J . Chem. SOC., Faraday Trans. 1, 1981, 77, 1755-1762Pyrolysis of EthylpyridinesRelative Stability of Isomeric Picolinyl RadicalsBY BARRIE D. BARTON AND STEPHEN E. STEIN*Department of Chemistry, West Virginia University, Morgantown, West Virginia26506, U.S.A.Received 19th December, 1979Decomposition rates for C-C bond homolysis in 2-, 3-, and 4-ethylpyridine were measured in a verylow pressure pyrolysis (VLPP) system from 1099 to 1227 K, and relative stabilities of the three isomericpicolinyl radicals (C,H,NCH, * ) were derived. Bond dissociation enthalpies at 298 K, D(C,H,NCH,-CH,),were found to be 316, 309 and 312 kJ mo1-I for 2-, 3-, and 4-ethylpyridine, relative to 310.2 kJ mol-l forethylbenzene, assuming that the A factor for all of these reactions is 1015-0 s-l.Differences in activationenergies of the three ethylpyridine decompositions were the same as the differences in A% (298 K) of thecorresponding methylpyridines ; hence, the A% (298 K) values of all three picolinyl radicals are computedto be the same within ca. 2 kJ mol-’. These results were related to H-atom abstraction studies in solution.Very low pressure pyrolysis (VLPP) is a technique well suited for determining thethermodynamic stability of free radica1s.l The technique is presently being employedin our laboratory to determine structure-stability relationships for a range of benzylicradicals. The present report presents results for the three isomeric picolinyl radicals.These results were derived from measured rates of C-C bond homolysis in thepyrolysis of 2-, 3-, and 4-ethylpyridine.A kinetic study of the unimolecular decomposition of ethylpyridines has not beenpreviously reported in the literature.However, Roberts and Szwarc determined therate of formation of the isomeric picolinyl radicals by C-H scission in thedecomposition of the three picolines (methylpyridines) by the toluene-carrier method.2They reported activation energies of 3 15.9, 320.1, and 324.3 kJ mol-l for 2-, 3-, and$-methylpyridine, respectively, based on the assumption that log,,(A/s-l) was 13.3.These values have since been revised by Benson and O’Nea13 to values of 360.7,364.8,and 369.0 kJ mol-l, by assuming a more reasonable log,,(A/s-l) of 15.5. These valuesare lower than their preferred value of 369.4 kJ mol-1 for the activation energy ofC-H bond scission in toluene, with log,,(A/s-l) = 15.5 assumed.Whereas the Afactors and activation energies found by Roberts and Szwarc2 may be incorrect,Benson and O’Nea13 feel that the rate constants are probably reliable. However, therate differences are small and may not be significant in view of the likely uncertaintiesassociated with this type of determination. Reactivity studies, on the other hand,indicate that the p C-H bond is weaker in toluene. Johnston and Williams4 havestudied the relative reactivities of 3- and 4-methylpyridine toward t-butoxy radicalscompared with toluene in solution. They reported values of 0.44 and 0.33 for 3- and4-methylpyridine relative to 1 .OO for toluene. Gritter and Chriss5 also found tolueneto be more reactive than 4-methylpyridine toward free radical attack.Bridger andRussell6 studied the attack of phenyl radicals on various hydrocarbons and reportedvalues of 4.0 for 3- and 4-methylpyridine compared with 9.0 for toluene as relative1751756 PYROLYSIS OF ETHYLPYRIDINESabstraction rates. The present studies resolve the discrepancy by providing accuratevalues for relative radical stabilities.EXPERIMENTALAPPARATUSThe VLPP apparatus is described elsewhere.’ Operating principles and practices of the VLPPtechnique have been reviewed elsewhere.’ The fused silica reactor used had a collision numberof 1160 and an escape aperture of area 9.08 mm2. Past experience with the technique has shownthis reactor geometry to be optimal for VLPP experiments since (1) too small a collision numbermay lead to transient effectss and the necessity of calculating accurate reactor Clausing factorsand (2) too large a collision number makes recombination and secondary bimolecular reactionsmore likely and accurate measurement of the escape aperture area more difficult.A singleaperture reactor eliminates errors in collision numbers for small apertures that may occur inmultiple aperture reactors due to improper seating of the aperture plate.REAGENTS2-, 3-, and 4-ethylpyridine were obtained from Aldrich and used without further purification.The purities were given as 98% for 2- and 4-ethylpyridine. Gas chromatographic analysis ofthe three reagents in our laboratory indicated 99.3”/,, 98.2% and 99.9% for 2-, 3- and4-ethylpyridine, respectively.RESULTS AND DISCUSSIONExamination of the mass spectra of reactant and products indicated the expected/3’-bond scission, as in eqn (1).mle 107 92 1512At 50% decomposition, the decrease in m/e = 107 corresponded well with the increasein m/e = 15 and with the cracking pattern around m/e = 92.A simple productspectrum was obtained for 3-ethylpyridine at 1 173 K, with a prominent m/e = 92 peakand minor m/e = 91 and 93 peaks. 2- and 4-ethylpyridine exhibited more complexproduct spectra with peaks at 91-94 in similar amounts. These peaks are believed tobe caused by secondary reaction of the picolinyl radica1s.l~ Product peaks at m/e = 16and 78, corresponding to CH, and C,H,N, respectively, were absent in all cases.Rate constants were measured by observing the relative intensities of the parent peakat m/e = 107 under conditions in which (1) the reactant was flowing through the hotreactor and (2) the reactant was flowing directly into the mass spectrometer withoutpassing through the reactor.Slight differences in parent peak intensity in the twB. D. BARTON AND S. E. STEIN 1757modes, under conditions of no reaction, were taken into account. These intensitydifferences arise from small differences in the alignment of the ionization region withthe inlet ports associated with the two measuring modes.Unimolecular rate constants were found to be independent of flow rate betweenl 013 and l 015 molecules s-l for each ethylpyridine pyrolysis, indicating the absenceof significant heterogeneous and bimolecular processes.Some illustrative results for2-ethylpyridine at I148 K are shown in table 1. No systematic trends were found forTABLE 1 .-RESULTS OF FLOW STUDY AT 1 148 Kflow rate % decomposition/molecule s-l (decrease in parent peak)2.5 x 10132.0 x 10141.3 x 101518.0 f 2.017.6 & 0.916.7 & 0.9any of the compounds over the range of flow rates studied (a factor of 50-100). Flowrates chosen for rate studies were in keeping with our own and previous experimenters'experience with VLPP techniques. A flow rate of 2 x IOl4 molecule s-l is near optimalsince (1) at lower flow rates, signal to background ratios become increasingly smallerand (2) at significantly greater flow rates, mass spectrometric intensity becomesnon-linear with concentration.Heterogeneous reactions of a number of polar moleculesin VLPP have been found to follow half-order kinetic^.^The conversion of mass spectral intensity to kuni has been described previous1y.lThe reactor escape rate k, = 4.0 (T/M)i includes a correction for a Clausing factorof 0.91 for the exit nozzle. This Clausing factor was computed by means of a MonteCarlo technique.1°The present work was carried out in the temperature range 1099-1227 K and a flowrate of 2.0 x 1014 molecule s-l. Experimental studies of ethylbenzene decompositionin the same reactor, carried out by Robaugh and Stein' and also by the presentauthors, are well represented by log,,(k/s-l) = 15.0-300.8/8 (8 = 2.303 RT) in thisnarrow temperature range.The experimental data are presented in fig.1, along with RRKM fits to the data,shown as solid lines. The model used in the RRKM calculations is given in theAppendix. The degree of falloff depends primarily on the values of the Arrheniusparameters and is rather insensitive to exact details of the transition-state model. l1Representative data for ethylbenzene obtained by Robaugh and Stein7 are includedfor comparison. Relative rates under identical experimental conditions indicate thatrates of decomposition are in the following order: ethylbenzene = 3-ethylpyridine > 4-ethylpyridine > 2-ethylpyridine. The dissociation rates of the ethylpyridines increaseas the enthalpies of formation of the corresponding methylpyridines increase (table3). We will now show that, if the above rate differences are ascribed primarily todifferences in bond strengths, the enthalpies of formation of the three methylpyridylradicals are virtually identical.An A factor of 1015.0 s-l was chosen for each of the ethylpyridine decompositionsbased on thermochemical kinetic estimates3 and combined toluene carrier and VLPPresults7 for the decomposition of ethylbenzene, shown in eqn (2).- a- + CH31758 PYROLYSIS OF ETHYLPYRIDINES1100 1150 1 2007’1 KFIG. 1 .-Rate constants measured for the decomposition of ethylpyridines. Representative points forethylbenzene are shown for comparison; e, 2-ethylpyridine; ., 3-ethylpyridine; A, 4-ethylpyridine; 0,ethylbenzene.Solid curves shown are RRKM fits to the data.Computed values of the activation energies and differences relative to ethylbenzeneare given in table 2, where A& is the activation energy of the ethylpyridine minusthat of ethylbenzene. A different choice of A factor leads to negligible changes inactivation energy differences calculated for the three ethylpyridines. The assumptionof identical A factors for ethylbenzene decomposition and for each of the threeethylpyridine decompositions is supported by the use of the best methods currentlyavailable.12 These methods yield identical A factors for the four decompositions.Typically, one may make absolute A factor estimates to within k0.3 unit in log,,A,as was done recently for the p-bond scissions in N-methylaniline and N,N-dimeth~1aniline.l~ One may expect relative A factor estimates to be considerably moreaccurate for cases where the reaction of interest is similar to the reference reaction.Such is the case for the decompositions of 3- and 4-ethylpyridine, with ethylbenzenedecomposition as the reference reaction, since the only differences between the threereactions involve small differences in ring frequencies far removed from the reactioncoordinate.(Alternatively, Benson’s AQ parameter is identical for each of these threedecompositions.12) We believe a fair uncertainty estimate is f 0.04 in log,, A ,amounting to 1.0 kJ mol-1 in AE,, for these very similar reactions. The A factorfor decomposition of 2-ethylpyridine is less certain, due to the proximity of the N atoB.D. BARTON AND S. E. STEIN 1759to the breaking bond. Steric interactions may be slightly different, since an H atomortho to the ethyl group in ethylbenzene is absent in 2-ethylpyridine (Le. Benson's ASZparameter may be slightly larger for 2-ethylpyridine dissociation than for the otherethylpyridine dissociations). Allowing for an uncertainty of f 0.1 unit in the relativevalue of log,, A results in an uncertainty of 2.2 kJ mol-l in AE, for the 2-ethylpyridinedecomposition. Some recent calculations concerning A factors in the @-bond scissionsof some methyl-substituted ethylbenzenes provide some insight into possible uncer-tainty in A factors attributable to differences in steric interactions.l* For a model inwhich there was complete relaxation of steric interaction in the transition state for2-methylethylbenzene, a value of log,,A of 15.04 was obtained, compared with a valueof 15.00 for a model in which there was no such relaxation.In this case, for a significantchange in the transition state model, the change in the A factor was modest. Finally,the data may be easily fitted to within kO.4 kJ mol-l for the RRKM models used.Hence, the total uncertainty in AEa resulting from (1) relative differences in A factorsand (2) experimental error and fitting are shown in table 2 for the three ethylpyridinedecompositions.TABLE 2.-ACTIVATION ENERGIES; lOg,,(A/S-') = 15.02-ethylpyridine 307.0 & 12 5.9 3.04-ethy lpyridine 303.0 & 1 1 2.1 + 1.83-et hylpyridine 300.0 f 1 1 - 1.2+ 1.8et hylbenzene 300.8 & 9.2a 0a Ref.(7).The AE, values were used to derive values of Afl(298 K) for the picolinylrelative to a value of 196.6 kJ mol-l derived for the benzyl radi~a1.l~ We make theassumption that the difference in heat capacities between the benzyl radical and thepicolinyl radicals is the same as the differences in heat capacities between ethylbenzeneand the ethylpyridine, i.e. AC;, [reaction (2)]. Since thevibrational frequencies associated with the N-atom in pyridine rings are not expectedto dramatically differ in the molecule and radical, this assumption is expected to bean excellent one. Then, the difference in heats of formation at 298 K between picolinylradicals and benzyl radical is given by eqn (3)[reaction (l)] = AC;,A(AH,O) (R.) = AH: (picolinyl) - AH: (benzyl)= AE, + A% (ethylpyridine) - AH: (ethylbenzene)= AEa + A(AH,O) (R).(3)Since A% values for the ethylpyridines have not been reported in the literature, A(AH;)(R) was taken to be the same as the difference between AH: (methylpyridine)and AH: (toluene). Values of the heats of formation are presented in table 3. The errorestimates for AH: of the benzyl radical were those of Rossi and G01den.l~ The resultsshow that, on the basis of the above assumptions, the heats of formation of the threeisomeric picolinyl radicals are virtually identical. Stated differently, the computeddifference in activation energies for p C-C bond scission in isomeric ethylpyridinesis simply equal to their estimated differences in AH:.This interesting result implie1760 PYROLYSIS OF ETHYLPYRIDINESTABLE 3.-HEATS OF FORMATION AT 298 KAH: A(AEIp)/kJ mol-l /kJ mol-l2-meth ylp yridine3 -me th ylp yridine4-meth ylp yridinetoluene2-picolinyl(l)3-picolinyl(2)4-picolinyl(3)benzyl99.16 & 0.75" 49.0 f 1.2106.36 +0.58" 56.2 & 1 .O102.10f 1.40" 51.9f 1.825 1.50 8.4 54.9 f 4.225 1.60 f 7.0 55.0 f 2.8250.60 & 7.8 54.0 & 3.650.17 k0.42" -196.60 f 4.2b -" J. D. Cox and G. Pilcher, Thermochemistry of Organic and Organometallic Compounds(Academic Press, London and New York, 1970). Ref. (15).that the factor(s) responsible for the differences in stability of isomeric methylpyridinemolecules is somehow negated in their radicals.While the error limits given in tables2 and 3 cast some doubt on the certainty of this conclusion, these error limits are,we feel, conservative, especially in the present context where relative, not absolute,errors are of primary significance. The key fact that we feel is of importance is thatdifferences in enthalpies of formation almost exactly match differences in activationenergies computed on the basis of the very reasonable assumption that all A factorsfor dissociation of these species are the same.One may define the bond dissociation enthalpy in a simple bond fission reaction(4)The difference in bond dissociation enthalpies between reaction (1) and (2), AD, isthen given by eqn ( 5 )AD =A% (picolinyl) - A S (benzyl) + AG (ethylpyridine) - A S (ethylbenzene)( 5 )The derived bond dissociation enthalpies relative to a value of 310.2 kJ mol-l forethylbenzene' are presented in table 4.by eq* (4)D(R-R') = AH: = ZAHfo-CAHfo.products reactants= A(AK) (Re) - A(AK) (R) = AE,.TABLE 4.-BOND DISSOCIATION ENTHALPY AT 298 KD(C,H,NCH,-CH,)/kJ mol-l2-ethy lp yridine 316f 123-ethylpyridine 309f 114-ethylp yridine 312& 11ethylbenzene 310.2k9.2"a Ref.(7)B. D. BARTON A N D S. E. STEIN 1761One may use the results of the present study to estimate relative H-atom abstractionrates, compared with toluene, as in eqn (6)The ordering of estimated abstraction rates at 298 K would be: toluene z 3-methylpyridine > 4-methylpyridine > 2-methylpyridine.Our results are in accordwith the finding that H-atom abstraction rates from methylpyridines are slower thanthat of toluene. However, measured relative abstraction rates from isomericmethylpyridines are not in the same order as their dissociation rates.The relative stabilities of the isomeric picolinyl radicals are not in accord with theolder work of Roberts and Szwarc,2 with regard to the order of relative rates of thethree methylpyridines. The present value of 251 kJ mot1 for the AH: (298 K) ofpicolinyl radicals is 14-23 kJ mol-1 higher than that derived by Benson and O’Nea13from the data of Roberts and Szwarc, although part of the difference is attributableto the choice of A factor, We believe that the older toluene-carrier method used byRoberts and Szwarc is simply not accurate enough to obtain the small rate differencesinvolved.Fig. 1 clearly shows the high level of precision available in this type ofcomparative kinetics experiment using VLPP.CONCLUSIONSThe /?-bond strengths in ethylpyridines relative to that in ethylbenzene were foundto be in the following order: ethylbenzene = 3-ethylpyridine < 4-ethylpyridine < 2-ethylpyridine, with bond dissociation enthalpies at 298 K of 310.2, 309, 312 and316 kJ mol-l, respectively. An interesting conclusion drawn from the present studyis that the relative activation energies of the three decompositions are the same as therelative AH: (298 K) values of the parent compounds; hence, within experimentalerror, the three isomeric picolinyl radicals have identical AH: values.Further studiesare underway to discover the extent of validity of this finding for other relateddecompositions.We thank the donors of the Petroleum Research Fund, administered by theAmerican Chemical Society, for partial support of this research. We are also gratefulto the Research Corporation for partial support, and to the West Virginia EnergyResearch Center for a postdoctoral fellowship for B.D.B.APPENDIXThe following model was used for the RRKM calculations for the three ethylpyridines.Complex: P = 1.22 x 1 0 - 1 ~ ~ (g cm2)3, a: = 3.Vibrational frequencies: 3000(9), 1620(3), 1380(9), 1 180(4), 1040(4), 960(2), 870(2), 780( l),660(2), 450(3), 410(1), 350(2), 250(1), 70(2), lO(1).Molecule: I = 7.29 x (g ~ r n ~ ) ~ , CT = 3.Vibrational frequencies: 3000(9), 1620(3), 1380(9), 1180(2), 1040(4), 960(3), 870(2), 760(3),660(2), 450(3), 350(2), 180( l), 50( I), 30( 1).58 F A R 1762 PYROLYSIS OF ETHYLPYRIDINESD. M. Golden, G. N. Spokes and S. W. Benson, Angew. Chem., Int. Ed. Engl., 1973, 12, 534.J. S. Roberts and M. Szwarc, J. Chem. Phys., 1948, 16, 981.S. W. Benson and H. E. O’Neal, Kinetic Data on Gas Phase Unimolecular Reactions; NSRDS-NBS,21, 1970.K. M. Johnston and G. H. Williams, J. Chem. SOC., 1960, 1446.R. J. Gritter and R. J. Chriss, J. Org. Chem., 1964, 29, 1163.R. F. Bridger and G. A. Russell, J. Am. Chem. SOC., 1963, 85, 3754.D. A. Robaugh and S. E. Stein, Int. J. Chem. Kinet., in press.* D. F. Kelley, B. D. Barton, L. Zalotai and B. S. Rabinovitch, J. Chem. Phys., 1979, 71, 538.* S. E. Stein, S. W. Benson and D. M. Golden, J. Catal., 1976, 44, 429.lo D. H. Davis, J. Appl. Phys., 1960, 31, 1169; B. D. Barton, Ph.D. Dissertation (University ofl1 P. J. Robinson and K. A. Holbrook, Unimolecular Reactions (Wiley-Interscience; London, Newl2 S. W. Benson, Thermochemical Kinetics (Wiley-Interscience, New York, London, Sydney, Toronto,l3 A. J. Colussi and S. W. Benson, Int. J. Chern. Kinet., 1978, 10, 11 39.l4 B. D. Barton and S. E. Stein, J. Phys. Chem., 1980, 84, 2141.l5 M. Rossi and D. M. Golden, J. Am. Chem. SOC., 1979, 101, 1230.Washington, Seattle, WA, 1979).York, Sydney, Toronto, 1972).2nd edn, 1976), pp. 90-100.(PAPER 9/2011