Kinetics of Decomposition of Chemically ActivatedAlkyl FluoridesBY J. A. KERR, A. W. KIRK, €3. V. O'GRADY, D. C. PHILLIPS ANDA. F. TROTMAN-DICKENSONChemistry Dept., University of Wales, AberystwythReceived 5th June, 1967Activated alkyl fluorides have been made (a) by the reactions of suitable radicals with fluorine,(b) by the combination of radicals, (c) by the insertion of methylene into fluorinated alkanes. Therates of elimination of hydrogen halide have been compared with the rates of stabilization by collisionfor C2H5F, n-C3H7F, iso-C3H7F, s-C4H9F, t-C4H9F, CH2FCH2F, CH3CHFz, CH3CF2CH3,CH3CH2CHF2, CHF2CHFz, CH2C1CH2F, CH3CHClF and CH3CClF2. The rates have beenstudied over a range of temperature for six compounds. The results can be interpreted by the RRKtreatment but a RRKM treatment is desirable.The stepwise deactivation of n-alkyl fluorides isconsidered.Fluorinated compounds offer many advantages as components of systems for thestudy of chemical activation. The advantages derive principally from the strengthof the carbon-fluorine bond which is not attacked by either mono or biradicals andwhich is not involved in isomerizations. The first series of reactions in which thesepropertiss were exploited was that of the fluorinated cyclopropanes. Activatedcyclopropanes were made by the addition of methylene from the photolysis of ketento the appropriate fluorinated ethylene.' Studies were made of the isomerizationof the chemically activated molecules to fluorinated propenes. Complementarystudies of the thermal reactions were also made.2 The results are summarized inthe table 1.At first sight, these are a consistent set of data in which the relative ratesTABLE 1 .-THERMAL ISOMERIZATION OF FLUORINATED CYCLOPROPANESA n E4 Emolecule (SeC-1) (kcal moIe-1)cyclopr opane 15.2 13 65.0 104monofl uorocyclopropane 14-6 15 61-0 1121,l-difluorocyclopropane 14.1 17 56.4 1151,1,2-trifluorocyclopropane 14.4 18 50.5 1071,1,2,2-tetrafluorocyclopropane 15.3 21 48.5 116R and E are the number of effective oscillators and energy content derived from RRK theory.of isomerization of the chemically activated species can be understood in terms of theA factors, activation energies and number of effective oscillators in the molecules.It is, however, possible that the thermal reactions were not the same as those inducedby chemical activation. The isomerization products of the chemically activatedcyclopropanes were shown to be the expected fluorinated propylenes, but thesecompounds were not produced in stoichiometric quantities in the thermal reactions.At the time, their absence was assigned to thermal polymerization but it now seemslikely that isomerization was accompanied by decomposition of the cyclopropanesto yield difl~oromethylene.~ Although the nature of the reaction is uncertain themeasurements of the thermal unimolecular reactions at low pressures served for thedetermination of the number of effective Kassel oscillators as well as the Arrheniusparameters, as given in table 1.26264 DECOMPOSITION OF ALKYL FLUORIDESActivated alkyl fluorides have been produced in three ways which yield moleculeswith widely different energy contents.Molecules with an energy content around115 kcal mole-l have been formed by the addition of singlet methylene usuallyproduced by the photolysis of keten with light of 3130A to suitable alkyl fluorides.These experiments are normally carried out in the presence of considerable concentra-tions of oxygen to eliminate reactions of monoradicals and of triplet methylene.In the absence of good thermochemical data it is impossible to estimate accuratelythe energy contents of the various molecules. The value of 115 kcal mole-' wasderived for the conversion of methyl fluoride to ethyl fluoride for which the thermo-chemistry is satisfactory. It is unlikely that the heats of the other reactions differby more than 1-2 kcal mole-'.Molecules with an energy conent between 85 and 95 kcal mole-' can be made bythe combination of alkyl radicals, one at least of which contains fluorine.Thefluorinated radicals can be made either by the abstraction of a hydrogen atom froman alkyl fluoride by methylene (in some systems this abstraction predominates overthe insertion reactions), or by the photolysis of a suitably fluorinated ketone. It isone of the convenient properties of fluorine chemistry that fluorinated acetone photo-lyzes by rupture into alkyl radicals like the parent compound, but chlorinated ketoneslose chlorine atoms. The energy content of the products is uncertain because evenwhen the heats of formation of the smaller radicals are known those of their com-bination products are not.Alkyl fluorides of lower energy content can be made by reactions of which thefirst member of the series isThis is the reaction by which ethyl fluoride is formed when ethane is fluorinated bythe element. The exothermicity of the reaction is 69 kcal mole-l but it is not knownhow the energy is distributed between the products.Since the C-F bond is strongand its formation is responsible for the exothermicity of the reaction, which probablyhas no activation energy, most of the energy must become vibrational energy of ethylfluoride. Reactions of this type in which a bimolecular reaction yields a chemicallyactivated complex molecule should, in the future, provide considerable insight intothe energy distribution in chemical reactions.C2H5fF2 = C,H,F+F.ETHYL FLUORIDEThe first reported observation of the decomposition of an activated alkyl fluoridewas that of ethyl fluoride formed in the fluorination of ethane.4 This report wasmisleading in that decomposition was caused by self-heating of the reaction mixture.At much lower pressures than were first used, ethyl fluoride formed by reaction ( 1 ) as aC2H5 + Fz = C2H5F* + Fstep in the reaction of ethane with fluorine does decompose by loss of hydrogenfluoride.The rate of reaction was found by comparison with the rate of stabilizationof the fluoride k,,(1)CZHSF* = CzH,+HF, keC2HsF*+M = CZHSF+M, ksThe experiments were carried out in Pyrex vessels at room temperature as previ~usly.~Reaction occurred rapidly on mixing.The products were analyzed for ethylene andethyl fluoride and an excellent plot of the functionwas obtained between 0-08 and 2.0 torr. k, can be calculated from ke/k, on theassumption that log k, (torr-l sec-') is equal to the collision frequency 7.0.[C,H4I/[C,H5FI = (k?/kS)(l /[MIKERR, KIRK, O'GRADY, PHILLIPS AND TROTMAN-DICKENSON 265Activated ethyl fluoride was also formed by the combination of methyl andfluoromethyl radicals from the photolysis of 1 : 1 mixtures of acetone and 1,3-difluoroacetone between 10 and 120 torr at 40°C. Over this pressure range thestabilization could not be described in terms of a single collision.A step-laddermodel was adopted as described below. From the limiting effect of higher pressuresa value of kJk, was deduced. The energy content of the activated molecules wastaken as 85-4 kcal mole-' following Benson and Haugen's recommendation fora similar reaction.Ethyl fluoride containing 115 kcal mole-' was made by the photolysis of ketenin mixtures of methyl fluoride and neopentane (14-7 : 1) in the presence of about 10oxygen. The neopentane was added because large amounts of ethylene were formedin the photolysis of keten alone. The decomposition of the ethyl fluoride couldonly be estimated by comparison with the stable 2,2-dimethylbutane that was alsoformed by insertion. Again there was evidence of step-wise deactivation which hadto be allowed for when estimating k, for molecules with the full amount of energy.The values of k, found in each of the three systems can be correlated as shown belowusing classical RRK treatment with YE = 12.5 (by analogy with the cyclopropanes,A = 13.5 sec-' (as for other elimination reations of alkyl halides 6, if the activationenergy is 51 kcal mole-'.energy contentkcal mole-1 log k (obs.) log k (calc.)abstraction 69 6.78 6-79combination 85 9.0 8.96insertion 115 10.46 10.57The RRK treatment underestimates the rate of reaction of molecules with low energycontent so that the true activation energy must be greater than 51 kcal mole-'.PROPYL FLUORIDESThe photolysis of a 1 : 4 mixture of diethyl ketone and 1,3-difluoroacetone wasinvestigated between 0.1 and 20 torr and 30 and 350°C.The yields of propyleneand n-propyl fluoride were determined. Good straight line relations of Cpropylene]/[n-propyl fluoride] against 1 /[MI were obtained at each temperature.The relative yields of n-propyl fluoride, isopropyl fluoride and propylene weremeasured at room temperature when keten was photolyzed in the presence of ethylfluoride between 6 and 350 torr with oxygen. In some runs neopentane was addedin amount equal to one seventh of the ethyl fluoride. Insertion by singlet methylenein the 2-position occurs 1.11 times as rapidly as attack on the I-position. The re-activity of the C-H bond in the 2-position is 3.5 times that in neopentane. Thesteady-state equations describing the system were readily solved.The deactivationplot for n-propyl fluoride was a straight line, but that for isopropyl fluoride was not,which indicated stepwise deactivation. The interpretation of the isopropyl fluorideresults was therefore slightly ambiguous and a large uncertainty must be placed onthe result.BUTYL FLUORIDESTwo series of experiments were carried out on the photolysis of keten in thepresence of 1 : 10 neopentane+isopropyl fluoride mixtures, over a pressure range of0.4-25 torr. In the second series about 10 % of oxygen was present in the mixtures.Both s-butyl and t-butyl fluoride were formed in the presence of oxygen. The relativ266 DECOMPOSITION OF ALKYL FLUORIDESrates of insertion of singlet methylene into the C-H bonds are given in table 2.The analysis of the results obtained with oxygen was comparatively simple ; the ratesof decomposition of the activated molecules are given in table 3.In the absence ofTABLE 2.-RELATIVE RATES OF INSERTION OF METHYLENE INTO C-H BONDSbond relative rateHXH2C(CH3)3 1H-CH2F 0.21H-CH2CH2F 0-29H--(CHs)CHF 0-26H-(CH3)2CF 0.72H-CH2(CHs)CHF 0.52TABLE 3 .-ELIMINATION OF HYDROGEN HALIDES FROM CHEMICALLY ACTIVATEDALKYL HALIDES AT ROOiM TEMPERATUREkelkscm alkyl halide means of formationn-C3H7Fi-C3H7FS - C ~ H ~ Ft-C4H,FFCHpCHiFCH3CHF2CHj CF2CH3 aCHjCF3LOSS OF HYDROGEN FLUORIDEabstraction 0.07combination 14insertion 351combination 0.1insertion 2.0insertion 20 rrt5insertion 0-065Combination 1.7insertion 16.0abstraction 0*00055combination 4.8abstraction 0.135combinat ion 20.4combination 0-43insertion 1 *24combination 4.1insertion 19.0combination 1.4CHFZCHF;? combination 0-008 dCH2FCH2Cl abstraction 0.00096CH3CHFC1 abstraction 0.009CH3CF2C1 combination 0.56LOSS OF HYDROGEN CHLORIDECH2FCH2CI abstraction 0.000026CH3CHFC1 abstraction 0-08CH3CF2C1 combination 0.16n is number of effective classical RRK oscillators.I t E*12.5 512121- -14 52455443Ea is calculated activation energy on the assumption that log A = 13.5 sec-'.a ref.(8) ; b ref. (1 1) ; C at 100°C ; d at 200°C.oxygen, butyl fluorides were also formed by abstraction of hydrogen atoms by methyl-ene followed by combination. The quantitative interpretation of the results isdoubtful but an approximate rate of decomposition of t-butyl fluoride formed bycombination was foundKERR, KIRK, O’GRADY, PHILLIPS AND TROTMAN-DICKENSON 2671,2-D I F L U 0 R 0 E T H A N EThe attack of fluorine atoms on ethyl fluoride occurs preferentially at the 1-position.Hydrogen atoms in the 2-position are some 0.5 times as reactive per Hatom. Activated 1 ,Zdifluoroethane is, however, formed in sufficient quantities forstudy between 0.01 and 1 torr at room temperature. The decomposition of thisactivated ethane was the first such decomposition to be reported for a molecule formedby the combination of two radical^.^ We have considerably extended the work inthe range from 7 to 70 torr and -33 to 402°C. The values of kJkd are plotted infig.1 from which it can be seen that the simple treatment well describes the results.-2-8 I 1 I300 400 5 0 0 6 0 0 700T “KFIG. 1.-Relations between temperature and the function equal to Z / A . (E/E--Ea)”-’. Solid linesare calculated values for activation energies between 49 and 54 kcal mole-’ for the elimination ofhydrogen fluoride from 1,Zdifluoroethane. Open circles are this work ; filled circles are results ofPritchard, Venugoplan and Graham.’ Calculated lines on basis of A = sec-’, number ofeffective oscillators = 14 ; EZg8 = 85.4 kcal mole-’.1, l - ~ IF LU o ROETH A NESince this ethane is preferentially formed in monofluorination by atomic fluorineand is comparatively reactive, the elimination reaction was studied at room temperaturebetween 0.1 and 1 torr.Two methods were used to make this ethane by combination.The first was the photolysis of mixtures of acetone and 1,1,3,3-tetrafluoroacetonebetween 10 and 50 torr and 112 and 347°C. The second was by the photolysis ofketen in difluoromethaae between 40 and 500torr and 17 and 252”. No ethaneswere found in the presence of oxygen so that insertion is ruled out. The resultsobtained in the two systems were similar but not identical. Different values of k,/kswere found because the deactivating efficiencies of the bath gases varied. This wasconfirmed by using C2F6 as the bath gas. Results obtained at 500°K were as follows :ketone system + C2F6 44.0 cm ;keten system + C2F6 44.8 cm ;ketone alone 38-8 cm;keten alone 45-9 cm268 DECOMPOSITION OF ALKY L FLUORIDESI \ I I I3 0 0 4 0 0 5 0 0 6 0 0 7'T "KI I I I0FIG.2.-Relation between temperature and the function equal to Z/A . (E/E-E&'. Solid linesare calculated for activation energies for elimination of hydrogen fluoride from 1,l -difluoroethaneproduced by radical combination from photolysis of ketones ; A = 1013.5 sec-I ; number of effectiveoscillators = 14 ; EZgs = 85.4 kcal mole-'.I I I\300 400 500 6 0 0 7(I I IT"K3RG. 3.4imilar to fig. 2 but with 1,l-difluoroethane produced by combination of radicals from thehydrogen abstraction by methylene from difluoromethaneKERR, KIRK, O'GRADY, PHILLIPS AND TROTMAN-DICKENSON 269Hence we have deduced that the relative collision efficiencies are ketone 1.0; C2Fs0.80, CH2F2 0.38.The overall results are shown in fig. 2 and 3. A lower approxi-mate value of kJk, has been calculated from earlier re~ults.~DIFLUOROPROPANESStudies of the photolysis of keten in the presence of 1,l-difluoroethane at roomtemperature have been described.8 The propanes are formed both by combinationof radicals and by insertion and it is difficult to derive accurate values of kJk,.Further experiments have been done between 40 and 500 torr and -35" and 220°Cin the presence of oxygen when only insertion is important. The results which areplotted in fig. 4 differ slightly from those of the earlier work.I I I300 4 00 5 0 0T "KFIG. 4.-Relation between temperature and the function equal to Z/A . (E/E-E&' for eliminationof hydrogen fluoride from difluoropropanes formed by the insertion of methylene into 1,l-difluoro-ethane.Calculated lines are based on, for both reactions, A = sec-' ; Ezy8 = 115 kcalmole-' and number of effective oscillators = 21.1,1,2,2-T E T R AF LU 0 R 0 E T HA N EThe photolysis of 1,1,3,3-tetrafluoroacetone has been studied previously lo but noelimination of HF from the ethane was found between 20 and 100 torr. We havestudied the system at 2-18 torr and found elimination as shown in the table of resultsand in fig. 5.CHLOROFLUOROETHANEThe attack of atomic fluorine on ethyl chloride occurs preferentially on the CHzClgroup. The hydrogen atoms in this group are 2.6 times as reactive as those in th270 DECOMPOSITION OF ALKYL FLUORIDEST "KFIG.5.-Relation between temperature and the function equal to Z/A . @/I?- for elimination ofhydrogen fluoride from 1,1,2,2-tetrafluoroethane formed by the combination of radicals. Calculatedlines on the basis of A = = 90 kcal sec-' ; number of effective oscillators = 17 ; andmole- l.methyl group. The system is complex as both the fluorochloroethanes can eliminatein two ways. Information on all four eliminations can be obtained from observationsbetween 0.01 and 1 torr, but the rate constants are not determined with great accuracy.The low rate of elimination of hydrogen chloride from CH2C1CH2F was particularlydifficult to measure and the rate listed in table 3 should be regarded as an upper limit.Nevertheless, the relative sizes of the four elimination rate constants from moleculescontaining equal amounts of energy and with similar vibrational properties arefirmly established.1,l -D I F L u o R o c H L o R o E T H A NEThis ethane was formed by the photolysis of mixtures of acetone and 1,1,3,3-tetrafluorodichloroacetone between 1 and 10 torr and 100 and 230°C.Eliminationof both hydrogen chloride and hydrogen fluoride is observed. The results are shownin fig. 6.EFFECT OF STRUCTURE ON RATEIt is probable that all the reactions in which alkyl fluorides are formed by insertionare of equal exothermicity. The differences in their rates of elimination can thereforebe attributed to the two factors : the activation energies of the elimination and thevibrational properties of the molecules, which incidentally determine the A factors.It is probable, though we have not yet completed the full calculations that are neces-sary to prove the point, that the vibrational properties of n-propyl fluoride and iso-propyl fluoride are extremely similar.It is further known l2 that methyl substitutionat a point far removed from the chlorine atom makes only a small difference to therate of elimination. Accordingly the rate of elimination from ethyl fluoride formedby insertion (288 cm) is greater than that from n-propyl fluoride (2 cm) by a factoKERR, KIRK, O'GRADY, PHILLIPS AND TROTMAN-DICKENSON 271I I =' \ I300 400 So0 600 7 0 0T "KFIG. 6.-Relation between temperature and the function equal to Z/A . (E/E-Ea)"-' for 1,l-difluoro-chloroethane formed by the combination of radicals.Open points refer to the elimination of hydrogenchloride ; filled points to the elimination of hydrogen fluoride. Calculated lines for the differentactivation energies were in both cases based on A = sec-I ; number of effective oscillators =15.5 and EZg8 = 95 kcal mole-1 (based on a consideration of bond strengths in CF3--CF3 andof 144 which can be broken down to a factor of 260 caused by the greater molecularcomplexity slightly offset by a factor of 1-8 for the methyl substitution. The resultfor isopropyl fluoride should therefore be multiplied by 260 for the purposes of isolat-ing the effect on activation energy of changing from a primary to a secondary fluoride.A similar argument can be applied to the butyl fluorides.Hence we find that therelative reactivities for otherwise similar fluorides formed by insertion reactions are :CHj-CFS).primary1secondary tertiary fluoride27 940These can be compared with the relative rates of elimination from chlorides at 360"and bromides at 320°C :primary secondary1 1431 292tertiary26000 chlorides61000 bromidesThe trend in the fluoride series is of the size expected for molecules of much higheraverage energy content. A further parallelism between the bromides and the fluoridescan be seen in the effect of a- and P-halogen substitution. Thus, at 437°C the relativerates of elimination from the bromides are :CHsCBr3 CH3CIIBrz CHzCH2Br32 8 1The activation energies for the fluorides from this work are :CHzBrCH2Br0.5CH3CHF2 CH3CH2F CH2FCH2F45 51 52 kcal mole-272 DECOMPOSITION OF ALKYL FLUORIDESDEACTIVATION BY COLLISIONIn the systems in which the elimination from ethyl fluoride was studied, the varia-tion of the yields of ethylene and ethyl fluoride with pressure could not be fitted toa scheme in which a single collision was completely effective in deactivating themolecule. The simplest model that would explain the results was one in whichstepwise deactivation occurred.The observations on the methylene insertion andradical combination systems were treated in this way using an RRK function torelate the rate constant to the energy content. The inadequacies of this function areleast serious when dealing with molecules containing such high energies. The resultsare summarized below.systemCH2(S)+CH3F (18")CH2F+ CH3 (40")CH2F+ CH3 (30")CH2F+ CH3 (42")CH2F+ CH3 (90")CH,F+ CH3 (1 60")CHZF+ CH3 (235")CH2F+ CH3 (284")deactivator step-sizekcal mole-'The results near room temperature are similar to those obtained by Rabinovitchand his co-workers by more precise methods in other systems.13 No studies overa large range of temperature appear to have been published.It is difficult to acceptthe results obtained here with nitrogen at their face value, but there is no doubt thattemperature has a big effect on the deactivation process. The results have beenexpressed in terms of step-size but it would be possible to describe them in terms ofvariation of the proportion of the collisions that remove energy. The effect ofstepwise deactivation is apparent at high pressures with ethyl fluoride. Probablythe comparatively low A factor for the reaction lends added importance to the energyfactors.'B. A. Grzybowska, J. H. KnoxandA. F. Trotman-Dickenson, J. Chem. SOC., 1963, 746.F. Casas, J. A. Kerr and A. F. Trotman-Dickenson, ibid., 1965,1141.F. Casas, J. A. Kerr and A. F. Trotman-Dickenson, J. Chem. SOC., 1964, 3655.F. P. Herbert, J. A. Kerr and A. F. Trotman-Dickenson, ibid., p. 571.J. M. Birchall, R. N. Haszeldine and D. W. Roberts, Chem. Comm., 1967, 287.G. C. Fettis, J. H. Knox and A. F. Trotman-Dickenson, J. Chem. Soc., 1960, 1064.S. W. Benson and G. R. Haugen, J. Physic. Chem., 1965,69, 3898.A. Maccoll in Studies on Chemical Structure and Reactivity (Methuen, London, 1966). ' G. 0. Pritchard, M. Venugoplan and T. F. Graham, J. Physic. Chem., 1964, 68, 1786.* J. A. Kerr, B. V. O'Grady and A. F. Trotman-Dickenson, J. Chem. SOC. A , 1966, 1621.G. 0. Pritchard, J. T. Bryant and R. L. Thommasson, J. Physic. Chem., 1965, 69,2804.lo G. 0. Pritchard and J. T. Bryant, J. Physic. Chem., 1965, 69, 1085. ' W. G. Alcock and E. Whittle, Trans. Faraday Sac., 1965,61,244. R. D. Gila and E. Whittle.ibid., 1425.l2 A. Maccoll in Studies on Chemical Structure and Reactivity (Methuen, London, 1966), p. 53.l3 G. H. Kohlmeier and B. S. Rabinovitch, J. Chem. Physics, 1963, 38, 1692