Photochemical Intermittency and the Photochlorination of Tetrachloroethylene in Carbon Tetrachloride Solution Citation Larsen, Delmar H. (1936) Photochemical Intermittency and the Photochlorination of Tetrachloroethylene in Carbon Tetrachloride Solution. Master's thesis, California Institute of Technology. doi:10.7907/qndx-yz67. https://resolver.caltech.edu/CaltechTHESIS:04012025-153556923 Abstract No abstract. Item Type: Thesis (Master's thesis) 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): Dickinson, Roscoe Gilkey Thesis Committee: Unknown, Unknown Defense Date: 1 January 1936 Additional Information: No page 5. Record Number: CaltechTHESIS:04012025-153556923 Persistent URL: https://resolver.caltech.edu/CaltechTHESIS:04012025-153556923 DOI: 10.7907/qndx-yz67 Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided. ID Code: 17122 Collection: CaltechTHESIS Deposited By: Benjamin Perez Deposited On: 01 Apr 2025 23:33 Last Modified: 01 Apr 2025 23:34 Thesis Files PDF - Final Version See Usage Policy. 8MB Repository Staff Only: item control page

PHOTO CHEMICAL I NTERM I T'l'ENCY AND 'r HE PHO'I'O CHLORI NAT I ON OF' TitTRA CHLOHOE'l'HYLENE I N CARBON TE'l'RACHLOR IDE SOLUTI ON Thesis by De l mar H. Lar se n In Partia l Fu lfillme nt of the Re quirements for t he De gree of Ma s ter of Science Ca l iforni a Ins titute of Technology Pasade na , Ca liforni a 1 9 36 - 2 - INTRODUCTION If a photochemical reaction involving an intermediate step or steps proceeds at a rate proportional to the rate of absorption of light to some power other than one, it is readily seen that if intermittent insolation is used, the reaction-rate will be a function of the frequency thereof. For, speaking rather loosely, intermittent light of frequency very high compared to the duration of the intermediate processes of the reaction will be in effect continuous light of a certain intensity; whereas intermittent light of very low frequency will produce effectively alternate li3ht and dark periods, the former of a higher intensity than the "continuous" radiation. Although the total amount of radiation absorbed is the same in both cases, different rates will be obtained because of the non-linear dependence upon light intensity. The work to be described was undertaken to verify the general results predicted by the intermittency theory as developed below, and to obtain some notion of the rate-constant of the assumed intermediate step in the particular reaction studied. This last possibility is perhaps the most useful and interesting feature of the intermittency effect, since it puts one in perhaps a more favorable position to discuss the mechanism of the reaction. - 3 THEORETICAL A. Derivation of the iaction rate under intermittent illumination. We shall consider the following general reaction: + \.,v ~ 2. A A ~ C, ~ c.. -t- A 2A ~ A~ A2. Now ~ ! "'·. k3. '2.,.. - ;l_Jl--= ~:s (~)(a) . (1) = 2. @ (I-.-,.), assuming for the moment that ~a.) each light quantum absorbed produces dissociation. - d<A) ~ dt = In the steady state evidently I/ -'4, (A) = ( 1: .J.1.) 'l. b.-. . whenc.(. finally ( 2) where Cio~s.> is the rate of absorption of radiation in einsteins per liter minute. However, if the illumination is intermittent (we shall assume unlformity of intermittency; i.e., all light p eriods are eaual in duration, as are the dark perlods), we must wrj_te t wo differential equations, one for the li ght period and one for the dark: Light: - 4 - Separati n g t he v a riables and integr a ting , we find: v/:J;: ) 1~':)] 2 +llMh-1 Gt.) Letting C( ~ l = + ~0nst. 1,., t '2~ for con venience in (J:.J.,) wr iting , we have: (5) Again inte g r at i ng , we h a ve: (6) It is evident that t he c onc entrati on of A bui l ds up in the lig ht and de cay s in t h e dar k in srich a manner as indic a t ed in fi gure 2 , ( figure l wi ll be dis - c u ssed in detail later), ult imat ely reaching a steady state , wh e re in t h e de cay e xa ctly balances t he growth. Leaving f or the moment t h e ques ti on of how soon this steady state is reache d , we sha ll enquire into t he limit ing v a lues of ( A) after it has b e en at t a ined, steady Il l umina t ion: Placi ng t-=- o when (A):: 0 in e quation ( 4 ), we find c ons t . : 0 ) whe nce (7) Int e r mi ttent Il luminati on With Very Long Peri ods. Let ~ and h'Y be the durations of light and dark periods res pectively . It is readi l y se en that the average value of (A) is the .area under the curve for one cy cle, divided by 'l' 4- h~ • In the limit this is seen to be: ( 8) Intermit t ent Illumination With Very Short Pe r iods . (A) wil l evident l y reach a steady state whe n 'l"~(slope of li ght curve vs . time) -:. ,..,.,..,.(- s lope of dar k curve). These slopes are g iven directly by e quations (3} and (5), which so lve to I ol ( 9) Thus if a rotating se ctor is used to produce the int ermi ttency , for a 1:1 se ctor we have from (8 ) and ( 9 ) , if A~ a ver age value of ( A), Slow rotati on Fast rotation A :: A - ---\ o(_ • I 2.. -:: .J_ · IQl. 1la - 7 - And for a 1 t 3 sector , i.e., lig ht : dark = 1 t 3, we ha v e Slow rotation Fast rotation .. A - A :,~·* -:: I -· -2..I oi.., But from (1 ) we see that f or g iven values of {B) and ( I ab&) to l-'-1} the ov erall rea ction rate is pro p ortiona l We must then der ive an expression f or A under v ary i ng condit i ons of ( I abs) , h, dnd ""'t-'. - From the figure, whi c h is drawn for a steady state , A is e vidently the area under two conse cu t i v e segments of the c ur v e di v ided by "Y + n"l"' . Light Peri od Area = r;;dt Chang ing t h e v arj_abl e from -t to ( A) by mean s of ( 3) , and omitting parentheses on (A) h ence for t h , we ha v e ~ 2. \~~ l A~ A I - oi.~ A>- d( at Ar) A, wher e as bef o r e a... = ,I ~':\ . y ('t .J.r..) This g i ves on inte gration ( 10 ) - 8 - Dark Per iod Us ing ( 5 ) we have ( 11) (12 ) Now putt i ng limit s into ( 4) a nd ( 6 ) we may write (13 ) l I - - - -; 2.1~ h'°Y A, Al.. Letting x -:: (14) V4 ~(tJts.) 'Y ; we may eliminate fro m (1 2 ) us ing (1 3 ) and (1 4 ). A1 The r esult is: A, =2 ~,'l'(o+i) l')!,+ ½e (1 + I: )o'..L ) 7 (15 ) ~"'~ 'J It is of i nterest to no te that ol.. A2. is the steady st ate value of A2 for any g ive n x d ivided by t he maxi mum v alu e of A for steady il lumina tion. Now what we shall me asure i s t he ra tio of the rates of r ea cti on under a g iven s peed of r o tat i on and under ste a dy illumination, which r a tio e quals A . Since As~OtA : 7)(1:....t-J A~k-Jj j ( --b."i from (7), we may write - A :: (16 ) - 9 It is p ossible to solve (13) and (14) for rL A2, The result is: where a-=: (n--t-l) +t""-X-tl)e'l.1' b - n-x:.(,-e"-10 ) C \- - e (17) 't. -,c. A We may then obtain a plot of A - vs. x for $~ any g iven n b y obtaining c( A2 from tl7) and substituting the value into (16). Fi gure I shows such a plot for n = 3, the v a lue used in the experimental work to be describ ed. A t a ble of /4 A2. and ~s~ for selected values of x follows: TABLE 0.01 0.10 0.20 0.50 1.00 2.00 3.00 5.00 10.00 14.00 20.00 22.00 26.00 30.00 34.00 40.00 100.00 500.00 I /(1(.A:l ¼~~ 1.985 1.858107 1.732 1.4457 1.1811 1.024 1.004 1.00057 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 0.4999 0.4995335 0.498 0.48914 0.466 0.42 2 0.392 0.3572 0.3192 0.3052 0. 2925 0.290 0.285 0.282 0.279 0.2757 o. 2625 0. 2533 - 10 We must now determ ine how soon afte r the star t of il lumination t h e steady state for (A) i s re a c hed . may derive from ( 4 ) , setting t -=-o when °'A:: o ( 6 ), setting t-=o Da r k period : J We and from v,rhen OLA= 1, the fo llowing ex nres s ions ~ r1-A = -v ,._ ~ .,., -+ I = -_-_-_-_-_-_-_-_------ 11V't("t...-...) b."i t 4- I which expressions ar e fr e e from 'T' , a s i s cl ear they must be . -t for We may the n p lot cu1° ve s for ( A) a ga inst ";t, ~ any des ir ed v a lu es of n and x , and starting wi th ( A) when t = O. =O 'r his has been d one in fi gure 2 , where the l9"er se gmen ted cu rve i s for x = 1, the smal l er for X.::: an d n = 3 in both cases. !o- , The dotted lines indic a t e t he upp e r limit s in the steady states . It is t hus seen that by changin g ~ only, the ...L 10 steady state i s quic k ly rea c hed i n about the same length if x i s v a ri ed fr om 1 to of time in each c a se, whi c h time is auite short . It is seen tha t be c a use the growth curve is a hyper bolic tange nt, for x =5 o(,A has attained 9 9 . 99% of it s maximum v a lue when t ="1-', i.e. at the close of the fir s t light period. Further , it is s een from fi gur e 1 t hat t he optimum range of x to work in li es to t he ri ght of x = l; so tha t one or two r e v olutions of the sector wi ll suffi ce t o effect a steady state for ( A). It is to be noted that the dimensions of x are those of pure numb e r. - 11 B. The Reaction Studied The photooxidat ion and pho toc h lorination of tetra c h l oroethylene ha v e been s t udied by Dickinson and his co-wor kers . 1 , 2 , 3 It has been found tha t c h lorine- tion does not sensibly pro c eed unt il al l the oxy ge n pre s ent has been reduced by a chlorine sens itized oxidation. The ch lorina ti on then takes p l a c e substantia lly in accordance with the e quation: (18 ) This r ea ction will pr e sent l y be treated in de ta il . EXPERIMENTAL A. Apparatus. The source of light was an atmospheric pressure me rcury lamp, air-cooled, whi ch was found to have a surfa c e intensity of ~4358 of 6 or 8 times tha t of an ordinary comme rcial evacuated arc, when oper a ting with ha lf t he p owe r consump tion of the latter. A p ortion of the arc 4x9 mm was used; light from this passed thru the se ctor ( p laced quite close to t he ar c), then thru a wa terc oo led filter whi ch isol a ted A4 35 8 1. Dickins on and I:eemake rs J . A. o .S 54 by means of Corning 3852 ( 1 932 ) 2 . Leermakers and Dickinson II 54 4 648 (1 932 ) 3 . Dic k inson and Ca r ri co II 56 1 473 ( 1932 ) - 12 - g l asses Novi e l A and blue pu r p l e u l tr a , a n d 12 mm . of 1 m. Cuz.S0 4 • 'r he li gh t was c ollimated by a p a ir of 8 d i op~er l ens e s , p l a c ed after the filters, a fter which a p ortion of t he li cht was refl e cted off at ri ght a n g les by t wo p arallel g l a ss p l a t es . This reflect ed light pass ed thru a c ompensa ting tank of xylene a nd on to a t he r mop ile . The d irect beam passed into t he thermost at , a b rass box fil l ed 1Ni th t hree windo·11s. one of 35 mm. integr a l Two dia phragms were provi ded; with t h e box, and one of 12 mm. which coul d be slipl)e d into a p ositi on c on c entric with the othe r. The li gh t finally pas s ed out onto a se c ond t he rmopile. The reacti on cell e mploye d was of pyrex g lass of t he sha p e below: - 13 The A1 s and B ' s a r e fl at glass p lates , with AA = 5 . 36 cm. and BB • 1 . 4 4 cm. Thus t he re is a con - siderable rang e of ch lorinee con c entration (6.07 0 m. to 0 . 014 m.) in whi c h the absorption of A 4358 for the .s h e.rt pa t h i s less t han t h irt y pe rcent whi le tr a n sm ission for t he long pat h is le s s than 70%. Th us it wa s a dvan - ta ge ous to c a rry out the rea c tion pas s ing lig h t thru the face s B ( d iam . of beam 35 mm .) and t o m<~ asure thEJ c h l orine c onc ent ra tion by trans mi s sion mea s u re ments t hru t he faces A ( diam. of beam 1 2 mm. ). A mirro:r was p l a c e d b e h ind one , of the faces B whe n t he rea c tion was being c arried out, t h us mainta ining a substantial l y uniform intens ity of radi a tion thruout t he cy l inder illuminated . rrhe cell wa s so mounted t ha t it c ou ld be easily turne d thr ou gh 90 ° in ea c h p o s iti on one of t h e f a c es A or B was with1n a fe w millimeters of the diaphl~agm used . A. g l a ss bead C v7as enc l osed in t h e c e ll to pro - vide stirr ing . Long c ap i l l ary tubes D made possible rela t i v e l y permanent remov a l of exy g e n from the ma in body of t he c e ll. The rotating se ct or wa s d ri v en by a s y nc hronous motor t h ru a gear c hain . :C i ght e en different se ctor speed s of fro m 1. 736 to 562 . 5 r . p . m. were p ossible . ·:v01•k All of the to d a te has b e en c arried out wit h a one t o three s e ctor . The t a n k wa s p r ovid e d with a stirrer , heati ng coil , and the rmome t e r . Th e t a n k wa s fill ed with xy l e ne, - 14 xylene , g lass and c a rbon tetrachloride h a v ing nearly t he same indice s of refra c tion . Prelimi n a r y illumination wa s c a rri ed out thru a window in one end of the box by me ans of a 500 watt projection l amp , a one in c h lay er of 10% cu pric chl or ide solution serving as an inf r a red filter . 'I'he filter system was found t o b e effecti v e by s p ec t rosco p ic examina tion; the ap p aratus was tested as a phot omete r with K2 Cr 0 4 and proved satisf a ctory . u, The absorption coefficient of Cl 2 ~ CC1 4 was determined; subsequent ± i J tration of t he solutions ga v e E = 2 . 17 ::t - ic.d ~. ~10 defined by - ( in mo l s /1 i t e w, din cm. ) . Leer - makers and Dickinson2 found 2 . 14; Dic k ins on and Leerma kers l found 2 . 26 . Be e r' s Law was tes ted by a dilu ti on of the solution used within the r a n g e of Cl 2 concentra tion s subsequently emp l oyed and found to hold . The c hlorine , CC1 4 , and C2Cl4 we re purified as des c ribed by Dick ins on and Leerma ke rs . l The the rmopiles were c onnected to t wo D' Arsonv al gal v anometers with a c ommon sc a le . were teste d for linear ;+j and a ppl ied to all readings . The galvanometers proper corre c tions we re During the exper iment s illumina - tion wa s pro v ided by a y ellow safe l i ght found to tr a nsmit no v isib l e radiation absorbable by c h lorine . The experiment s were c a r ri ed out as follows : Known amounts of C;2Cl 4 and Cl2 in CC1 4 we re d iluted with CC 1 4 to a kno wn v olume ; the CE:,11 was fille d , and after - 15 in some c ases pre l iminary e x posure to sunligh t for½ ' to l ' exposed to t he proj e ction l amp , wit h shakin g eve ry minute . The Cl 2 c on ce ntration was follo wed by means of an automobile head li ght bulb replacing the mer c ury l a mp . As soon as the fo rme r began to dimini s h, short exposures with c a r e ful shaf:.ing we re made until it was ascertained that t h e c h lorine was d isappearing uniformly , thus i ns uring comple te remov a l of oxygen . The mer cury l amp wa s t h en star te d , a nd the c1 2 me a sured , us ing t hree 20 11 e xposures tbru t h e long p a th. Illumina tions of from 3 1 to 1 2 ' we re ma.de , usua l ly ea c h at a different sec tor s peed . "Infin i t e" sec to1• speed wa s obtained us ua l ly by means of a wire screen of 24% tr ansmiss ion . Th e c h lorine c oncentrat ion was of c ourse d e t erm j_ned b e t ween ea ch run, and t he solution wa s stir red befor e and after e a c h det e r mina tion . It was f ound pos sible to make a satisfa ct ory corre ction for t he reacti on c au sed by t he meas urements of Cl 2 con c entrat i on , s ince each succeeding ga lva nome ter s wi ng was found t o suffer a n early c onstant increme nt . Corr e cti on was also made for t he unilluminated p ortion of the solution under the a ssump t i on that no mixing occ urred duri ng the run, a n ass-:.n;1pt i on whic h of c our se be c a me le ss v a l i d as t he t i me of the run wa s i n cre ase d. TREAT~ENT OF DATA If Ic i s t h e s e ns ibly c onstan t ( be c a use of - 16 - t he mirror) inte n s j_ty of illumination pre v a iling thr u out t he solution , we may r e adily d erive t h e following ex p r ess ion : () , . c ·("n -v .O.iffl ..La..t-. : .1.c,. ,;,.. u.z.) .A,~ ,ye 10 . whe re A)t., i s t he path leng t h . If I o j_s furt h er t he num - be r of einsteins per rn.inute entering the reaction vessel from the lamp , we may set whe1·e f3 i s e v idently a function of the tran smis s ion of the solution . Tak ing 0 . [~ 75 as the r e fle c tive power of si lver for A4358 , we may de rive t h e follovYing e mp i ri c al e xp ression f or the a vera ge p correspondin g to a given We then may derive the following expression : ( 19 ) We may inte g rate t he rate e xp ression given by (1 6 ) , r egarding ~ as constant thruout a run . ( 'l1he s mall v aria tion in ~ is over shad owed by v a riations in 1 0 , f or example . ) R= We der ive the following expression : I (~i.) ;!a. 11.-Y l~ro where t:.. t (Uh.), •Ta:. e:. t is t he time of the r un in minutes . The da ta of a sample experime nt ( 5 -12a ) ~ produced be l ow, where I 0 ( and thus k ) i s expre s sed in arb i trary units ( crn . of g a lvctnome ter deflection .) re- - 17 TABLE II (Cl "G ) before illumination = 0 .158 5 m. II ( C2 Cl4) 'I'emp . .Jhill.. "Y 1 0 ~~5 . 0 At II 1 . 370 - 26 . 0 ° c. 'Io Au. Io 3 . 00 1 4 . 58 1/20 . 85 3 . 06 1 0 4 . 70 3 . 25 1 4 . 38 1/46 . 88 4 . 09 1 0 4 . 54 7 , 50 1 4 . 40 1/46 . 88 9 . 47 0 . 856 1 . 24 0 . 0105 0 . 630 1 . Z>4 0 . 007 2 0 . 9 59 1.43 0 . 0 10 4 0 . 822 1.51 0 . 0067 2 . 07 3 1. 60 0 . 00 91 1 . 926 1. 68 0 . 0064 . 0 1 64 . 0206 4 . 65 4 . 66 ~ . 0 281 . 0310 4 . 64 6 Au./3 . 0309 . 0387 4 . 53 5 ·Y~ . 0462 . 0509 4 . 54 4 -►.. ( (<.t ..)~ - QJ, . 05 43 . 0634 4 . 58 3 cu~) . 0847 4 . 78 4 . 81 2 m. . 0124 The runs vv i th 'l' = o vrere with a s c reen of 24% transmission ; the av . Io was thus reduced by 49'b to correspond to a 2 51; sector of infinite s p eed of rotat i on . The fact that t he (Cl 2 ) is less a t t he end of one run t han a t t he beginning of the next a r ises from the mi xing of pr e sumably unillumina ted r ea ctants bet ween runs by shaking . Fr om a c ause not understood , but due pe rha ps in part to c onve c t ion i n the solut i on during t he run and in part to approxima ti ons i n the method of tre atmen t which wer e we ll ni g h una v oidable , t he k ' s found with the - 18 screen, which shou ld hav e been constant , fe ll off with decreasing c h lor:Lne c oncentration and with in cr ea sing ~t. 'l111e decrease wa s as mu ch as 1 8% for the ver y lo west ( Cl 2 ) and long est runs . This phe nome non ap peared to be a func- tion of bo th At and ( Cl2), for when an exp eriment wa s c ar ried out in which as the ( Cl2 ) diminished the length of the runs was also dimin i shed , thu s reversing the usua l treatment , the k 1 s tended to b e constant . This was a p p arently n ot due to a lack of C2Cl 4 , for Leermakers and Dickinson observed no such effect working with the same amount s of Cl 2 and C2 c 1 4 ; further , increasing the r a tio of C2Cl 4 to Cl 2 initially present by a factor of two in one expe ri me nt (5 - 1 8a ; see table III ) c a used no differe nce in be h a vior . The r mal reac tion of ap pre - c iable magnitude would have caused an increase in the k obser v ed with increasing time of run , instead of a decrease . The effect was still pres ent when the k 1 s were calculat ed neglecting the r ather lar g e corr e ction for unilluminated solution . No correction wa s made for therma l reaction; t h is was a ssumed to be ne g li g ibly small , in a ccordance with t he r esults of Leermakers and Dic k inson . These workers found the thermal rate to be of such a magnitude that probably less than 1% of t he chang e of chlorine concentration in the present experiments could be ascribed to this source . A hi ghl y empiri c al but prac tic ab le me thod of - 19 - correlating t h e v a rious k t s for 'Y-:o with ~t (C1 2 ) was found to be the following: a g ainst(a~¼t . a nd The k's a re plotted All of the points corresponding to a g ivenAt are t h en found to lie on nearly strai ght lines, all of which have the same general slope. By this de- vice we can then c a lculate what the k for~-: o would have been for any particula r run in which the sector was rotated at a finite rate. A Thus for each s u ch run a Va lue of A-;- 1 stea:a-:r \ Ct~ was obtained. It will be rec a lled from our p revious de•iations that for a 1:3 s e ctor A/Asteady varies ,. from O. 50 to O. 25 as 'l' goes from O to 00. This was found to be the case. Since x=-V1t1a. 11 c~'l'and log x = log 2. -t- log'Ji"-+ log(Vt.&»s'T'), k.. c a n be found by the difference in abcissae re quired to bring about coincidence of a p lot of log x vs. 1/Aste a dy a s calculated from figure I, and a plot of lo~,r)vs. A/A steady as determined from experiment. Such a plot, with the curve in the p osition chosen as best, i s shown by figure 3 . There follows a table presenting t h e essential data of the individua l experiment s . For tho s e runs where '1' -::j: o , l o g ~ 'I-'), t h e calculated k for 'r' -::. o , and the value of A/A steady therefrom are given. Columns 2 and 3 g ive the concentrations of chlorine and tetrachloroethylene in mols per liter with which the solution was made up; c olumns 4 and 5 the averag e intensity (in cm. of g alvanometer deflection) and chlorine concentration - 20 - prevailing during each run; column 6 the reci procal of the duration of the light flash in minutes; column 7 the duration of the run in minutes, column 8 the observed reaction rate constant, (Iabs) being e xpressed in cm. of galvanometer deflection, column 9 the calculated k which would have obtained under similar conditions but with infinitely fast sector rotation, column 10 the sum of the logarithms of the square root of (Iabs) (the particular values quoted must be multiplied by log@_ 10 and divided by the volume of illumina ted solution in liters, and moreover include Io expressed in cm. of galvanometer deflection) and1", and column 11 the ratio of the rate observed to the rate under similar conditions but without the sector. - 21 - TABLE III Expt. ~~ (c2 c1 4 ) ..:9__ (Cl2) '/-y 4-24a 1 0.1220 0.'783 4.23 2 4.23 3 4.23 4 4.42 5 4.37 6"t R.. .01015 .0563 00 31 .0405 2250 4.05 1 .01122 .0292 bO 51 .00985 .0209 2250 61 .009'70 .0140 gio 81 .01118 \l (,. :o) to-,~rr) A/A,~~ 31 4-25a 1 0.1220 0.783 4.25 .0481 00 2 4.23 .0355 750 51 51 3 4.07 .0251 e,o 4 4.34 .016'7 '750 91 5 4.42 . . 0112 1/4 16 1 .01020 • . 00976 .01057 3.0488 .457 .00908 .009'73 .00523 4-30a 1 0.1220 o.783 4.35 4.57 2 3 4.46 3.94 4 31 .0451 oO .0333 333 4.!. I .0243 1667 5l1 2 .01'72 00 8' .00953 .00970 .01010 3.4014 .480 .00795 .00933 3.6435 .427 .00901 4-30b 1 0.1220 0.783 4.10 4.03 2 3.91 3 4 3.72 5 4.60 6 5.00 .0593 1667 31 .008'74 .0468 333 31 .00920 .0360 166!7 41 .00926 51 .0268 00 .00997 .0188 333 '7.17 1 .00914 81 .0130 di" .00895 .01025 3.7767 .425 .00979 i.4346 .469 .01025 3.6866 .451 - .00840 3.3032 5-4a 1 0.1220 0.783 4.05 3 : 89 2 3.96 3 3.75 4 5-5a 1 0.1587 1.175 5.36 .0421 4688 3.98 1 .00668 .01057 2.3324 .316 4.68 .0296 00 61 .01010 2 4.64 .0207 4688 9.13 1 .00572 .00933 ~-5263 .306 3 5-'7a 1 0.1587 1.370 5.28 .0283 4688 6.05 1 .00764 .00995 ~~2589 .384 4.73 .0196 00 7.04 1 .00849 2 .00423 .00870 2.1186 .242 4.'76 .0149 4688 8.9 1 3 .0388 4688 4' 61 .0283 00 .0195 93.72 9 I .0130 00 12 1 .00814 .0104'7 ~.2577 .390 .00936 .00779 .00933 3.8293 .428 .00815 5-12a 1 o.158'7 1.370 4.58 4.70 2 4.38 3 4.54 4 4.40 5 6 4.65 .0695 00 31 .01053 .0540 2085 3.06 1 .01000 .01019 2.6955 .354 .0419 00 3.25 1 .01034 .0329 4688 4.09 1 .00675 .00990 ~-2543 .339 .0234 00 7.50 1 .00914 .0162 46.88 9.47 1 .00637 . 008'79 2 .1287 .362 5-13a 1 0.158'7 1.3'70 4.76 4.47 2 4.50 3 4.25 4 4.47 5 .065'7 2085 3.0'7 1 .0507 00 31 .0385 93.75 41 51 .0281 ~ .0176 93.75 8.02 1 .00690 .01048 2.'7267 .329 .01038 .00840 .01040 3.9782 .401 .01018 .00758 . 00898 3. 8363 .422 - 22 - TABLE III (Cont'd) Ex.pt. (Cl2) ( C2Cl4) I 0 --- (Cl 2 ) '('r' A"'t 5-18a 1 0.0900 1.568 4.58 4.60 2 4.23 3 4.28 4 .0378 CJP 41 .0282 4688 5.54 1 7.17 1 -0203 t00 .0152 4688 7.51 1 5-25a 1 0.1587 1.370 J.013 0003 2 3.70 3 15.52 4 1552 5 15..94 6 .0598 oo .0448 ~ .0316 oo .0226 oo .0160 00 .0124 00 6.5 1 91 6' 31 ~' 3½' R. Rl'l"=o> to,(~~) A/As~ .01038 .00695 .00996 2.2304 .350 .00673 .00581 . 00867 2. 0997 .334 .00979 .00839 .01017 .00885 .00985 .0056 - 23 - No determination was made of the absolute light intensity corre s ponding to unit ga lvanometer deflection; instead this quantity was calculated from the rate constant determined by Leermakers and Dickinson2. When this value (5.1 in mols, einsteins, liters, and seconds) is used it is found that one cm. of g alvanometer deflection, that is, one unit of Io as used above, corres ponds to a tot a l flux of 6.54 x 10-8 einsteins per minute entering the reaction cell when the 35 mm. diaphr agm is employed, if we take as a mean re a ction rate constant .e1e1 in the arbitrary units heretofore used. Using t he same beam splitting device and the same galvanometers and thermopiles, though p ossibly in a permuted arrangement, Le ermakers and Dickinson found a flux of 13.56 x 10-8 einsteins pe r minute passing a n equal area in a similar position. This difference could probably easily be accounted for if we ta ke into consideration the different sensitivities of the thermopiles and galvanometers, and the variation caused by slight changes in ali gnment of the different pieces of apparatus. Using then t h e v a lue thus obtained for the absolute light intensity, we readily find from fi gure 3 the following va lues of k4, where the first column of figures g ives the difference in abcissae of the calculated curve and of the plot of the group of points (~is in minutes, and (Iabs) is only parti a lly converted to cm. of ga lva nometer intensity, as expl a ined in connection with - 24 - column 10 of Table III), and the second column. the values of k4 in cubic centimeters per mol second: TABLE IV Difference in logarithms -3 + o.56 Best fi't Extremes: sec. -3 + 0.38 -3 t 0.60 DISCUSSION OF RESULTS Two mechanisms have been proposed for the reaction studie4 ? They are: Mechanism A Cl2 + hll ➔2 Cl (1) Cl + C2Cl4 ~ C2Cl5 ( 2) C2Cl5-t- Cl2 -+ C2Cl6 + Cl ( 3) 2 C2Cl5 ➔ 2 C2Cl 4 -+ C1 2 ( orC2Cl6-+ C2Cl 4 ) ( 4) Mechanism B Cl2+ hv ~ 2Cl ( 1) Cl + Cl2 ~ Cl3 (2) Cl3 + C2Cl4 ~ C2Cl6 + Cl (3) 2 Cl ~ Cl2 ( 4) Now it will be recalled that k4 was obtained from an expression of the type 7Jb.t.tT~'Y, under t h e assumption that the quantum yield of the first reaction in the mechanisms above was unity. It is readily seen, however, that if - 25 - this is not the case, the values of k 4 above are really those for r1k4 , where r 1 is the primary quantum yield. The much disputed question of the actual occurrence of such a "primary recombination" of halogens in solution has not been settled 4 • The presumption is of course that the dissociated atoms recombine before they have had a chance to become separated by diffusion. We may calculate in the usual way (see for example ref. 3) the specific collision rates at 250 for the reactions of interest in the mechanisms above. Using the estimated molecular diameters of Dickinson and Carrico3 and taking 5.8 ! as the diameter of 013, we find the following collision 1•ates, in mols, c. cm., and seconds: Mechanism A Mechanism B 2.ox1014 2.1x1014 9.5xlo 13 l6.4xlol 3 If r 4 is the fraction of collisions of the reactants in (4) which are effective in producing combination, we may readily derive the follovving, using the 11 best 11 value of (Mech. B) _l_ Thus r 1 and r 4 lie somewhere between unity and 1.oO I or~'-'o depending on the mechanism chosen, and subject to the restraint of the equations above. There is no reason to suppose that r1 and r 4 are greater than unity. 4. Fr~nck and Rabinowitch: Trans. Far. Soc., 30, 120 {1934) Ogg: Chem. Reviews, 17 423 (1935) Rabinowitch and Wood, Trana. Far. Soc. 32, 555 (1936) - 26 Dickinson and Carrico3 have given an exhaustive analysis of the two mechanisms in terms of collision rates and collision efficiency. They were unable to make a choice between the two mechanisms on the basis of the data at hand. The partial lmowledge gained in this experiment concerning the quantities of interest does not yet permit a conclusive statement to be made concerning which mechanism is the more likely. Further work on the problem will perhaps include a determination of the temperature coefficient of k 4 ; a more precise determination of the thermal reaction; use of sectors of different ratios of light and dark periods; actual determination of the absolute amount of radiation entering the reaction-cell, and an investigation of the cause of the observed falling-off of the rate constant. It is not unlikely that the question of primary recombination of halogens will be settled before long, in which case the actual k 4 determined in this experiment will be more precisely known. SOURCES OF ERROR The principal uncertainty in the measurements arose in the chlorine concentration measurements. Gal- vanometer drift, unknown amounts of convection within the solution, and slight fluctuations of lamp intensity during the measurement combined to make the accurate - 27 - determination of a second-order effect such as t he one under investigation difficult. This was particularly so under conditions of low chlorine concentration, of course. The fluctuation of the lamp during a run likewise led to uncertainty in the results, although this did n ot appear to have been a difficulty of first magnitude. The method of treating the data by taking an average of radiation intensity and chlorine concentration in order to calculate (Iabs) probably did not introduce serious error, nor did the assumption of uniform intensity thruout the solution. Dark reaction has been mentioned as considered negligible. Convection and mixing of the solution was an unknown factor; but calculation of reaction rate constantBunder the assumption of complete mixing and under the assumption of no mixing did not lead to widely different results. No great lack of uniformity of intensity over the cross section of the beam was apparent; however, this will be investigated later. ACKNOWLEDGEMENT The writer wishes to express his profound indebtedness to Professor Roscoe G. Dickinson for the suggestion of the problem, and for his great helpfulness in carrying out its investigation. - 28 - SUMMARY The theory of intermittency in certain types of photochemical reaction has been developed. A rela- tion between reaction rate and certain experimental variables has been explicitjy derived for illumination through a 90 degree sector. The determination of the rate constant of the chain breaking step in the photochlorination of tetrachlorolthylene in carbontetrachloride solution has been made; the most probable value is k 4 r 1 = 3.7 x 1011 c cm/mol second, where k4 is the rate constant and r1 is the quantum yield of the primary photochemical process. r 1 has been shown to lie probably between 1 and 1/260 or 1/440, depending on which of two mechanisms proposed is chosen. Sources of error are discussed. Figure 1 F'igure 2 30 0 I T 1-'---t----i---t--+------lr---t------+---i---+-------1--r- - I 0 .'.L l,. :, ...... ""' 11 --+ {I ;; .!J I{; 0 + <1 I Figure 3