, A 2 Fv(J) can be ca.lculated ( J + V2) 2 a.nd plott ed ago.inst ( J + 1/2) , and a stra.ight line fitted to the po ints. Then, t he valuG of the ordinate at ( J + V2) slope of the line giv es SD • v 2 • -o/4. g ives 4Bv a.nd t he Actually, in these experiments the ex- '-:11 . . ' was t r a;)o 1 a t· ion t o r\ J~ .:.. 11"::'. . .) 2 = 0 , 6D n:ia. d e b e ing . . comnE:g 1.J.t':_;l. bl e in v po ri son with 4B • v The L1 2 F~( J) e..nd ~ 2 F ~( J) values obtained fro1tl the analysis of the v a rious bands hav e a lr ee.dy been t abul a t ed in Tab l e s 3 t h rough 11. From th ese , or from the avera.ge ..L1 F where several bands have a. cormno.n 2 sta.t c, tl:e 4 l (J) J + v a l~ es were calculated . The r e sults of these 1/ 2 calculated are a lso g iv en i n Ta.bles 3 t:t.rough 11 . On p lotting the A 2 Fr.7J J+Vz it \·ras fow1d that the data in general were too sca.tt8red and were not avail ab l e for sufficiently high values of J to pe rmit the dete rmination of D v with filly a ccurs..cy . Date. is BV8.il a ble for only two l e vels of t he lo we r state v 11 == 0 . 15 Here , as sugge sted by Herzb e rg , a theoret ic sl v 2 lue for D G.nd 1. v ':<: LlB ..,; fro m the equr:.1.tion D = e e. was used in c a lcul~tin g D i s ":.'lot knmm, the e rror ir..troduced by usin g G( l) - G( o) v • \'l11 il e W 11 e --- (2 6) in its p l a ce is sri.al 1, o.s is a l so th e error i· :tr oduc ed by nsgls cting ,8, and so ase ssuming Dv = De . A pre l i!".ainary e s t i rn.2.te of Bs ' n:a de by a v ery rough oztr ap olt::tion of L1 l~eJ) and ~ F 11 l(J), yi e lding B/ 2 Dn = 4 ( 0 .5 2) ( 77~); 3 = 0.52, ·,;e hs.ve = ,. ., "4 • 10-6 v.~ 2 Using SD" as the slop e , th e best straight line through t he expc::rime nt a 1 ..L1 F 11 v a lues ws..s tLen drc:.1-m . 2 The i n t er c e p ts on the A~:I'" 2.036 . .J + a.x is we r e 4Bn 0 l/z ::. 2 . 061 , es timate of the probable error , we obtain:II B 0 .,-, Dl !I -1 = 0 . 5152 ± 0 . 0005 cm .:::. 0 . 5090 + 0 . 0015 cm -1 FrOJ.~j these , r-c;.ak ing an 42 . The rote.tional constants in th e g round state of ClF ar c t he r efore B II = 0 . 5183 ± 0 . 0010 e -1 CI£. ae..n = 0.0062 ± 0.0017 cm-1 e From these the inte r nucle a.r distance can be ~t e rrnined by the use ·;'; e f i I1d of e qu ation 10 . r r e 11 II = 1.6255 + 0 .002 - l..63ot1 ± o .co1 0 10 -8 cm -R 10 - cm. Tl: s <la.ta for the isotopic bands are too few to p er.mi t a very ac- cur a te de termin.ation of B u'l . Ifo1:13v e r , ta.kir..g 0 4 D ul = p D 11 0 ex pe ri ~ ental 0 d a ta Ll-B nt = 2 .03, 0 B ul = 0 . 5075 0 ment with = C.90 • 10- 6 , we obtain from the f 25 -1 0.005 cm , in s at i sfactory agree- -1 11 = C.5050 cm 0 To obt 2.in th e rota ti onal constants in t he u p p e r sta t e , t hs value s A Ci' of ~ 2J. v J + r (J) we re plotte d a.gs. inst (J + lfl) 2. Le t u s fi r st consider y2 thos e levsls with v 1 = 6 , to 10. The s e l e v e ls a r e in th e r egion of large a.nha r monicity of the potential function , as is evidenced by the e qu ation ne cess a ry to fit the band origins , Equation 23 . Hence it is not to be expecte d t h at th e va lue s for D will be even approximately v e qual to D e (A} e 2 The method u s ed to obta in D1 and B 1 v v first dr a w thrcugh t he exp erimenta l va lue for A 2F' was to a. straight line J + 1/ 2 tha t s e emed :r e asona bl e a nd det6.rr:1ine a tEmta.tive B v i) LJ v , was ca.l culated anu then a D. , 8 .Cquation 23 . and D • v From these using Equation 26 an d W ~ from The value s for D ue:r e t h0n pl ctte d against V an d a smooth v curv e taken . It vm s found that the si gn of DI") 1 ws s OPl::o s i te to the sign. for 0 43. Dv' for v' =f:-8; this arises from the overlap1J ing of the f e.nd. l-1 branches in this band, so that a mean value was measured on the comparator. Enough data is avail a.bl e on this band to r.:iake quite a good deterr.:ination of B 1 without assuming a D 1 8.Yld it is seen that this value agrees 8 8 well with the B 's for the other bands v The curve taken appears quite a.rbitrary, as an attempt wa.s made to weight the various points on a basis of how definitely the A2F (J) J + 1/2 Using the smoothed values determined the slope of the 1 ine through then:.. va.lues of D , the final values of B v v a.re sum.ma.rized in Table 17. tentative values of 0" These calculations In the second and third colu.Yffi1s are the selected, in the fourth column the smoothed and D v were obtained. v value of D and in the f if~h column the final va.lues of ','.< • v' v Also listed J..J are the values of B and D as calculated for v = v v - lj2, that is/B e and 0 De • From B ', we calculate e r 1 e = 1.92 A. There is no method of es- timating the departure from linearity of the de~·;endence of Bv' on v' + so no estimate can be made as to the accuracy of B ' and r e 1 • e In fact, all that c1rn be said is that the B ' and e r'e given represent the ex- pcrimental data over the limited range 6 <_ v' ~ 10. prob2.ble error in the B v -1 values is + 0.001 cm y2, The esti~ated , except for v' c:i 8, where the error might be larger. • t1 Taking Dv'"' = p~Dv'' it is fou.vid that B 1 8 Fron the Bv' of Table 17 we have f 2 l 10."' = 0.219. = 0.253, B B8 1 = 0.250 aJ1d r 10 2 3 = 0.222, which is satisfactory agreement. There is no simple way of predicting Dv' f'or the levels with v' = 14 and 15. A H'I From the few values ofµ2~ , available for these J + 1/2 bands it is estimated that 44. -..::::~.::.::;:::__F::..~--=:=-~==-"":..::..5=-==~'.=~::::-.:.-;:::.___.:. __::;:_;_-.:._::-rt~?:~ 0~ __;__~L=- .__:::...:;-:.=:.::..-:c.:....:..:.::---=-___:_:_~1.: ._.:. _:_ ___::_·--- -=:__;=_...:.:::.:..__;_-_...:.:.___.:.~-·-::·~ ~---~ ··-:1 1 _ -1 : -1 i -1 l B , CE . • rB 1 calc. cm-1 1) v 1 1 4B 1 cm I BD 1 cm ! SD 1 cm I v I v : v ' v , . v . 1 • II s mo ot ne d f i na.l v a lue !, t entative i tentative I I -·--·· · -------- -r------ -----··-··---1·-----=·~ ---- --+I ------~~- . 6 ·1 1.122 I i ! l 2.10 v 1.7 • 10 .... -5 I ·- - - -------·-- ---- - .. -- ··---------·------ --0.280 0.282 i. I i.o76 s * I (i.01s ) I 9 I 0.968 I 2.8 10=: 0.243 0.241 I 10 ; 0 .899 I 5.6 3.4 10 ..., I 5 .2 o--22-4 o .221 I o .312 1 r-·. --·-· -··---:--·~ 1. 493 4 C( 14 0~530 15 0 . 450 e -~ ' 2.0 .270 0.268 ! C-2. !J; • io) ---- (0.254) 0.255 i.5.10 1. 7 • 10' • 0.056 . 5 ; ~- --····--------·-··- -· -·-· ·------··---·---·----·--··- -·-· -1---·---·-· -·-·· - ·--·-----~-----: 1 !' : II a e • = o.013s -----·-·-t-----··--- ·---r------ ---·---- -·--·-·---·--· ·--- ---·-· ··- ···--·- --··-- ----· - 1 _J _______ __ ________ *Se e discu ssion in t ext on pg,, 45, v ca 1 c = 0.372 - o.Ol3 S(v' + 1/2) . fB ' -- i 0 .132 J _____ ~-·-~~-~ ·----·-·--·- ·----·---·-- -· ---· -·-·- --- - 45. -1 Bl 4 Bl 5 ' = 0 .132 ± 0 . 003 cm -1 = 0.112 + 0.005 cm l For conv eni ence , t b es e al s o ar e li s t ed in Ta bl e· 17. 1 t h at l\ is no H v e ry rap i d ly a pi-: roach ing ze ro. On e se es This c an b e exp l a i ne d cl a ssically a s r es ultin g fro m t he vibrationa l en ergy of t he molecule b e i n g v e ry ne a r th s dissoci a ti on limit, so t h at the particl e spends a l a rge tL~ie p art of its at l a r ge r wh e r e the pot entia.1 energy cha n g es v e ry slowly with r. The r e sults of t hcsse calcul a ti ons of B v ve.lues a.r e i n a g r eement wi th t h e value s for B' • B11 obtai ned in the d e t e r minat i on of th e band Th e v a l u e s of B 1 ori g i n s . - B11 foLm d i n thi s ma nne r , t a.bul a t e d Ln Ta bl e 12 ar o [ i v cm aga in in Ta bl e 18 , whe r e t hey a r e compar e d with the A 2F differen ce cf B 1 and B11 a s obtai-n.ed fro m the extrapol a ti on of £-.I v2 J + In vi e w of t he .i:J ro ba.bl e e rror }'r evious ly assi gn e d to t he quan.ti ti e s a.g r e e n~ ent listed , t he i s e ntir e l y s 2.t isf a ctory . It ha s b e en menti oned s e v E.ral times 8 ~ 0, ~; r e viously t h at in th e bands ( 8 ~ 0) l and 8 'C!-- l t he r= 311d ?. b r anch liLs s ove rlap . a.s ons f or and r e sult s of t h i s ove rl a pf i ng are h e r e des crib e d . c a lc ul a t e t he se;: a.ration of the line s R(J + n) and :F(J) • 9 The r e Le t us From Equ a tions we ob t a in -:i H(J -!- n) - F (J) 2l3 ' + (38 1 == = ( 2J +n +l) - 2 2 B11 )(J +nW. B '-B 11 )( J+ n) +( B 1 +B 11)J-( B 1 -B 11 )c2 [2'2 1 +n( B 1 -B 11 ~ ( 27) If t hese lin es of t he P and .' br anch e s exa ctly coincide , it is e vid ent bI = t h at we must ha.ve n n+2 B. -.-ihen t h i s r e l a tions h i p i s ap~:J ro x ima. t e ly s a tisf i e d, say 3 11 = n+2 B' - ~ n ' H( J -) - Fr J) +IJ. \ then = n q ( 2J + n + 1) ••••••• •••• • ( 28) • From the rotationa l sna.lys is i t is f ound t ha t B 1 ~B 11 a...'Yld n = 2 for t he 46 . Ta ble 18 From th·o Bn Band BI A¥£_~221'~_lue_~s__ - - - · -- - - · - - - - - - - 0 . 515 0 . 280 . 515 . 210 0 . 515 . 254 9 ~0 . 515 ic --- o . 515 . 224 14 ~ () • .515 .132 15 ~ 0 . 515 . 112 6 ~ o 7 +- 0 s ~ .l t 3' ··-·------ --·-·---- I i -·--··· - ··- - - - ·- & B 11 - -· · ··- - - B 0 . 235 0 . 233 . 245 . 243 . 261 . 258 . 272 . 272 • '- ........ ..i. 0 Q, . 288 . 383 . 384 . 403 . L1:.:07 0 . 509 . 254 ! . 255 .254 1 0 ~1 . 5C9 . 224: i . 285 . 287 . sc s :1 c . 253 \ I . 255 .248 . 229 • 295 r s ~ o) ( 10 ~o)f., . 5os 1 . 219 - - -____ _ ____ _ _ _ ___ __L _ _. I i _L__~------·-· ------···- 1 - - - - - - - · -- ----·--·--- , 8 -+- .L ·~ \ i 13"- Z/ 0 C8]c . I Fro1" j I : __ ________ _ _ _J 46a. It is obvi o ·~1::; t-na.t this relation cannot hold bands mentioned above. e xactly for both the bands ' 8 + - 0 and 8 ~ 1, for they have the same B 1 and diff'erent B11 • On examining the photographic plates it was observed that the lin e s of the ba.nd 8 ~ 1 were quite narrow, only slightly broader than normal single lines of the san1e intensity, but that the lines of the band 8 --E- 0 were very much broadened, ~ the lines for high J (J > 30) having a. width of about one wave nurnber. what one calculates from the values of B found. This is exactly de have -1 B 1 = 0.254 cm 8 B II 0 ~ 0.515 B 11 = 0.509 1 It is seeh that B 1 error. 11 is exactly twice B 1 wi thin l es s than experimental 8 For the 8 ..;.--- 0 band we c a lcul a te from Equation 28, putting n and J = 30, R(J+ 2) - r>(J) If R(J + l) - P (J) ~2 = 2 ~ (2J + 3) = 2(-.007) (63) =-0.9 is calculated from Equation 3 without neglecting the terms involving D1 and D11 , n it is found that even if 3 1 = B, n + 2 R(J + n) - P ( J) is not equal to zero but is a function of D! D11 and terms 3 in J up to J • This se paration is negligible for sma.11 J and need not be cons id e red here, but if it Hould c aus e the coinciding line s to be separat6d if th e bands could be followed out to J = 40. It is evid ent that the quant i ty determined by ms asureir:ent on a ba.nd with overl app ing P and R branch lines i s a ctua lly R ( J +n) + 2 F ( J) From Equat ion 28, we see tha t for n = 2, R(J +t~- r(J) approaches zero a.s J app ro a ches -3/2, and will be v ery s~nal 1 at J = 0. Hence it is ex ectecl that a.ny quantity involving R(J) to J or F(J) which is extrapolated = 0 will g ive very nearly the true value . However, any quantity 47. determinE::d from the r a.t e of change of the R( J) or ? ( J) , that is, any quantity derived fr om t he slo pe of a. curve of a. function of R(J) and P(~ p lotted a ga inst J is like ly to be in great e rror. founO. that the vs_lue s of B1 , and B11 Thus it was Z/ obtai ned f or the bands and 0 :o.enti oned. ':!e r e cons is t ent vii th the r es ults . on the othe r bands, but the values for D1 , obtained fro m t he slo pes of curves we r e meaningl ess . c a lcul at ing L'.l 2 F 11 0 va.l u es , where data is a vail ab l e fro m s e ven b ands , it Ll F 11 is s een from Tab l e 8 that t he 2 0 obta ined fro m band 8 ~ 0 do not de vi a t e systerr.a.t ic a lly for lo u J fro m th e .A F0 11 obtained fro m the other 2 ban ds , but that for h i ~h J t he r e i s a consider able deviation. the .4 2 F0 \J) In obta ined from th e band 8 ~ C for i gno r ed in calcu l at ing 3 11 0 Hence, J g r eat e r t h2n 28 we r e • 111. s a chec1-: on. the corr 3ctiless of the ~via l y si.s we n1i gh. t s ee 1-1ov.r well 3a.dger 1 s rule on ths r '3l ation of intornucl ear distan ce for frequ ency ,,, holds in this Gol e cule. The ru l e stat e s that r e where for ClF, a 8o l e cul e composed of one a tom fro m the first row end (c ij;\ i/3 • O. 5 3 s one at om f.. rom th e second row 01 t1e pe r iod ic t ao, l e , .D o.nd d .. = 0.9·1. lJ It ho.s been s ho Hn earli e r in this rape r that (Jj c f:=, 773 . 40 II Cr.1 -1 w' = 3 13 .5 e r 1 e = 1.92 From Er~P.J.at 2.on 15 , He obtain, i.ntroduc i nz av~.) ro p r i at e v 2.l u.es for the constants snd the r e duced mass 2 2 2 h e = 'in rr /,/ c ;,() e k 8TIJ ~e nce e = 0 . 133 • ti ~r iin • .) -~ = • 725 w ..... 2 ~ ayne s cm. . ' 7, 'fv e ' 0 . 0712 ·10 6 we calcul ate r C " C8 .l C • • re' cslc . = 2.23 48. no re8.l 3 i.=_,Tiif.icance ce.n be given to the values of w v. ' snd De 1 (or r e '') ' a s there is no re~son to bsli sv c th at the ex~rapol ations fro~ v' = 6 to = -1/2 are the correct one. v' r 1 e co.1 c. b~!dly The 19.rgs differ ence bet . . rncm r 1 e and leads one to believe that at l east one of the extrapol ations is off. It wou.1::1 sur e ly be possible to find va.LJ.0 s for ()J 1 e l3 3.Dd e I consist (mt \-:ith the: s p -sctr03cor ic d:;.ta. ;::rr:d satist'yi::.g 3ad6 e r' s ru".'.e, but for such value s to hav e ::--.ea.ning it HO'.Jld b e unique. H~s n <~c e ssary to s:10\'l tte solution As tht;re is no g:ceo. t L1t~rest i::1 th;:; v a luc:s of .3 ' and ,_; not thought wo rUrn!1 il c to VII. o.t te ;:r.r~ t W 1 e it , this. Tll c Chlo .cinc Fhwride Spectrun . De toi::c':lin3.tio::.1 of th:::; Dissoci3.ti'.:m 2nergy ar~d Potentis.l 8urve. The dissoci9.ti ·:m e n 0 rgy o-!: a 1uol c cul s , o. qu.2ntity of v e ry g r eat with ve17 gr0s.t a ccuracy sl; ectr c s copicn.l ly. The dissoci9.tion t:nt:::ct_;y D 0 is defi ned to be the energy T0qui:c c;d to s :;;i:,a.rate iYl-:,o t "\-vo normal ato::.'.1.s 9.. mol&cule in its loivest vibrational l·3 V G1. If a (v' ~ o) :; rogrcssbn is o os '~ rv ~ d ·11hi ch per:-u.i ts a re as one.b l e extra1; 0 l 3.tion -:o ::;, c.onvcrg c·m ce 1 i"Tii t, at wh ich ths molecule in th e excited state dissociate s i~to utcms , it is ss en that if ths Gnergiss rel 3t iv e tc the nor~~l atoms of the products of thi s d issocis.ti::m are knOi'm tb.:.: :lissocis..tion li::-.G.i t c 2. .n b s d ir a ctly ob- t e.incd. li~it is easily obtained. Follo ~ in g ( .t":'I' /!.' · AGy-115. 9 l7yf1-v-v lbr th:: usual procsedurt:/'~~r the- (v'~ 0) ;, ro g r e ssion i s p lotted ageinst v'. In ClF, th s conv e rgence i\. ·::; lot of the V8.lU '.:; S of the LI '.} 1 v + v2 obtai•.1 e d from Table 15 ags. inst v 1 49. is given in Figure 5. experi~ G ntal The the size of th0 larger circles p oints, the s~aller points are reprcsen~ing re ~ ressn~cd by circles, the proba.ble error iYl. those -1 circles representing a probable error of 0.1 cm ':'h e curve dravm £'or v'< 9 is the one ca.lcu1.a.tcd from Squation 23. 9 For < v' .< 12, th3 perturba.tions in the vibrational lev e ls are too large to :; errni t any re2.sonable curve to be dravm, but for v 1 ;> 12 a straight line fits all the points within exp eri~ental error. line cuts the axis AG = 0 at v 1 = 13.6. It is seen that this From this it will be concluded that the 17 -E-- 0 band is the last band of the progr e ssion. Allo wing for the rather large possible error in the :::isssurer.J.ent of the be.n d head. on the microphotometer trace and for the distance be hre en band orig:in and band head, it may be ~. rri tten v conv. l .· = 21,507.8 ± 2 cm-1 lrJ.. The positions of the band origins of th e ,., • lLtd '.~ ~ ~: rogress J.on r 0 ..... e V,/' against v 1 are also given in Figure 5. Bere the curve w2s dravm to cor- respond to ~qu.9 .tion 23 for v' for v' < 10, and to fit the exl:' erimental points >lC. There api; ea.rs to be a very faint narrow region of absorption im- mediat e ly to the viol et, of' the 17 ~ 0 band which might possibly correspond ·to another bru1d. 1"'1t It is too weak to say definitely ~ it. is real, so the evidence given by the is n0gl e ctsd. than .t1 Gv+l 2 p lot is taken a.'Yld this possible band In any case, it is sey; arated from the 17 ~ 0 band by less the probabl e error given for the convergence limit. ',Ie can i mn1ediately obtain D 1 and also D 1 , the latter being the o e dissociation er.ergy in th ·3 u~~Jer state ~easured from the miniYaum of the potentfa.l curve. Using Equation 23 to obtain the value of P" extrai)(>l- at e d to v 1 = -1/2 we obtain D I = 21 507.8 - 18561.6 = 2946 cm e -1 49a . 0 0 (0 (\j 0 IO 0 0 0 f'C') (\j 0 0 I") ~ .8 ~ s - b ~ +o ~ ~o 0 ~ 0 @ C\I v' Figure 5 . Det cr~n inat i on of th e Convergenc e Limit 18 50. D0 ' = 21507.8 - 18561.6 + l,;12( 313.5) - 1;"4(2.2 - lj's (. 4) = 27 9C cm-l A.s has be en 21entioncd b efore, the e xtr apo l a tion to v' == -·V2 i s quit e a rbi tr a ry, so th a t t he s.bov e fi gur e mi ght b e g r eatly in e rror. A r ea son- -1 a bl e v a lu e for the r rob a bl e e rror in D' is 50 cm 0 '.i. .11 . J.J ·~ 11 1I a n d D11) oo' t a.1.n o ' e it is nec e ss a ry to kno w the p roducts of th e dissociation a t the convergence lirni t. lfo attempt will be made in this t h esis to discuss the e l e ctronic lev e ls an d their dissoci3..tion p rodu cts, for that in its e lf is a subject for an entir e th e sis. b e ma.d e that c h lorin e fluorid e e l e ctronic sta t e s The assv.2Ilp tion will r e s e ~bl e t h e io d i n e chlor- id e sta tes and., 1 e s s clos e ly, t he chlo:cine, bromine a.i.vid iodine sta tes. -/6 ~ullik en has ~ iv en a g ene r a l d is cussi on of the halogen s pe ctr a , in which h e conclu de s t h at the a bsor p tion i n th e visibl e r e gi on i s due to a tr ans ition, Hith th e c a s e C c oup ling . T L.~ r .i. t.;; 1 <;' + ~ 3 Tl + ground state is f o r1·::.e d fro m t i·10 norme..l a t or::.i s , + both in the 2? / 3 2 stat e, whilG t he atom :Ln th e 2 ~· a l s o e xists a lev e l ar.. e x a:np l e of Hund 1 s 0 is f or me d from one normal 312 st ate, 3.n d on e e xcite d atom i n th e 2F 1 / 2 s t G.t e . Th e r e !Tr st at~·dhich li e s below th e 3 n 0 + sta t e and which3 is 3 r- foca.e d fro m two nor~ual a.tort1s in th e 2P / sta t e . 3 2 TT 1 The tr ansition g iv e s b ands h aving F , Q and R branches whil e t he transition :1 --E-- n; --..-- /s-.,. L .,.~ If~ giv e s bands having only F and R branches. In a.11 visibl e and infra-r ed h a lo gen s pe ctra , t he only st 2~t e s of th e h a lo gen atoms wh ich n e0 d b e consid e r e d a re grom1d state s, a ll other states having f :-: ir too hi r.:h t e r m value s. ,na 1 ogen avol!l -+. e r t e d ( 2P a' ouo' 1 e t s ar e inv a r a tions I - 7598, Br - 3685, 312 All th e 2 -, :: 1 A) , 1 o we r -+-1.-.1a:n. 1r ·,1i th s ep... -/ c;., Cl - 881 and F - 407 -1 cm The s pe ct ra of a ll i n t e rh a l ogcm c ompound s s tud i ed: a r e quite sirf.il a r to the hs. lo gen s :.; e ctr a. . l'T ifo rc at;<J in, in IC1 1 Brown and Gibson find two -..+ ' L. 51. ~ ' ·--L.+ an ct t o a 3 b an d sys t ems corresponding to a 3fl 1 ~ TT + ~ 0 ' ~~ + transition, the former having f , Q. and H branches, the latter only :: and Rbr anches . Also there is very good evid ence that the dis so ciation 3Tf/ products of the 11 excited state are a normal 2 :? / 3 2 iodine atom and an ex- cite d 2D chlorine a.tom. It is observed in the I Cl spectr\.:Lu that the • 1;2 ,.-, + + .._. , 2... bands in the 3 system b e come very diffuse for v 1 7 4 • This n- 0 B.rovm and Gibson asc ribe to the intersection of the potential curv e; of state by a repulsiv e O+ state arising from two normal a.to:ns . 2F'?;:::> 0/ - /8 Van Vleck concurs \vi th this explanation of the e:xperiment2..l ob- servations, both a.s to the ge:-oera.l nature of the states of the halogen aJ1d inte rhalogen mo lecul es and as to the nature of the p erturbation in the 3TT + state. Th e spectru.::11 of iodine bromide has also been inv e st0 igated by Brown'; who reports the spectrum to be siE1ilar in all particulars includL:g the ) Brturbation of certe.in levels in the the iodine chloride spectrum . 3T7 + state, to 0 As in the halof3en and io dine chloride s pe ctra, th e two upper states have different convergence limits, the ir 2 sep aration in IBr correspon:ling to the ? / 3 2 - 2 ? v2 ser. arat ion in bro- mine. It is evident that the chlorine fluoride transition reported in this th esis resern.bles closely the s.n d. inte rhalogen ""::olecules. 3n +.,, 0 - , ___ -t .. L. transl tions in th e halogen If a second ba.Yld syst em ha.d been observed in ClF with a convergence 1 imi t separated fro::n the observed convergence limit by the the abov e 2 ') P ; - . . . F l / separation in either chlorine or fluorin8, 3 2 2 assignr~ent could b e made vri th certa inty. Al thou gh this was not obs e rv e d (unfortuna tely, no serious attempt was made to locate it), the follo wing points of similar ity between the chlorine fluoride states and s pe ctrum and the halogen states a:.i.d s ~: e ctra C8.n be :·nade: - 52 . ( 1) The chlo rine fluoride bands have only F and H bran.ch es , a.s do the halogen a.Yld other inte.rhalogen 17 + ~ I: bands . 3 1 If one only 0 assumes the ground state of ClF to be 'I:+ the l ack of a Q. branch requires that the u pp er state be a.no+ state . '7.. ( 2) I n th e chlorine fluoride spectru:.n, as in the 0 ff; ~ 'L. band spectrum of iodj_ne ch lo r i de and iodi.ne bromide t here is a. p erturb at ion of ce rtain of t he vibr a ti onal l evels of the up per state. ( 3) In the u pper s t ate of ch lo r ine fluor ide a.s in the 3 [T 0 + s t a te of of the halogen and other inte rhalo gen 2nolecules, th e energy of dissoci at ion eof :i:4; is low compared to the energy of dissociation of the ground sta te. r 1 e over r e There i s a corresponding large increase of 11 • Since t here is no reason to b e li eve chlo r ine fluoride different-from the other molecules mentioned, there is sure ly no doubt but that the chlorine fluoride u pper state :i.s a 9 and one to one .... p / 3 2 3 rr 0 + state , 1·1hich dissoc i ates in- 2, . . ;;: 1/ 2 a.tom . It is stili necessa.ry to decide v:h ich of the two a toms i n chlor- ine fluorid e goes into the 2 ?1;/2 s t ate on the dissociation of the uprz per state . I t seems r easonab l e to assu.rne the products are Cl ( ..;Pr ) 2 2 and F( F ; ) f or the follo wing two re as ons: 1 2 (1) In iodine chloride and iodine bromide the dissociation of the 3 T1 o state gives t h e lowes t possible energy stat e for the divf 2 ided sys t em wi th one atom exd ted, na.n e ly, I( P / ) 3 2 Br ( 2 ? 1/'~. with Cl( ( 2) The two and Cl or I n ClF the ene r gy of the divided system is l owe st 2 ~) 3/ 2) and F( 2P l/2) • 3 TT + states of ClF formed from 2PV2 + 2? 3/ 2 will have 0 the s ame symmetry. For ls.rge int ,~ rnucl ear distance the separation is the d i fference in th e doublet s eparations of chlorine and 53. ' -1 fluorine, 4 7 4 cm • This is quit e s ma ll, s o that a s the atoms ap p roach one another t h er e will b e r e sonance be tween the states 2 r esulting in first a p proxi r;.mti on fro m Cl( P 2 Cl( P y 2) + F( 2P 1/2) and y 2) + F ( 2P3/ 2) • This resonance will c a use a sep aration of t h ese s t a t e s, c a.us ing th e f ormer (th e lower) t o be d e e p ened and It has b e en observ e d by l-'IulU.ken that the latte r to be r a is e d. in halogens the dissociation en e r gy of the r I' ·" .. . . 3/T + state ( vrhere, 0 . . of course, this r e sonance ca:rtn~~ex-i-a-t, t he re being only one (l - . _," is of the same order of map;nitude as the doubl et s p l itt i ng state) in the ha lo gen atom . So we find:- Cl Br I 3 150 3730 6850 881 3685 7598 F D I = 0 2D J. y2- 2p3/2 = 407 I n chlorine fluoride D 1 = 2790, just slightly small er than in chlor0 ine . ine, If the chlorine fluorid e state observ e d ris es from excited chlort~is D 0 corres p onds to the up ~ e r of the t wo possible state s, for which the pot ential curv e h a s be en r a is e d by its int e raction with the othe r ch lorine fluoride state so de cre as ing the depth of t he minj_mum . But then, the observed D 0 is much l a :c ger t han on e vrnuld eXI)e ct from the ab ove de. t a on th e h 0.lo gen s. On the other hand , if th e ch lorine fluoride state obs e rv e d arises from e xcited fluorine, this D 1 is for 0 the lo wer of the two p o,s sible st a t e s. He r e , n e.:;lecting resonance , a. -1 D 1 somewha t less th an 3000 cm ;,rnuld be expe cted, whi.. l e the r eo sonan c e would incr ease this de pth so ~ewhat~ This could easi l y g ive a dissoci a tion enerE.,y co'"!lp2 r a b l e to the obs e rv ed one . I n iodin e chloride , -1 3 + _. , r. '1': . '_~' D' is only 1060 cm for the sta t e , only one-fourth- the expect ed /To v a l u f_, .s.o _the v alid ity of the qu anti t e.ti ve comparison just ~'la d e is 54. v e ry doubtful. kno~:m On the other hand, iodine chloride is e xcited chlorine on dissociation. as de e p ~or in chlorine fluoride is as to g iv e Hence, as th e dep th of the min imum. deep er th an exne cted it se ems c er- t a in that the u p per sta te involved in the observ ed transitions is th e lo v1e r of the two states p ostul a t e d. Having decided on the di ssociation p roducts of the up ) er sta te as 2 being Cl( P¥ ) + F(;, l / ), the dissociation energy of the grotrad 2 2 state, measured from the first vibrational level, v 11 = O, is easily obtained:2 2 vconv. lin.- C \12 - ?3/2) //1 Do" a D 11 = 21101 ± 2 cm-l 0 F Also of interest is t he dissoci G.t i on Gnc rgy D/ of t he ground state as r;::ea.s un; d from the rn.inimuhi. of the pot enti 2.l curve. It is e vi d e nt t hat it is obta i ned from D " by addin ~ to the l at t er the zero po int e nergy. 0 tram Hence," Equation 2 ·w0 h s.ve ..... s.11 .that is obtained by the analysis of th e {,() 11 .s ~ sr e ctrL-li.1 i s G(1) - '"'(o) =11 - ,, ,, w e x, e 2 l() '' x 11 ( n e g l e cting (1it;he r ter::o s) • To obt a in U) and e e e . . , f ". !20.~ , '<~ :=: I.+ 4 :ze a.n empir1c 2 J. equation. o ·· bir g e is u seo. · Be On c a lcul s .ting t he v a lu es of t b& const ant halo [~ cns it i s f' ound to f or c1 , 2 1.01 f'or Br , 2 i n the, -; xr r e ssion for the 4e a:i:-iP'E cia.bly .:;mall er thru1 1.4, nauc ly 0. 98 1.15 f o r I Cl , a.n d 1.16 for r 8..b l s extra~; ol 2.tion f or ClF the vs_ lue C.95 i s t 31::en. 2: 11 e We frorr. t h.:; rota.tiona l ana.lysis, it i s fo und ths.t U) 3u bst i tuting ir.. the equation ~or G( l) - G( o) w . II :<: ,_., II .=. 9 • 9 cm-1 t;; {lj ..; II e 793.2 -1 cir. 111e ou t 2.in 2 • As a re a son- ae" and TeJdr._g 11 1f.J x e e/"'-"e 11 11 :::::: 80 . .I 55. and the dissocis..tion energy is:n -e II -1 = 21495i 3 er.:.; It is no w possibl e to C:raw s.. re a sonabl e ap pro:dm.ation tc the pot- ential curves of th e chlqrine fluorid e mol e cul e . Figure 6. The curves for 1 2.:+ and The s e are given in T7 + states ar0 obtained using 3 0 th(; 1-~orse function calculated from the ve.luss of B , W and D • e e e + Th(; repulsive 0 state , lHtG'·r.n dottoc\ is arbitrarily dravm. as all that is kno 1,m for it i9 its horizontal asymptote and the; approximate loc. .p l~S in t~ srsec t'ion Wl "th +vne a t ion O..t. •..t. 1 • 3 -,-r-+ I Io· -~ cuive. 'rh.P. - - l_.!'."ll'rst tl·.•,'o ui"bv rational levels are dr awn in the lowe r state, and the vibrational lev e ls v' = G and v 1 = 10 ar e indicated in th s U}") F e r state. It is seen by ap- plication of the Franck-Condon principle that the most intense trans- itions from v 11 v 1 = 3. = C should be to s.round v 1 = 5, EJnd from v 11 = 1 to about From tLc obs e rved sr: c ct:rum tl1 e bands ( 10 <-- 0) 8-Yld ( 8 ~ 1) are estimated to be the nost intens e bands of their res p ective v' progressions. 7hi s deviation is not surprising in view of the large from the simple qua dratic in v + 1/2 g iven by the Korsa pot ential curve. It is seen that if the vibrational qua.ntum nurnbers ass i. gncd to the lower state bmds a.re incr eased by one this deviation of the expe cted intensities from the obs 8rved intensities would be increas e d, for the effect of the new nt~bering would be to inercass the internuclear distance at which transitions would be Dost likely to t3ke place. 'o ihile ths postulated p erturbation of the 3 rr 0 + state gives a 11:uite reasona ble expla.na.tion of the diffuseness of tl~e bands 11 ~ O, 12+-- 0 and 13~0, i t does not e xp lain vrhy the lir..es of llf+-0. 1 sha.rf! n e e..r the origin, beco;ne cliffuse for J greater than 14. The i 0 0 -i 0 0 IO 0 0 0 ~ 0 0 0 0 ~ li £ =>o 0 0 0 C\I 0 0 0 Q ·1: .. · .~ . The Fotential Curv e s of Chlorine Fluoride Dis s ociation Froducts 2 -U p~ e r asymp tote--Cl( ~p / 2 ) + F( P ) 3 2 112 Vi br at i onc l l eve l s i nd i cated a r 8 Lowe r stat e--vtt = 0, 1 Upp er s t a t e- - v' = 6, 10 56 ~ 21 diffuseness ca.nnct be ca.us ed by dissociation by ro ts.tion, for th e rot- ational energy lev e ls Aff ected start thirty wave ntu:o.be rs be l ow the dis- sociat i on limit. On the ot her hei.nd ,, the effect of rotation on the eff e ctive potentia.l curve for vibration i s such that on the h i gh ene rgy s i de of a region of pe rturbation one expe ct s the lev e ls with low J, not those ui th high J, to be ~- erturbed . e::.~~:; lanation lfo other found for this obs ervation which is in any way satisfactory . co'.!lp l et e s the discussion o f the ClF s~- :. ectr1..un clusions:- ( 1) The band system observed results from a has been This To swr.:c:m rize the con- 3n o+ ""' . L - 1 )'+ ,_, transit- i on, the u pper state belonging to Hund 1 s ca.se c. The up pe r state is pe rturbed by a repulsiv e a+ state. ( 2) The dissociation products ar e :1 r;+ ground state and the o+ r epulsiv e state :- rz .:;r-ro+ state:1 1 (3) The constants for the ground state , determi ned from the l evels v n :: (' I.., J 1, are:D It 0 ~ II '"' ( 1) _ = 211Cl ± 2 cm-1 = 2 .616 vo lt s -1 D ti e :: B II e = 0.5183 :!: c.001 cm q eII - ~ ( o) r ·- 1 -1 0.0062 + 0.002 Cill z 793.2 cm-1 11 II e ·w'x e e 0 = 1.6255 ± 0 .002A 0 r 2.665 volts :: = 773. 4 ere-1 "ti fA) e 21495 ± 3 cm 0 = 1.6304 ± 0 .COLCI. II rv """" -1 9.9 cm 57 . ( 4) Th e constants for the excite d s t a te be lc w t he per turb a t i on r egion, d et e r mi n ed frO l:J. the bands (VI~ 0) With 6 ~ V I ~ 10 , ( 8 <--- 1) , ( 8 ~ 0) i,., and ( 10 ~ 0) l a r e :- ( l C 40:-- 1) , _, B e -· 0 . 37 2 cm ..i. r = 1.92 cm- 1 e -1 0 .0138 cm /Y '-'t e -- Fer the pro gr es sion v'~ O, V= 18561.65 + 313. 484 (v'+:Y2) - 2 .217(v 1 +1/2) 2 - 0. 400(v 1 + J/2) 3 We ~ ~ v~ J.3 . 4o4 cm-1 I: ..., / v e = 18956 ( 5) W 1X I -1 2 . 217 cm :: e e , ~ I , y I =- 0 . 4A 00 cm-1 ""' 8 6 cm·1 The re ma i ninE con stant s for the u ppe r state , det ermi n ed from the bands of ( 4) and a l so the b ends (v'~ C) with 11 ~ v' ~ 17 a.re :- V:.onv . lim. . = 215 07 . 8 ± D0 ' = 27 90 ..L ' D I = 29L1.i:6 e J G -1 cm -1 50 cm = C. 346 volts 50 cm = 0.365 volt s + I -1 Bl5 = 0 .11 2 c rn -1 J l~ :::: C.132 cm VIII . t} Tl:e Ch lorine Fl u oride Spe ctr um Some App lic ati ons It is of i nt er e st to compare the constants of t he chlor i ne f luoride e ol e cul e with t he const::mts for t he othe r hal ogen an d i n t erbal ogcn iao l e cul e s . I n Ta bl e 19 a r e lis t ed the val ue s of t he dissoc i ati on en- OJ, D0 , th e vibr at i onal const ant , er mr ti ,, We_ ti, th3 f orc e co11 s t ant , k , e and the inte rnucl ear di s tan ce r ", f or t h e s eve ral molecul e s . e e Also g iven a r e the mol e cu l a r we i gh ts an d t he r e· duced mass e s in atomic weight unit s . I n Figure 7 ul a r u e i :::;ht. ' h; 11 e ' r e 11 an d D ti a re p lott ed a g a inst t he rnol eco 57a Tab le 19 . D II o(volts) II -1 w e ( cm ) k 10 -7 e r e ( A) r.lo 1 e cul a.r ·.it . 1. 44 38 9.5 54.5 12 . 4 71 17 .7 115 . 5 24 . 6 ll o dymes cm . --a&-..: F2 OlF 2 . 616 793 . 2 4 . 59 1.625 Cl 1. 989 2 Br Cl 2 . 48 1 564 . 9 3 . 32 2 . 26 ( 430) 2 . 68 _ Br 1. 971 323 .2 2.45 2~284 I Cl 2 .153 384 . 2 2 .36 2 .3 21 IBr 1 . 818 268 . 4 2 . 00 I2 1 . 542 214 . 4 1. 74. 2 _,~ -- 160 I ;U- 40 .0 162 . 5 27. 8 ~- - - 207 49 .0 2 . S6 7 25!.J: 63.5 57 "· 90 150 2f0 58. Some regularity is s een in the v2.lu e s for k , but the values for e r e appe ar to be rathe r erratic. No attempt has been made to de termine possible exp lana tions for th e obs erved values. It is s een that all th e values for D 11 li e fairly near a s traight 0 line . However, it is probably mor e r ea sona bl e to drav1 a smooth curve through the points for iodine, bromine, and chlorine~ Then, the dev- iations from this curve for the interhalogen compounds can be int e rpret e d in t erms of difference in electronegativity of the two halogen atoL1s. So, in iodine bromide and bromine chloride, this diff erence is s rn:a.11, whil e in io dine prediction. chlorid e~ it is :;.:.iuch larger, in agreement wi th However, the point for chlorine fluoride devi ates only -P +h • ' '1 y no t· in ' a.ccora811ce ' s l i• gh..i. .~v l y .1.rom . ., e curve, d eciaea It has be en reported by Ruff and La.ass 23 of chlorine fluoride i s 25.7k cal. pe r mole. w i'th 22 n rormer l'd eas • ttat the hea t of formation This t hey determined fro m the heats of the reactions of chlorine fluoride, fluorine, and chlorin e with hydr ogen. They r ep ort:- ClF + H = HCl + HF + 58 .6 k~al. 2 l/2F l/2Cl Hence, l/2F 2 + l/2H 2 + i/2012 EF + 6 2 .3 kc a l ::::. 2 + l/2H 2 = 2 = HCl + 22.0 kc3.l ClF + 25.7 kcal. ' . a. mo r e accurate de tere 124 rep or t ing I n a. 1a t er paper, R\U f .-P. a.na, -Nenz mination of the above hea.t s of r eaction calculate~ E = -27 .4 kcal. for the for mation of chlorine fluori de . Since t h is heat of' reaction will be r r a ctically independent of t sm;, era.ture, it may be co:ubb.ed with th e spectroscopic values for the energy of dissocia tion of chlorine and chlorine fluoride to obtain the energy of dissociation of fluorin e. /?'rom Herzberg 25 ' we obtain 59. Cl 2Cl, 2 f hi.5 rlf!'sR~rcfJ/ Fro!J7 ClF .= D c = 57190 ca l .. D = 60307 cal Cl + F, 0 Cor.abining t he se l ast thr ee equati ons with th e pro per factors we obt a. in:- 2F, D 0 = 8.6 kcal. (0.37 volts) This is of a differ ent ord er of ma:;nitude from a ll va.lue s previously r eported. From the continuum obs erved in th s abs or ption s pe ctrum. of fluorin e it has been calcul at ed t hat D is 2.8 volts 0 26 , or 64.6 kcal. It is s een from Fi gur e 7 t hat any r ee..sonab l e extr apol ation of t he D 0 curv e vrill giv e a va lue bet ween 2.5 and 2.9 volts. Th e work of Ruff end his colla.bora.tor s on chlorine fluoride app ear s t o be we ll don e. The ir work on th e he at of r eaction of hydro gen and fluorin e a gre e s we ll \·: ith othe r det er mina tions. Th e s pe ctro- s co r ic va lue fo r t he dissociat ion en er gy of ch lorine is c erta inly corr ect, t he chlorine s pe ctrum having been thoroughly studi ed. The dis- sociat ion ene rgy of chlo r ine f luorid e obtained in this inve stigation can at most be high by the doublet s eparation in chlo r ine, or low by the cloub.let s ep aration in fluorine. These possi bl e corr e ctions are f ar too small to account for the discre r; an cy, b eing only 2.52 and 1.16 kc a l., r espe ctive ly. It is true that there is~a efinit c r eason why on e s}).ould be able to extr apolate from the oth er halo gens to obt 2.in a. dis s ociati on ener gy for fluorin e; it is a.l s o tru e t hat one cannot usua lly Dak e an entir e ly definit e calculation of t he dissociation. limit from meas ur e:nents on a. continuum, but it is difficult t o be li eve t hat either of t he s e methods could be off by th e ext ent indicat ed by the differ enc e bet ween 2.8 volts and 0.37 volts. Henc e , the conclusion i s re a ched 60. that the value of Ruff for the heat of formation of chlorine fluoride i~ far too la;ge, the error probably being caused by too low a value for the heat of the reaction•- Ha + C1 F := HCl + HF. This cone lusion is a.l s o reached by an cxaniina.tion of the heats of form- ation of iodine chloride, iodine bromide 8-11.d iodine chloride, res pe ctively 4.0, 1.7 2.nd o.7 kca.l. ,_. er rnole. Th ere see.:ns to be no ether way of obtaining a r eas onabl e va lue for the heat of formation of chlorine fluoride. cr,emic31 evidence available to I: ermi t a quantity. si~n.:; lc There is no other de termination of this The electronegativity of fluorine was selected by ? auling 27 to fit the previous values for the heat of fornat ion of chlorine fluoride and the dissociation energy of fluorine; the electronegativity va.lues a.s presently assigned cannot be used. It perhaps would be pos- si.ble to find l..L.Ylique solutions for the elcctronegativity of fluorine and the heat of formation of chlorine fluoride consistent with the date. on other fluorine compoi..mds. If the value for the heat of for mation of chlorine fluoride is ca.lculate d using the dissociation en ergy of fluorine obtained from the 8 bsorpt ion continuum, it is found that d BClF = +C .6 kca 1. ~liis too is certainly incorrect, for chlorine and fluorine form an explosive n ixture. Hence, it appears that the value reported by ~uff for the heat of forme.tion of chlorine fluoride is too hi:-;h by a rather large a!rlount, a:ic1 the value g:iveri by ·.fartenberg 26 for the dissociation energy of fluorine is a.lso at lea.st slightly too hi:-::;h. sa.id only tha.t It can be definit e ly Sl. + Jc~ .. 2 - 20 Ov.J..•' 01l'.;' 2 0 .... P. 1,11': ...t ._..1r. _. -AFJ..ClF :::: +1:e lv"Bt of -Porm...,t .....,_,!. ..... . . . . . c. _ o ·~ on of chlorine fluoride i n kcs.l • ... .,...&.. .. per mo le (at o°K, but a lso s.t roo?n tern er8ture with very little error). D ~ = the dis sociation 0~2 IX. ener ~y ~ of fluori.ne. The Chlorin e Trif'Juorid e Spectrum An attemr: t to find soro.e absorp tion bands of chlorine trifluoride was na.de . Th e chlorine trif'luoridc 1:rns Section II of this th e sis. '::i. th 22.stinan !l:O p lat e s. The ~uartz p:ce~12 r e.d a s describ ed in th e p rism s;sctromet e r was us e d 'rhc 3 :r:ieter sbson)tion cel1 2n:l auxili a ry system was exactly as described for chl o rin e fluorid e . Pressures from fiv e to s e v enty e x~-:; c)s;). r e S'Jry c e nt~neters:were us ed; the neces- ti:--ies varied. f r o ~:i one-ha lf to t en minutes. Un.der these condi ti01;.s, no b.J nds :.re rc f01.md in the :re,s ion 1 to A 5550. A4800 To th e r ed s idc of /\ 5550 the s ensi ti vi ty of th ::: Eastman this latte r regi on . To the vi8l et of \HS too g r :.:::a.t tc sbsoq:-ti::m by id e 2.::.lu C:J. Ch. chlo1~ine ch lori~e i~t~nsity ,c;r::nit dec i s i on .:: s t:) tte ~:res c::nc •.:;; or :J.bssnce of trifluorid e . trifluorid 0 t ai:en unch::r s J..m:.i.12.r co::_d i t i ::;::.-1s t1l:1t of the chlorine dioxii e bsnda arc upf roxi~ately the -~he s 2~ e on Tl: u.s i t th e ch l orine d io~i ie ~:i2.J b:; con cl v.d e: d tl:.a t the a~;-1o unt of' ch lo rin.8 di0 xi '.le i n ti~i fl u o.c i.:l e j_9 th e ss;:Jc a s t he a ,::cunt of th e chlorine in chl orine fluorid0 , n ~nc ly, on~ ~cr c ent or l es s 2s de ci2 e d in s ~ ct i on :rr of this the sis. if ~oss ib~ s , the TI0~ 2 nts of inerti a of chlorine trifluor i i e , t o do 30 ~ou l d ~e qu it ~ diffi cul t . qu.a:c tz :':tor:t the sJ .: ;t c:''.1 . ~indo~s ~n Li tl! i um fluorid e cou l d i.:- r obn.bly b e u3c:d f or th e abso r pt i on c e ll. I t is ross i b l c tl-:2.t 31 l th e obs ·2 rv e d y e llo 1 color of th e ch l orl~ e trifl~ori d e arcss fr o~ t h E chlorine diox1 L L:; , so t'.1.at t~ 1e f o rnu· compound ::-:i.;ht t!o t absorb in the n ea:c u l tra- vi o l e t . ·:-.'.l i [~h+, ~. I n t hat ca s e the R8.n-:::"?.n s ~ · c ctr w_-: o {' th e chl or in e trifl J.o:ride 0 be obtai nab l e , ) rovid i::;ci. only it d id not r03_ct with t he quartz The 2lectron Diffrection I nv e sti~ ati o n of Chlorin e Fluorids The inter nuc l eRr distan ce in ch l orine fluorid o a s det e r~ in o d ~:;.i:_; e ct.rosco _:? ic a lly h a .3 b <;e n g iv ..:;n C8.rli(:;r in t h i s t he sis ( 3~ cti o n VI). In chlorine trifluorid e , it is ex;G ctcd t ha.t t he t h ere i s no Getho d of s:..n gl e s ts b e F r e dicti~\C ·Jx ~1 ect 8 d. c ~ l o ri~ e -flu o ri ne the s.m0 mt of l engt hen ing Qr t he bond 1 Eer e is d s scrib e d s.n incl epend.snt cl e t e:rc.i nat ion of the chlorine-fluorine distance i n ch l o r::.n '~ flu or id e; 2.nd an ex s.~:l in- ation of the possible structur e s of chlorine trif l uorid e by the e l- s ctron diffra ction metho d. * ,/ith :Jr . Verne r F. I-I. Scho mBk.~ r. 63 . Th e e l e ctron :iiffre_cti on a,i:-p a.ratus us e d in thi.s i:'.1.v e st igation has bee n de scribed by Bro ckway 28 • The came ra dists.nce vrn.s a bout 11 cm. a n d th e wa ve l ength of th e e l e ctrons, d Gt e r r.Gine d in the usua l way 28 f r o:r:i trans mission photo[£ra{Jhs of s o l d foU ( a app roxi ~ at e ly 0 = 4 . 070) , was 0 O. OSA. The chl o rin e fluorid e and chlorine tr ifluoride samp l es ua e d we re po r tions of the s e c c ~T.po un d s p rei; ar e d for t he s pe ctrosco p ic i nv e stigation as described i n Se ction II of this t he sis . short l y b e for e u se Hl a. ~uartz The s ampl e s we r e d is tilled sy st em a.nd k e pt i n qua rtz trap s . The v a lv e d e scribed in S s ction II as value A of Figure 2 was used to control the introduction of t h e gas into the e l e ctron diffr a ction ma ch ine . Fh otografhs of ch lorine fluor i d e were taken with the tem; erature of the l iquid s a.mp le ro_n g ing fro .ca t h e t empe r3_tur e of li quid. air to -1C5°C; the vhotogr ap hs of chlorine trifluorid e we r e t ak e n with the temp erat ur e of the li qu id s 2.mp l e ran~ ing fror!l -75°0 to +12°C. The bo iling p oints a n d fr ee z i n g po ints of th e s e substanc e s a r e list e d in Tabl e 1. I r1;-vi_edi2. . tely b e:E'o r s e a ch s e t of photo gr aphs v;a s t ak en a sma ll amount of th e s a!-n.1J l e w2.s distill e d off in ord er t o r emove a.ny h i R:h ly vol a til e s u bsta nc :::s for me d as the r e sult of rea.ction of t h e s amp l e with the quartz of the tr8.p or t h e l ubric B.n t on the t ape rs. I t is b '2 lieved that n e ither 88.r,::1 l e contained a lar ge emotmt of i mpur ity. 3 oth th e chlo rin e fluorid e s nd th e chl o rin e trifluo ri ds c ont a ined some chlor ine d io xid e , but the amount of this 1 2.tt e r s u b stance , s s stated e arli e r in this thesis , is be li eve d t o h ~v e been th e extent of one pe rcent or l e ss , a n e gl i g ibl e 8Ftount . ~ r e sent to Th e r. : h oto- ,;r a phs of: chlorin e fl uo rid e 1:.9.V C a Si mp l e di a tomi C Fat t e rn, i r:.d :i.c ~d:, ing the '.3. iJ s en ce o f l ar ge amount s of the t etr a.f l uoride s of c 8rbon e.n d silicon. These 12..tter impurities are surely not rresent in the chlorine trifluoride because of the very larre differences in boiling Chlorine Fluoride. The photographs of chlorine fluoride, as has been mentioned, show S<?von 1 1essure ble r:i.ngs were obs "'::rved, for which the ~sasured s 0 value~ ninime. are civen in Table 20. ~wo correspond.in:; to the maxima and sets of ~essure~ents are listed, set A the avera~e of ~easure~ents on three photo3raphs, set B tbs &verage ~ado nine ~onths after ttosc of set A. v:=i.l ~).C::3 of -, In t~1c last coL.u~ll'l of J.'3_blc 20 in te rnuc 1 (; a:r d.istsnc& in chlorine: .fTJ.or- r cuterJ..a.ost fss..tures in on elcctroG diffraction r;hotograi:h arc often t.m- reliable. ths3e extreme vsi~0s, ons obt~ins the avera;cs N0glsctl~g given in Te.ble .20 for t;2.Ch of ths sets of .::i.e3)3U.rc::1snts, or combining the tv10., 0 r = 1.63 ~ C.OlA 0 8..tion., ·,·fr1cre it wo..r:~ found fox· the lov1est vibr8.tional level in the grou.nd sts.tc., r 0 = 1.6304 + 0.001. The agree~ent is excellsnt. Ghlo:cine Trifluoride. The photogrs.phs of chlorine trifluoride arc; :1uits s·:;tisf2cto:cy., ts.vine: seven i~ieasu.rsabls rings. si·c z;iven in Table 21. T~E: / lC , The r:,,essured aJ.O v~::.lu2s \ q = -fF s) c sti:nated intsns i ti es of tte; vJ..rious ~r.ax- r.c:e. and i:::iinina. a:rc giv011 in the ·coh.L'Tirl he9cled I, 2.nd the corrc:srondi:'.lg ,s. Table 20 ~·fax ~.· ~ in 8 r 0 .Set A 3 3 4 4 5 5 6 6 8.43 ( 1.669) 10.4:0 ( 1.656) 12.32 ( 1.653) l ·'± .36 1.638 16 .27 1.639 18.15 1.642 20.17 1.634 7 7 Average 1.638 ± 0.002 2 .. 99 ( 1.501) ·L 55 ( 1.698) S.53 ( 1.670) 8 .6C 1 1.635 10.47 1.645 1 2 .62 1.614 H .. 36 1.638 16 .47 1.619 18 .30 1.629 20.47 1.610 22.20 1.627 24.23 1.621 213.3 (1.61) 28.2 (1.62) Average = 1.626 ± 0.009 Set B 2 2 3 3 4 4 ,5 5 6 6 7 7 8 8 66. Table 21 i-~a.x ~: in 2a 2a 2b 2b 3 3 4 4 5 5 6 6 7 7 8 8 J ck 20 20 12.9 19 18 14.8 -2 2 17.2 4 4 21.4 -20 17 ( 1.029) 27.0 20 15 1.007 32.5 -20 +14 1.000 38.1 23 14 0.992 44.2 -13 ( 6) 50.7 15 ( 6) 57 . 4: -20 6 0 .994: 62.5 20 5 0.998 67.7 -18 3 0.997 73.3 19 3 0.996 1s.0 -16 2 83.8 +16 1 qo 7.46 1111 Average l* 2.40 5* 5* 30 o/qo = 0.998 30 43.4 -3 45.0 -3 49.6 2 51.4 4 *Values used in ca lcul atin:::; the radi8.l distribution function. 67. co effici ents 0f the r a di 8.l dis tri bution func ti on co lu..mn h e a d e d c1_( • of q o 29 a r e l is t ed in the I n the lO'i! er part of th e t a bl e a r e g iven the va lues and I ass urned for t he first r.'.la.x i rn.um a s used in the r adial dist- r ibution function . The f i fth minimum aJ1.d max imum we r e e ach r e:r: r e- sented by the t wo pe aks listed in the lower s e ction of the t able a s i t was poss ibl e the r eby t o bett er repr e s ent t he n atur e of the observ ed f eatur e s of the photo8;raphs. The calcul at e d rad i a l dis tribution curve , 0 R.D. of Fi gure 8 , h a s stronz ~:: eaks at r equal to 1.69 and 2 . 31 A, and sma.11 p ea}:: s , which may or may n ot have si gnific ance, at 2 . 69 , 3 .36 0 and 3.66 A. The l as t p eak listed cannot h ave si gnificanc e f or any str ucture having the t hr e e flu orine atoms s bout a central chlorine atom . It i s obvious that the d i stances 1 . 69 and 2.3 1 mu s t corres pond to •- chlorine-fluorine and f l uo r inc-fl u or ine dis t anc es , res pe cti v e ly, giving a bond angle of 87 ° . Re duc ed th e or e ti cal c 1rv e s were ca lcul- a.ted for symmetric a l pyr 2..mida l str .Jctures wit h F-Cl- F bond angles of 1 84 °, 8 5.5 °, 8 7°, 8 8 .5 ° and 90 °. Fi e;ur e 8 . These a r e th e curves mar ke d A in The curves we r e c 3.lcul a t ed using the equ ation zCl I( q;\ = ] Cl ·'\ !j1 sin - J: z l 10 q ClF + __F_ . ZF n z fr srn 10 q __Q_l- Cl l r:i'- F J.- 1 . is th e d ista nc e between the ith a.ncl jth atoms , Z. i s th e vrher e ].lj '-' a to;:nic nu.rnb e r of the ith a tom, / . i s the X-r ay sc a tt e r i n g f a ctor as "' calcul ate d by Pauling and Sherman, and. q = 10 rt s =\ 0 I\ sin sP/ 2, 'f = sca tte ring angl e . The a pp e a r ance of the photogr aphs i s best r e ~· r es ented by t he cuve A-87, with A-85 . 5 only sli s h tly l e ss satisfactory. T~1e doubl e t ap~j ea.r- a.nee of the s e cond max i mu:'.n, rep r es ent ed by rn ea.surements 2a and 2b in FF 67a. 0 2. l 3 4 RD A 94 • 85.5 87 88.5. 90 · B c 30 Figu.re 8 Curve RD - The r adial distribution function fo:c ClF'7. v Curves A, B, C - Reduced theor etic a l intensity cu r v0s fo r ClF A--Sy;:D.Yaetrical pyra::-"idal model ~·!i th indicated F-Cl-F bond angle . B--Asynnne tr ice l pyr amidal rwc.e 1-h:o F-01-F bond a.n ~,. l0s = 87 ° one F-Cl- F bond angle = 1C5° C-~Flanar mociel- -two F-Cl-F bond angl e s= 87°. 68. Tab le 21, is definit e ly not indic a ted by the curv 8s A-90° or A-88.5° curve. The ap:?earance of the fifth maxirm. un on the photogra1'J hs indicates this feature is broad, \·dth the p e2k displaced ~.JO large q. This ex- eludes the ..A.-8 i 0 structure. 1 From these and other less obvious deviations of observed and calculated curves, it can be sai d that if chlorine trifluoride has the syrm::.etric2~l pyramidal structure, the F-Cl-F bond angle is 86 .5° ±. 1.5°. However, there is no siapl e theoretical re a son vrhi ch i-·equires that the structur0 of chlorine trifluoride have the abov e symrn.etry. It r.light be ass·lk"'Yled that one of the two subsidiary peaks in the radial 0 distribution function, at 2 .69 or 3 .36A, has :aeaning. Thus o:ne ob- ta.ins the unsymmetrical pyramidal structur e (B) having two F-Cl-F a 174° angle between the outer fluorines. In 0 both these structures the 01-F distances are a.11 assU:ned to be l.69A The reduced theoretical curves for these structures, labeled B and C are given· in Figure 7. The second max:i.mum on curve B does not agree with the photographs, nor does the fifth mini::num and maximu..rn on curve C. However, the deviations ar e sma ll a_~d it is not possible at pre- sent to exclude all the non-syrnrr.ietr ica.l structures which ;.aight be obta.ined by slight rn.odifica.tion of models B or C. In the la.st colu.:-nn of Ta ble 21 a.re g iven the values for qVl.jo /q /· as it is equal to one within expe rimental error , the r adi a l distribution function ye~\. s give the correct internuclear distances. The final result then is:If t he chlorine trifluoride molecule has a three--fold a.xis, the 0 Cl-F distance is 1.69 ± O.OlA and the F-Cl-F bond angle is 86 .5 ±1.5°. 69. If the chlorine trifluoride molecule does not have a three-fold 0 a.xis, the mean of the 01-F distances is l.69A, and there are at least 0 two F-F distances of ap proximately 2 .311\.• It is seen that the effect of having three fluorine atoms bonded to one chlorine atom is to increase the Cl-F distance from 1.63 to 0 0 l.69A, or 0.06A, v1hich is a. reasonabl e a..rnount. 8 '7° ,~ . is . . ra·,th er surprising. The bond angle, only However, the bond orbitals for this ca.se have not been studied sufficiently for one to say what is a r easonable value fo:r the bond angle. Ackx1owl edgem.ents I wish to thank Professor .t~. M. Badger for the many helpful sugge stions he has ma.de and the continued interest he has shown in my work. I arn indebted to Dr •. L • .R. Zurn.wa.l t, D:i:. S. Rasmussen for much helpful discussion. .L!.. H. 11.'yster and 1-lr. R. The assistance given by Dr. V. F. H. Scho::naker was largely responsibl e fo r the completion of the electron diffraction work . lJ!.uch valuable advi.ce on the handling of fluorine was obtained from Professor D. M. Yost, who also kindly loaned the fluorine generator us ·sd in this research. Th e stands.rd tapers of copper prepar e d by ~r. Alex Logatcheff greatly expedited the ex;erimental work. Reprinted from T11c Jo u irnAL or- C H1·::-.1rcAL P1-1ys1cs, Vol. 8, No. 4, 349-350, April, 1940 Printed in U. S. A. ' The Absorption Spectra of Liquid HF and of its Aqueous Solution in t!ie Region Ai\8000-11,000 The absorption spectra of liquid HF a nd of HF in aqueo us solution have been photographed in the region /../..8000-11 ,000 with a dispersion of about 60A per mm. In the case of the anhydrous HF a copper cell was used, with windows of MgO (artificia l crystal) clamped directly onto the ends of the tube. The MgO is not attacked by anhydrous HF and appears to have great er mechanical . strength than glass. · The absorption band of anhydrous HF in the liquid state is very similar to the ba nd of liquid water at /..9722 as reported by Badge r and Bauer, 1 though it is somewhat broader. The position of the band maximum determined at . several temperatures is given in the following ta ble. The large probable· error indicated arises from th e difficulty in estimating the maximum of a broad band in thi s region where the trend in plate sensitivity is large. TA BLE I . T he maximum of the band of liquid I-IF as a function of temperature . t<° C) Amax(A) 19 0 -30 -60 -75 9940±75 9990 10140 10300 10380 Salant a nd Kirkpa trick2 have found that the third harmonic band of unassociated HF gas is locat ed a t /..8790. From the data just given it is evidept that th e freq uency decrease in going from the gas to the liquid is very large eve n at the boiling point, na mely a bout 1320 cm- 1 • shift in the HF freq uency, which must be due mainly to the formation of hydrogen bonds of th e type F-H· · ·F, is large in comparison with the shift in OH frequency on the forma tion of th e 0-H · · · 0 linkage . The third .harmonic OH band of methyl alcohol, for example, is found at · /..9490 in.the vapor 3 and at /..10,070 in the liquid, 1 so tha~ the frequency shift is a bout 600 cm- 1 • · · The large difference in the magnitudes of th.e frequency shifts just mention ed is very interesting in view of the fact that the energies of the hydrogen bonds in associated HF and CH aOH a re so nearly. the same, being 6.7 and 6.2 kcal. per mole, 4 respectively. The absorption spectrum,of HF in aq ueous solution has also. been examined at HF conc~ntrations of 25 and 50 percent and at 0° and at 25°C._ in no case does a cursory ~xam,ination reveal any significant difference betweE;n the solution band and the /..9722 band of water in: either position or shape. AUSTIN L. vV AH RHAFTIG Gates and Crellin Laboratories of C hemistry, California Institute of Technology , Pasadena, California, ' March H, ·1940 . J. 1 Badger and Bauer, Chem. Phys. 5, 840 ( 1937). 2 Salant and Kirkpatrick, Phys. Rev. 48, 945 ( 1935 ). 3 Badger and Bauer, Che m . Phys . 4, 469 (1936). J. L. Pauling, The N ature of the Chemical Bond (Cornell U niversity. •· Press, 1939), p. 313. Summary Chlorine fluorid e and chlorine t rifluoride h ave been prepar e d, the ir a bsorption spec tr a i n the r eg i on AA4 500 - 5 500 exrunined , and t he ir mol e cular structur e s inv e stigated by the el e ctr on diff ract i on nethod . Chlorine fluoride g ives a band system with singl e F an d R bran ches resul t ing fror.1 a TT + ~ '}:+ t r a.n s i tion with the conv e r gence 1 i mi t 3 0 for th e p r ogre ssion v 1 +- 0 at 2150 7 . 8 ± 2 cm2 2 products are prob ably Cl( P / ) + F( P 1/ ) • 3 2 1 • Tl1e dis soci ation The val ues of t he variou s 2 s pe ctroscop ic constants a.r e : D II - 21101 + 2 cm = 2. 6 16 vo lt s , D II i:::: :, :::: 0 e ~' II e r + C.001 cm -1 0 . 0062 ± 0 .002 cm 1.6255 ± 0 .002 ., e D I e B I e Cr.:l 2790 + 50 cm cm = 0 . 36 5 volts -1 .:;: 0 . 346 v olts -1 -1 = 0 . 0138 cm e' (X. r -1 50 C. 372 ;:;; _,_ -1 I 1.92 cm e 0 :::: 0 1 .. 630·4 ± O. OO l A ti) I = 313 . 484 cm - 1 w n = 793 .. 2 cm-1 e fA_;exe = 2946 0 0 ll II 0 . 51 83 -1 '\;ll r -1 21495 + 3cm = 2.665 vol t s, I T"' J.J II G -1 = 9 .. 9 cm W 1x e W 'y e e -1 I 2. 217 cm I -G.400 cm e -1 Chlorine trifluoride has no absorpt ion bands of aprre ciabl e intens ity in the region ,A). 48CO - 55 50 . The el ectr on diffra ct"on investi gat i on of chl or i ne fl uo rid e yi e l ds 0 r ~ 1.63 ± O. OlA, in ex ce ll ent agr eement with the s pe ctros cop ic va lue . The el ect r on diffr action investigation of chlorine trifluorid e indicates t hat if t he s t r u ct u r e of t he 1.1ol e cul e is s y1~7!l et ric a l py r am- 0 idal, the Cl-F distanc e is 1.69 ~ C.OlA and tl e F-Cl-F angle is 86 .5° + 1.5°. If the stru ct u r e doe s not have the above symmetry, t he 0 mean of the Cl-F distanc e is l.69A and there a r e at least two F-F 0 distances equa l to about 2 . 31A. Th e absorption s pe ctrum of liquid hyd r ogen fluoride e,nd of i ts aque ous solution have been inv e stigated in th e re gi on 11000 . ~~ 8000 - Physica l Conste.nts The va lues for th e physic a l constants us e d in this thesis were t 9Jzei1 from Lierz ber t:: 9 • He g: i ves: - Yo -10 2 l octronic ch arge e 4. 8029.10 Pl anck's constant h oC' . o,, 26, • 1..1.. 0~27 e r .'.; . sec. e.s.u. 2.99776.lO+lOcm. s e c. Velocity of li ght 1 16 ma ss of the 0 16 gt.om I J..6600.10 Boltzmann' s constant -24 gm. 10-16 erg . de gr ee • .._; v 1 • Drzpf\7 k a Conve rsion f a ctors:-1 Unit erg molecul e cm ('. -1 '-''m 1 er g pe :r molecul e i::: • .__: 1 J_.9863.10- 16 15 0...,.14 j L:.. • 11"' V\ , mole l el e ctron volt 2.8 581 1 .4389.10 -'- 1 ca l pe r 0.349 88 a , -17 6 • a.;i4._,8..J.0 8066.0 1..6022 .10 - ;~ 1 -] ') 23,053 r e du ced r..1a ss. reduc e d r1ass in a tomic we i r:ht lm j ts . h /"/.\. ,5 • 8894 . -• .1 u 1"'\~2 -U , W /-! ~'.. B e ::: h 27.994.10 I e cal mole -4.0 ~· - e 2d~ vn c cm. '"' 16 c l e r-tron- vott L 2398 .10 -LL A ~~· ·-2Ll~· 16 • l('\u 11 4.3378.10 1 -5 References 11 0 xid a tion Fotentials 11 , 1. W. M. Lo.timer, p. 46. 2. Bax Bodenstein and Helmut Jockusch, Zeit. f .. anorg. u. allgem. Prentice-Hall ( 1938) Chemie, 231, 24 ( 1937) • ·ilartenberg, Sprenger and Taylor, Zei.t. f. f,hysik. Chem . Bodenstein F,#estband, p .· 61, (1931). 3. Gale and 1-:onk, A stro~o h. lJournal, 59, 125 (1924); 69, 77 (1929). rhys. Rev., 29, 211 {1927); 33, 114- ( 1929). 4. O. Ruff and H. Krug, Zeit. f. a.norg. u. allgem. Chemie, 190, 270 ( 193-0) • 5. J. Simons, J. Arne r. Chem. Soc., 46 , 2175 (1924). 6. lfassachussetts Institute of Technology ;·Jave length Tables, John ·; /iley and Sons ( 1939) ~ 7. International Cri tica.l Tables (l,IcGraw-Hill) 8. H. 9. G. Herzberg, ~olecular Spectra and Molecular Structure, Vol. I, Diatomic ifo lecules, Prentice-Hall ( 1939) • 9a.. R • .s. l lulli.ken, Fhys. P.ev., 36, 611 (1930) 10. ·d . Jevons, i\eport on Band Spectra of Diatomic 1.:olecules, (2hysical 3oc iety, London, 1932). 11. Reference 9, page 114. 12. Reference 9, page 154; reference 10, page 212. 13. Reference 9, page 213; 14:. Hahn, Ber. Deutsch. Cher1. Ges., 71A, 1 (1938). 15. R. l'J~ . Badger, J. C11em . Phys., ~' 71 0 (1935). 16. D 4'. ,;:, 17. l~ • G. Brown anc '_·";:' . 18. J. H. VanVLck, Phys . Hev., 40, 544 (1932). 19. 'd. G. 3rmm, Phys. ~~{e v., 42, 355 (1932). 20. Re ference 9, page 114; ( 1930) Vol. 7, p .. 5. c. Urey and H. ,Johnson, £hys. Rev., 38., 2131 (1931). 1 ('< . I"Iul 1 iken, Fhys. .._.. ref erence 10, page S5. ~\e v., 36, 699 ( 1930) . Gibson, Fhys . Rev ., 40, 529 ( 1932) ref e rence lC, page 29. . Heferences 21. Ref e r enc e 9, p a ge 4501 22 . Faulinp.:, 11 r efer enc e 10, ~ ag e 205 . The Nat ur e of the Chemical Bond 11 , Corn ell Universi t y 0) p .. 4 9 • ."!_"):. res s (\ 1. . 0'2 .., .:.1.., 23 . C. Ruf f and F . La.8 ss, Zeit . f . a.norg . u . a ll t!,em. Ch e::ni e ., 183 , 214 (19 29). 2L O. Huff and 'ii . Eenz e l , Zei t . f. a.norg . u . all gem . Ch emi e ., 198, ~'575 (1931). 25 . Ref e r ence 9 , 26 . ·J a.rtenh e r g , S ~ ren ge r a n d Tayl o r , Ze it . f. Phy s ike Chem . _, Bodenstein- F e stb a.n d , p . 61 ( 1931.) • 27. Ref e r ence 22, pase 28 . L. O. Brockvmy, Re v .. lfod. Phys. , ~' 231 (1936). 29 . L. ::::-auli.n ~ a n d L. O. 3 rockway , J . Ame r. Chem . 3o c ., 57, 2687 ( 193 5) • ; a~e 485 . 58 ~ Frcrno s i tion.s di still i n.;; L1 anhyd r ou~;;; hydr oger.. £T. 1or i ch~ th::;.n to use the EJ.G thods ;r ev i c~ sly 2. sug~ s sted . d i~fr a. ct.1 ic r1 VL~. l~ e Usin.g tb.s c l ectrcn f oy- t he F -F diste~ncc i:n chlorine tr i fluoride , it can be sxgued tha~ th~ structure of this ·no l c cul e is v s ry l:Lke ly not 3. sy~·nrne tric 8l pyrarnidal. T1-:,:; a.pp lic: s:!:..icl'l of t he: ? rs.J1ck- Con.clon l.Ti:'.lcipl c t.c ~1 8.. l ue s of r ! ~2..lc u l a t e d fro ;J. 3 1 v - rl! 1 V3.lu·,: ; s as is do:ic b:/ c3 . Brm'm s.nd G. s . Gib son . !. ii . ir:. their discussion of' t~1e i .o dLi-:; ch l oride s r:c ctrum i s ':fH,ho J.t 1 , Their syst -::~. is un.nc c c: ss-?ry s..nd sur e ly i nc or r e ct. 1 • .i.·-·· ·' r1ys . ~ . , oc::n.as as a s e ~~ a r ate =i J:\ 8V . ~ !'"\ ~' 9 .h..; r: :::, ~ ( ,1. ().:;1 '-''="' 2' ... • 1 consid erati on of th e work of Badge r on the Sf e ct r wu of ~e thyl , s.l soho l vc.r~oi~ i n. t hs )hotographic L.1::-'rs.-r c d ..i.. 8l1c:l th e i.-ro:ck of Eo:cdsn <') ::'.nd 31 :r~ :e r ""' on -:.he f a r i r.. ::'r a.-r cd :~bso :c pt i on L1 8. i cc.t e: tb.:i. t it i'lou ld a longer a bsorpt i on ;at h . ,J... 2e 6~ 8. ( 1936) 4 J . Chern.• ·r--hys ., _, ,... I , 07.P' J. Chem. Phys ., 2..1 c:'. \ J. ~· v -- ) C '7 ....;.._; ~, A vacu'. .l['J t:J.h e 9..i;ipl ifi s r C9.n be d0s 2. ,::;ne d to op ero.t c a. ::.'.lc chan ics.l r c co r dsr d ir vctly fro rn. a thermocoupie without U i. 6 u.ss of ? l1J' gal vo.no·,:ic t cr- . It s hould b~ possib l e to obta i n j us t as gr eat s ensitivi t y as i 3 pos si b l e with any other nethod . 7 • . The 3-bsorpticn 3iJ8 Ctru1:0. of cyc lopro) enE:; i n the pl'lOtographic infra r e d i s int e r f.=: sting in tha t the " e thylene b and 11 does not appe3r , but two strong bands s.re oos s rved at long e r vmve length s. 1. ~. H. qst e r. J. Chei:i . I hys ., §_,10 (1 93 8). 8. Ii. consideration of t h e relativ.:::: r o..t c s of the forw a rd an d r overs 8 rea ctions in the equil ibrium 7 L J. ,i. (, w_··,+ul,.l i·"' ..' !. i·.R Y'rJal ~ -4 _ T-1 • T bond\, - : ' ( J."''-' '" . ., . - ~·J. 1'nt , vc. r·:1~1 i a. Lr:;.1. 'oo..,..,d) ci leads to an e xp l ru1.at ion of th·s o o s : ::rvation that fiv e- memb ered ring s containing a hydrog en bond are ;'10r e stabl e thm1. s ix membe r ed rings conta.inin g s hydro zc n bond. 9. 1 Cne should Dot b e suq:: ris e d, as ar e 3us v1e ll, l:: '3. ycock, end :ledebush , a t the l a ck o:f' an as::;ocia.tion bond in t h e a.bsorption s::;.,e ctrurn. of hydrosen fluorid e 13~C S-l. i~ carbon tetr a ch lori de solution . Ch l or i n e Tr ifluoride , the p erf e ct oxidant for rocket eng i nes . J. Chem Phys. 11.., .96Z (1940)