Kinetics and Mechanism of Chloroprene Cycloaddition Reactions

Kinetics and Mechanism of Chloroprene Cycloaddition Reactions BY N. C . BILLINGHAM*, J. R. EBDON,?R. S. LEHRLE, J. L. MARKHAM, AND J. C . ROBB Chemistry Department, The University of Birmingham, Birmingham Receiued 10th Jualy, 1968 The thermal dimerization of chloroprene, containing diphenylpicrylhydrazylas polymerization inhibitor, has been studied in the temperature range 2560°C. Cyclobutane- and cyclohexenederivatives are formed in comparable yield ; the products were characterized as 1,2-dichloro-l,2 divinyl-cyclobutane,together with 1-chloro-4 (ct-chloroviny1)-cyclohexene-1and 2-chloro-4(ct-chlorovinyl)-cyclohexene-1. Small quantities of 1,6-dichloro-cyclo-octadiene1 : 5 are formed from part of the cyclobutane-derivative(Cope rearrangement). Dilatometric studies have been used to measure the overall rate and activation energy of the dimerizationprocess, and indicate that the overall reaction is second order. The kinetics of formation of the individual dimers have been investigated by using a pyrolysis-GLC technique to analyze the reaction products ; the energy of activation for formation of the cyclobutanederivative is identical with that for formation of the cyclohexene derivative, and the corresponding entropies of activation are similar. From a consideration of the kinetic parameters, the positions of the chlorine atoms in the products, and the absence of either solvent or photochemical effects, a general mechanism for the dimerization has been proposed. It is postulated that the transition state is formed by one-centre attack, and that the three possible types of attack (1,l; 1,4 ; 4,4) all occur to give intermediates which are probably diradical in character and which possess similar free energy. Published work on other cycloaddition systems provides some support for the mechanism suggested.

In this paper a simple notation is used for diene cycloaddition processes, which is symbolized by the size of the ring formed, Thus processes leading to the formation of the four possible Diels Alder adducts (I-IV) are called 6R.

I

I1

I11

IV

Other dimeric structures, chlorine-substituted divinyl cyclobutane and chlorinesubstituted cyclo-octadiene, are also considered ; the corresponding addition processes are symbolized by 4R and 8R respectively. The first detailed studies of chloroprene dimerization were reported by Cope et aZ.,1*2who used thiodiphenylamine as a polymerization inhibitor for the reaction at 80°C. The results indicated that 6R and 8R processes had occurred ; the products characterized were 111, IV (isolated as its dehydrochlorination product, 1-chloro-4, vinyl-cyclohexadiene 1 : 3) and 1,6-dichloro-cyclooctadiene1 : 5 (V). These results were generally supported by those of Klebanskii and D e n i ~ o v a and , ~ by Nazarov et u Z . , ~ who also believed I to be present in dimers prepared at 20°C and subsequently distilled at 90°C. All these workers prepared or isolated the dimers at high temperatures (up to 9OOC) and commented upon the instability of the products.

* present address, The Chemical Laboratory, The University of Sussex, Brighton. f- present address: Chemistry Dept., University of Lancaster, Lancaster. 470

View Article Online

B I L L I N G H A M , E B D O N , LEHRLE, MARKHAM A N D ROBB

47 I

The objectives of the present work were to study chloroprene dimerization and to separate and characterize the products, at temperatures lower than those previously used. It was also of interest to investigate the kinetics and mechanism of the individual dimerization reactions of chloroprene, especially since mechanistic proposals about the corresponding reactions of butadiene have recently been p ~ b l i s h e d . ~ . ~ Finally, studies of chloroprene dimerization reactions are relevant to investigations of the thermal polymerization of c h l ~ r o p r e n e . ~ Throughout -~ this paper all rates and rate constants are quoted in terms of the rate of disappearance of monomer. E X P E R I M E N T A L A N D RESULTS MATERIALS

CHLOROPRENE was supplied by Distillers (B.P.) Ltd., as a 50 % solution in p-xylene, inhibited with thiodiphenylamine and tert-butyl catechol. The crude material was fractionated at a reflux ratio of 35 : 1, using a 1.5 m column of Fenske helices under reduced nitrogen pressure. The nitrogen was " white spot " grade, and was fed through scrubbers containing chromous chloride solution to remove traces of oxygen, and then dried by passage through concentrated sulphuric acid, silica gel, and molecular sievestype 4A. The receiving vessel was cooled in Drikold +methanol during distillation ; only the middle distillation fraction (b.p. 34~4°Cat 41.7 kN m-2) was collected for use. GLC analysis of this fraction indicated better than 99-9 % purity. Further purification was performed on a high-vacuum line by degassing the monomer and thenallowing it to stand over l,l-diphenyl-2,picryl-hydrazylfor ca. 18 h at room temperature ; the monomerwas then redistilled twice on the vacuum line before being finally distilled into a reaction vessel. At all stages of the purification, storage, and handling of the monomer, precautions were taken to eliminate exposure to air and to maintain the material at as low a temperature as possible in order to retard any peroxidation. 1,~-DIPHENYL-~,PICRYL-HYDRAZYL (DPPH) was supplied by Koch-Light Laboratories Ltd., and was used without further purification other than drying on the high-vacuum line. METHYL ETHYL KETONE (MEK), used as a solvent in kinetic experiments, was B.D.H. " Special for chromatography " grade. No impurities were detected in it by GLC. The material was degassed and distilled on the high-vacuum line before use. CHARACTERIZATION O F LOW-TEMPERATURE DIMERIZATION P R O D U C T S

A preliminary account of our observations of the dimerization of DPPH-inhibited chloroprene at 35°C has been published.1° Detailed evidence was presented for the existence of an unexpected 4R addition process as a principal contributor, i.e., the product 1 ,2-dichloro-1,2divinyl-cyclobutane (VI) was isolated and characterized. The other principal product was a chlorovinyl-chlorocyclohexene,i.e., a 6R process was the other contributor. (a) P R E P A R A T I O N

O F THE D I M E R M I X T U R E

Large-scale preparations were performed in a two-necked flask of ca. 300 ml capacity. One neck of this flask was closed by a break-seal to which was attached a B19 socket. About 3 g of DPPH (polymerization inhibitor) was introduced into the flask via the second neck, which was then attached to the high-vacuum line. About 250 ml of purified chloroprene was distilled into the flask, which was then sealed off. Such flasks were thermostatted at 35°C for periods of up to 3 weeks, until the first indication of a change in the deep violet colour to brown could be detected. (This indicates imminent consumption of the inhibitor.) The flask was then attached to the high-vacuum line by the socket on the second neck, and its contents cooled in an ice+ salt bath. The break-seal was opened, and the monomer removed by distillation, leaving an oily residue in the flask. The flask was then allowed to warm to room temperature, and almost all of the dimeric yield was separated from the inhibitor residue by distilling the former into a liquid-nitrogen trap by prolonged high-vacuum pumping. This dimer mixture was stored under nitrogen at -45°C. The inhibitor residue, together with a little oil, was retained for subsequent analysis of its dimer content.

View Article Online

472

CHLOROPRENE DIMERIZATION

(b) I S O L A T I O N OF P R O D U C T S The mixed dimers have been separated into two components by the thin-layer chromatographic procedure already described,1° and also by a spinning-band distillation technique. l1 The latter technique is efficient and also offers the advantages of small hold-up volume and small pressure differential. In the present work, about 6 ml of the mixed dimers was fractionated at a pressure of 6.7 N m-2 using a " coiled-coil " stainless steel spinning band column at a steady 5 : 1 reflux ratio.12 The temperature profile of the distillation is shown in fig. 1. Two fractions were collected, each accounting for ca. 30 % of the original and corresponding to the plateaux on the curve; a third fraction corresponding to the intermediate region was also retained. During the distillation, the boiler-temperature reached ca. 70°C ; this represents a considerable advantage over conventional fractional distillation where temperatures exceeding 120°C were recorded. Nevertheless, the presence of a small tarry residue in the boiler at the end of the distillation indicated that a little decomposition of the originally colourless mixture had occurred.

' OO ~

1.0

2.0

3.0

time from start of take-off (h) FIG.1.-Temperature profile during spinning band distillation of dimer mixture.

Thus, the thin-layer chromatography and the spinning-band distillation indicated the existence of two principal components in the mixed dimers. This conclusion was supported by GLC analysis of the dimer mixture. In this low-temperature (50°C) GLC work, a 6.35 mm diam. column was used ; it was 0.3 m long and packed with 10 % dinonyl phthalate on crushed firebrick. The flow rate of the helium carrier gas was 1 ml sec-l, and the apparatus employed a katharometer detector. 20 pl. injections were made of (a)the dimer mixture, (b)the lower-boiling fraction (henceforth called fraction l), and (c) the higher-boilingfraction (fraction 2). The results are shown in fig. 2 ; no further peaks were observed during 18 h. Clearly the two components are present in the mixture in approximately equal proportions. The third (intermediate) fraction from the spinning-band distillation showed a chromatogram entirely consistent with that expected for a mixture of the two fractions. Gas chromatographic analysis of the oily inhibitor residue showed peaks corresponding to traces of the components in fractions 1 and 2, together with a new peak of greater retention time. (C) C H A R A C T E R I Z A T I O N OF P R O D U C T S

The elemental analysis and molecular weights of fractions 1 and 2 are consistent with dimeric chloroprene,lO and the structure of the component in fraction 1 was established as 1,2 dichloro-1,2 divinyl-cyclobutane (VI) by infra-red and n.m.r. analyses. The infra-red spectrum of fraction 2 shows two bands in the region 162-168 mm-l which may be attributed to two different forms of unsaturation. A band at 89 mm-I is attributed to the RlR2C = CH2 structure, which is also consistent with bands in the regions 4-2 and 4-8 z of the n.m.r. spectrum. (Chemical shifts in the latter were measured with respect to an internal standard, tetramethyl-silane, for which z = 10). The presence of three methylene

View Article Online

BILLINGHAM, EBDON, LEHRLE, MARKHAM A N D ROBB

473

groups was inferred from a series of bands in the n.m.r. spectrum between 7.0 and 8-07. These results are therefore consistent with conventional Diels-Alder dimer structures I and 111. At this stage it was not possible to assess whether fraction 2 consisted of one or both of these isomers. However, gas chromatographic work described later indicates that both isomers are probably present. Further work was performed to verify that the isolation of the fractions had not caused isomerization or decomposition of the original components. The infra-red and n.m.r. spectra of a mixture of the fractions were identical with those of the unseparated mixture.

I

I

I

I

0

I

I

I

2

3

4

5

I

1

I

I

I

I

I

0

I

2

3

4

5

6

I

]

6

I

retention time (h) FIG. 2.-Low-temperature (SOOC) GLC analyses of dimers ; (a), dimer mixture ; (b), fraction 1 ; fraction 2. This shows not only that no changes in the original components had occurred, but also that there is no detectable yield of dimers other than I, 111, and VI in the dimer mixture isolated as in (b). On the other hand the GLC analyses showed that an additional component was present in the oily inhibitor residues. From the weight of residue, the weight of DPPH, and the gas chromatographic analysis of the oil in the residue, it was calculated that the maximum yield of this component in the reaction of 35°Cwas 7 % of the total products. The infra-red for 1,6-dichloro-cyclospectrum of this component was identical with that published octadiene 1 : 5 (V) and its n.m.r. spectrum was also consistent with this structure. We may therefore conclude that at 35"C,chloroprene dimerizes to form principally 4R and 6R products in approximately equal amounts, and that at this temperature the 8R adduct is a comparativelyminor product. All the products identified are here listed together : 4R. (VI) : 1,2 dichloro-1,2 divinyl-cyclobutane; 6R. (I) : 2,chloro-4,(cr-chlorovinyl)-cyclohexene1 ; 6R. (111) : 1 ,chloro-4,(a-chlorovinyl)-cyclohexene1. 8R. (V) : 1,6 dichloro-cyclooctadiene1 : 5 DILATOMETRIC MEASUREMENTS OF THE OVERALL RATE OF D I M E R I Z A T I O N

Dilatometric measurements of the rate of dimerization were performed to evaluate the overall activation energy and the overall order with respect to monomer concentration. In order to eliminate the possibility of vapour-phase heterogeneous polymerization in the dilatometers, inverted dilatometers incorporating mercury seals were used. The construction

View Article Online

474


and the filling of these dilatometers have been described ;' in the present work the same procedure was followed but the dilatometers were constructed with smaller bore (1 mm) capillary stems. When a solvent was present, complete mixing of this with the monomer was achieved by movement of a glass-enclosed ball-bearing within the dilatometer by means of an external magnet, since vigorous agitation of mercury-seal dilatometers is not practicable.

TABLE 1 . - d ~ AND d~ (EXPT.) temp. ("C)

25-0 35.0 45.0

AT

25, 35 AND 45°C

dM(g/mU

dD(dml)

0.947 &0*001 0.934 f0.001 0.931 f0-001

1.157 f0.002 1 -148 0-002 1.140 f0-002

A knowledge of the component densities is required in dilatometric work. The density of monomer ( d ~and ) that of mixed dimers ( d ~were ) determined experimentally at several temperatures ; the values are listed in table I. The apparent density of the dimers in monomer solution was also measured, and the results were within the uncertainties quoted in the table. These values were used in activation energy calculations, and the same (35°C) values were used for all dilutions in the reaction order calculations. (a) T H E O V E R A L L A C T I V A T I O N E N E R G Y OF D I M E R I Z A T I O N The overall rate of dimerization was measured at 25,35, and 45°C. An Arrhenius plot of the results is linear (see fig. 3) and its slope corresponds to a value of ED = 90.0 f3.8 kJ/moIe for the overall activation energy. This value is concordant with that (21.8 A0.5 kcal/mole = 91.2 &2-0kJ/mole) of Hrabak and Webr.14 I

I

3.1

3'2

3.3

3.4

I / T X103 (OK)-*

FIG.3.-Overall

rate of dimerization : Arrhenius plot.

( b ) T H E O V E R A L L O R D E R OF R E A C T I O N

The results of a series of measurements of overall dimerization rate (35°C) for various concentrations of monomer in MEK are listed in table 2. In no case could any deviation from linearity in the conversion/time plots be detected, and the quoted ratescorrespond to the gradients of these plots. A log-log plot of these results is shown in fig. 4 ; a least squares

View Article Online

B I L L I N G H A M , E B D O N , L E H R L E , M A R K H A M A N D ROBB

475

evaluation of its slope gave a reaction order of 2.1 10.2, which indicates that the overall reaction is second order. Fig. 5 shows a conventional second-order plot of the data; its linearity and extrapolation through the origin again suggest that the reaction is second order. ___

- 60

-6.1

I

-..'1

- 6.8

I

/ / /

d

/

SLOPE=2-I + 0 - 2

LL.-L,-J--

0'4

0.5

0.6

FIG.4.-Overall

0.7

0.8

0.9

1.0

1.1

log1O M 10 rate of dimerization : log-log plot.

I-11 . 00

120

140

80

FIG.S.-Overall

[monomer]; rate of dimerization : second-order plot.

View Article Online

476

CHLOROPRENE D l M E R I Z A T I O N

Several explanations can be offered for the slight trend in the second-order rate constant with concentration (see table 2), i.e., the 0.1 deviation from second order in the least squares line (fig. 4) : (i) the apparent density of the dimer mixture may vary slightly with the concentration of MEK, (ii) there could be a specific solvent or cage effect, or (iii) a simultaneous reaction could be occurring. Explanation (iii) could involve the participation of an almost insignificant reaction of higher order, or the isomerization of the dimers to give a product of higher density. TABLE 2.-DIMERIZATION expt. no.

conversion %

bulk monomer 1 2 3 4

5 6 7 8 9 10 11 12 13 14 15

4.5 4.0 3.5 3.3 3.1 2.8 2.5 2.1 1.9 1.6 1.4 1.2 1.0 0-9 0.8 0.5

RATES IN

M.E.K AT 35"

initial monomer concentration moles /l.

dimerization rate x 107 moles/l. sec

10.56 10-03 9.25 9-44 8-88 8-58 8-05 7.36 7.21 6.48 6-28 5.79 5.26 5-15 4.72 3-58

15-59 14.10 11.67 11.51 10.74 9-63 8.58 7-20 6-65 5-48 4.94 4.11 3.35 3.22 2.90 1*62

109 R ~ I [ ~ I ; I./moles sec

13.98 14.00

13-62 12-93 13.62 13.09 13.21 13.28 12.79 13-11 12-53 12-24 12-10 12-14 12.99 12.61

The gradient of the line shown in fig. 5 corresponds to a rate constant kd = (13.4 f0.5)x 1. mole-1 sec-l at 35°C. Using the experimental value of ED determined above, the second-order rate constant for the overall dimerization reaction can be expressed as k = 2 5 x lo7 exp [-9O,OOO/RT] 1. mole-l sec-l. The uncertainty in the Arrhenius factor may be expressed as logloA = 7.4 f0.3. A is of the same order as values reported for the dimerization of other simple dienes.15 K I N E T I C S O F T H E F O R M A T I O N O F I N D I V I D U A L D I M E R S ; U S E OF T H E P Y R O L Y S I S - G L C TECHNIQUE

In order to examine the kinetics of the formation of the individual dimers, quantitative analysis of the dimeric products is necessary. Such analysis is complicated by the instability of the 4R product (isomerization,dehydrochlorination and decomposition),and any succesful method must either avoid this decomposition, or quantitatively complete the decomposition to give products which can themselves be resolved and measured. (a) L O W - T E M P E R A T U R E

GAS-LIQUID CHROMATOGRAPHY

Whilst the low-temperature GLC procedure described is adequate for examining the dimeric products isolated from monomer, it was unsuitable for measurement of the yields of the individual dimers in monomer (or monomer/solvent). This is because : (i) excessively long retention times were observed (e.g., up to 6 h even when using a short 0.3 m column), (ii) with the short columns necessary, the resolution of the large monomer peak in low conversion reactions was not adequate for quantitative analysis, and (iii) the resolution was also impaired because of the slow volatilization of the dimer components.

View Article Online

B I L L I N G H A M , E B D O N , LEIIRLE, M A R K H A M A N D R O B B

477

( 6 ) H I G H - T EM PERATURE G L C For the optimum resolution of monomer+ dimer mixtures, a 6 m GLC column (packed with 20 % Carbowax 20 m on a 60-80mesh acid-washed Chromosorb P support) was found to be desirable. Such a column must be operated at elevated temperature (up to 200"C), since low-temperature retention times of several days are not convenient in a routine analytical method. The behaviour of the dimer mixture at such high temperatures was therefore investigated. Fig. 6 (i) shows the chromatogram obtained from mixed dimers using the 6 m column at 180°C. M was characterized as the monomer peak, whilst the remaining peaks are associated with dimers, or their decomposition/isomerizationproducts. In order to relate these peaks to the 4R and 6R components, fractions 1 (4R)and 2 (6R)were examinedunder the same conditions ; the chromatograms are shown in fig. 6 (ii) and (iii).

M

0

FIG. 6.-High-temperature

0

20

40

60

80

100

120

140

retention time (min) (180°C) GLC analyses of dimers (i), dimer mixture; (ii), fraction 1 ; (iii), fraction 2.

Two partially resolved peaks (C) are observed for fraction 2, indicating that both structures I and I11 may be present. The possibility that these two peaks resulted from thermal decomposition of a single compound during GLC analysis was excluded by collecting a small quantity of the mixture from the GLC effluents ; the elemental analysis of this mixture and its i.-r. and n.m.r. spectra were identical with those of fraction 2 before injection. Thus the 6R products are thermally stable at 180°C. For fraction 1, however, the elemental analysis after elution (C= 64-7%, H = 6.6 %, C1 = 28.6 %) differed from that of the injected material (C = 54.5 %, H = 5 6 %, C1 = 39.2 %) demonstrating that some dehydrochlorination had occurred. Moreover, the fact that the eluted material displayed an i.-r. spectrum which differed completely from that of the original fraction indicated that extensive decomposition or rearrangement had occurred. By trapping peak D during several gas chromatographic runs, it was possible to collect an analyzable quantity of this component. Its elemental analysis, i.-r. spectrum, and n.m.r. spectrum showed it to be V, the 8R structure. (Spectral analysis also showed that V was not present in the original fraction). No attempt has been made to characterize peaks A and B, though probably at least one of these corresponds to a dehydrochlorination product.

View Article Online

478


(C) Q U A N T I T A T I V E A N A L Y S I S B Y P Y R O L Y S I S - G L C

The analytical problem was to measure the fractional yields of the 4R and 6R products when the total dimer content in monomer was only a few percent. Since the results in the previous section indicate that the 4R product pyrolyses rapidly (note the absence of tailing in the GLC peaks) it appeared probable that a technique could be developed in which the fraction of 4R originally present in the mixture could be measured from its decomposition products. (The stable 6R product presents no problem). The stages in this work were as follows : (i) to choose GLC conditions (especially injection zone temperature and column temperature) so that the 4R pyrolysis was rapid, quantitative, and reproducible, and (ii) to calibrate the apparatus with dimer+monomer mixtures of known composition, so that the areas of the monomer peak, 6R peaks,and 4R decomposition peaks could be used to calculate the composition of samples from kinetic experiments, 1

x *ri P/

,.!i3!I, L -5

i

0

2

I

I

I

I

0

2

4

6

8

FIG.7.-Pyrolysis-GLC

4

6

8

1 0

12

I

I

14

1 0 1 2 1 4

actual concentration % w/w calibration plots : (a), 6R dimers ; (b), dimer mixture.

The optimum analysis procedure was the following : the GLC injection zone is set at 270°C to ensure quantitative pyrolysis and rapid volatilization, and the column temperature initially set at 120°C. 0-5 ml of the mixture is injected into the pyrolysis zone, and the column temperature programmed upwards at 15"C/min until a temperature of 180°C is reached ; the column is then maintained at this temperature during the remainder of the analysis. The column used is identical with that described in (b) above. The calibration plotsare shown in fig. 7a and 7b. " Apparent dimer % " is that calculated from the sum of the areas of the dimer peaks and the sum of the areas of the dimer and monomer peaks. From these results it can be inferred that thedeterminationof the fractional yield of 4R product from a summation of the areas of pyrolysis peaks A, B, and D is quite justified. 6 R P R O D U C T FORMATION The orders of reaction for the formation of 4R and 6R products were found from experiments in which MEK was used as solvent for chloroprene ; the solutions contained 3 x loA2

( d ) T H E O R D E R S O F R E A C T I O N FOR 4 R A N D

View Article Online

B I L L I N G H A M , E B D O N , LEHRLE, MARKHAM A N D ROBB

479

moles/l. of DPPH as polymerization inhibitor. For each monomer concentration, six 3-ml glass ampoules were filled in the following way in order to ensure that each contained an identical composition. The ampoules were initially attached to a central vessel which contained a known weight of inhibitor. This vessel was then attached to the high-vacuum line and charged by distillation with the required volumes of solvent and monomer. After sealing off and removing the apparatus from the line, the individual ampoules were filled by tipping the apparatus. All the contents were then simultaneously frozen in liquid nitrogen, the ampoules were sealed off, and finally immersed in a thermostat for the required periods. The ampoules were removed in sequence at ca. 30-h intervals, and the conversions were measured by pyrolysis-GLC analysis. The GLC conditions and temperature programme described above did not permit good resolution of the monomer and solvent ; however, this was unnecessary since it was established by preliminary calibration experiments that the GLC detector displayed identical weight-response factors for both solvent and monomer. These partially-resolved peaks were therefore measured as a whole. A monomer concentration range of 5.76-10.56 M was examined at 35°C. The maximum conversion in any experiment was 8.5 %, and over this conversion range no curvature was detected in the rate plots; rates were therefore calculated by the initial gradient method. The results are listed in tables 3a and 36. TABLE 3(C2).-RATE

OF FORMATION OF CYCLOBUTANE ISOMERS IN METHYL ETHYL KETONE AT

expt. no.

TABLE3(b).-RATE

monomer concentration (moles /1.)

maximum conversion %

R4X 10'7 moles/l. sec

10.56 (bulk) 9.86 9.38 8-78 7.80 6.91 6.23 5.76

5.0

4.2 4.0 3.5 2-8 2.0 1.8 1.4

9.51 7.56 6.91 5.92 4.70 3.59 2.88 2.30

lO9X

R4/[M]g

8-52 7.77 7.86 7.69 4.74 7.05 7-42 6.93

OF FORMATION CYCLOHEXENE ISOMERS IN METHYL ETHYL KETONE AT

10.56 (bulk) 9.86 9.38 8-78 7-80 6.91 6.23 5-76

4.40 3.54 3.24 2.82 2.24 1.72 1.41 1.17

3.0 2.4 2.2 2-0 1.5 1.2 1.0 0.8

35°C

35°C

3.95 3.64 3.68 3.66 3-68 3.60 3.63 3-53

Log-log plots of the initial rates against initial monomer concentrations are shown in fig. 8. The gradients indicate an order 2.1 1 0 . 2 for 4R formation and 2.OrtO-1 for 6R formation. (2.1 f0.2 for overall dimer formation.) Conventional second-order plots of the results are shown in fig. 9 ; the linearity and extrapolation of these plots through the origin strongly suggests that the reactions are second order. The following 35°C rate-constants (1. mole-l sec-l) were calculated from the plots in fig. 9 : k4 = (7.8 f0.3)x . . . 4R formation, k6 = ( 3 - 7 f 0 . 3 ) ~lo-' . . . 6R formation, k = (11.6fO-5)x . . overall dimerization. The overall rate compares favourably with that determined dilatometrically.

.

View Article Online

480


0.7

0.9

0.8

1.0

~og,o[MIo

FIG.8 . 4 R a n d 6Rdimerizations :log-logplots : (a),overallreaction; (b),4Rreaction ;(c)6Rreaction.

0

20

40

60

80

100

120

[monomer]; FIG.9 . 4 R and 6R dimerizations : second-order plots : (a),overall reaction ; (b),4R reaction ; (d, 6R reaction.

View Article Online

BILLINGHAM,

481

EBDON, LEHRLE, MARKHAM A N D ROBB

(t?) T H E A C T I V A T I O N ENERGIES A N D ENTROPIES O F A C T I V A T I O N FOR PRODUCT FORMATION

4R

AND 6R

Dimerization reactions of pure chloroprene, inhibited with DPPH, were examined within the temperature range 30-60°C. The experimental technique was similar to that described in the previous section. Arrhenius plots of the results are shown in fig. 10, and the corresponding activation energies are listed in table 4. Assuming unit transmission coefficient (rc) in the conventional transition state theory expression, k = rc(kT/h)exp (AS* IR) exp (- EIRT), values of the entropy of activation (AS*)were also calculated, and these are shown in the same table, which includes the dilatometric results for comparison.

---__I

3.0

3.1

3.2

-

3.3

I / T X103 ("~-1) FIG.1 0 . 4 R and 6R dimerizations : Arrhenius plots : (a), overall reaction ; (b), 4R reaction; (c), 6R reaction.

TABLE 4.-KINETIC order of reaction

products

cyclobut anes cyclohexenes overall (GLC) overall (dilatornetry)

2-1 k0.2 2.0 s o - 1 2.1 SO-2 2.1 h0.2

PARAMETERS FOR CHLOROPRENE DIMERIZATION 109xk E log 1oA -AS+/R l./mole sec kJ/mole 7.8 f0.3 85.8 f4.2 6 . 4 h0.5 15.7 f1.1

3*7*0-3 11-6f0.5 13.4f0.5

84.1 *4*2 86-2f4-2 90*0&3*8

5*8*0*5 6-7fO-6 7.4&0.5

17.1 f l . 1 15.1 + l - 2 13.5fl.l

(f)S O L V E N T EFFECTS A few experiments were performed to assess whether there is an appreciable solvent effect on the ratio of products obtained. A series of reaction ampoules, each containing 66 % monomer in a different solvent, was prepared. In each case the system was inhibited with excess DPPH. For the solvents tetrahydrofuran, MEK, benzene, and cyclohexane, the percentages of the 4R dimer in the dimeric yields were respectively 62, 6 4 , 66, and 67 after thermostatting the ampoules at 35°C for 120 h. The corresponding percentage obtained with pure monomer under the same conditions was 64. Thus, there is no strong solvent dependence of the product distribution. 16

View Article Online

482


(9)I N F L U E N C E OF L IGHT

Since 1,2-dichloro-l,2-divinyl-cyclobutane (VI) has not previously been reported as a product of the thermal dimerization of chloroprene, it is desirable to check whether ambient (u.-v.) light could influence the formation of this isomer. This question arises because cyclobutane derivatives are formed when butadiene or isoprene are irradiated with u.-v. light in the presence of suitable photosensitizers,16*l7 and in the presentwork no precautions were taken to exclude light from the reaction ampoules in the thermostat bath. Thus, experiments were performed in which the reactions occurred in total darkness, and also in which they were subjected to u.-v. irradiation. (Osram MB/U lamp outside the window of the thermostat.) The results are listed in table 5, which includes the conventional measurements for comparison. The correspondence of the rates shows that the dimerization reactions are unaffected by ambient light and even by u.-v. irradiation.

TABLE 5.-EFFECT

OF LIGHT ON CHLOROPRENE DIMERIZATION AT reaction conditions R , %/h k4/k 6

total darkness normal u.-v.

0.048 10*003 0-04710-002 0.047 f0.002

35°C

2-10f0.06 2.11 &0.05 2.09 i-0.06

DISCUSSION OBSERVED P R O D U C T S

The present work shows that the principal products of chloroprene dimerization are 4R and 6R; only a few percent of an 8R product is formed at 35°C. These results differ from those of other workers, who have reported exclusively 6R and 8R products. It is most probable that the explanation of this lies in the fact that high temperatures were involved in either the preparations or separations performed by these workers ; at such temperatures a considerable proportion of the 4R product may undergo rearrangement reactions of the Cope type. Thus, Vogel l8 suggests that the production of cyclooctadiene-1 : 5 in butadiene dimerization results from the prior formation of cis 1,2-divinylcyclobutanewhich rearranges. He also predicted l9 that cis 1,2-dichloro-1,2-divinylcyclobutanewould more readily isomerize by this mechanism, owing to the ring strain imposed by the relative proximity of the chlorine atoms. Recently Hammond et aLZohave shown that trans 1,2-divinyl cyclobutane can undergo unimolecular isomerizations to form cyclooctadiene 1,5- and 4-vinyl cyclohexene. These authors also studied the rearrangement reactions of photochemically-produced 4R dimers of isoprene, and reported the following isomerizations of the cis and trans dimers respectively :

/ j R

',

/

CH3

CH3

\

Q

View Article Online


483

Similar results were obtained by Trecker et aL21 It therefore seems feasible to propose analogous reactions for the corresponding chloroprene dimers. The pyrolysis-GLC results in the present work provide direct evidence of the occurrence of a Cope rearrangement as above. A peak identified as due to 1,6dichloro-cyclooctadiene-1 : 5 was observed in pyrolysis chromatograms of dimer samples which had been shown spectroscopically to contain none of this compound prior to injection. Our characterization studies indicate that the 4R dimer formed by 1 : 2, 3 : 4 addition [ 1,chloro- 1,vinyl-2,(cc-chlorovinyl)-cyclobutane]is absent from the products of low-temperature dimerization. This compound would be expected, by analogy with the corresponding isoprene dimer rearrangement, to isomerize as follows : CL

VII

VIII

I1

The absence of isomer VIII from the products of high-temperature GLC is thus consistent with the absence of VII from the products of the low-temperature dimerization. OVERALL KINETICS

From the dilatometric results it is concluded that the overall dimerization reaction is second order in monomer, and that the Arrhenius factor is comparable with values reported for the dimerization of other simple dienes. Hrabak and Webr l4 have carried out dilatometric studies of the dimerization of chloroprene in the presence of DPPH, using n-hexane as solvent, and obtained a value of 1.7 for the order of reaction which they believed indicated a second-order reaction. They attributed their low experimental value to the participation of a concurrent reaction between DPPH and peroxy radicals ; the latter could be derived from the monomer and their reaction with DPPH would be first-order in m0n0mer.l~ This proposal implies that the overall observed reaction will be of mixed first and second order, although they did not verify this by a mixed order plot of rate/[M] against [MI. A second-order rate constant was calculated directly from the measurements ; the value obtained was 2-4 x lo-* 1. mole-l sec-l at 40°C. Our dilatometric rates and kinetic parameters are believed to be more reliable because (a) the concentration of monomer peroxide in our systems is known to be small, since negligible polymerization of monomer occurs in the absence of DPPH, and ( B ) the order observed is much closer to two. K I N E T I C S O F T H E I N D I V I D U A L P R O C E S S E S ; M E C H A N I S T I C C O N S I D ER ATIONS

(a) MECHANISTIC IMPLICATIONS OF THE KINETIC PARAMETERS. The activation energies for 4R and 6R product formation are identical within experimental error, and the corresponding entropies of activation are similar. An interpretation of this is that the transition states for both processes have similar structure; it might be

View Article Online

484


postulated that the same transition state (or intermediate) is common to both processes. These ideas imply that it is extremely unlikely that the cycloadditions proceed by a mechamism involving simultaneous two-centre-attack ; one-centre-attack transition states (represented here for convenience as diradicals) of the following types must be considered :

Ix

X

XI

There are two reasons why structure IX is to be preferred to X and XI : (i) in IX, both of the radical centres are stabilized by allylic interactions with the neighbouring vinyl groups, and (ii) the possible ring closures (shown by dotted lines) indicate that only IX may form the observed products ; X would lead to a 1,3-substituted cyclobutane and XI would not easily form a cyclohexene derivative. Since many Diels-Alder reactions are known in which the dienophile component retains its stereo-chemical integrity throughout the reaction, it has often been discussed whether Diels-Alder reactions might most generally occur by a mechanism involving two-centre attack.l If the argument presented above is accepted, it would appear that the dimerization of chloroprene is an example of the alternative onecentre-attack mechanism. The fact that the radical scavenger (DPPH) does not inhibit chloroprene dimer formation may indicate that the transition state is not correctly represented as having diradical character, or that the diradical intermediate cyclizes too rapidly to allow reaction with DPPH.

(b) MECHANISTIC IMPLICATIONS OF THE CHLORINE SUBSTITUTION. A consideration of the positions of the chlorine atoms in the observed products of chloroprene dimerization shows that no single transition state of type IX is sufficient. If chlorine migration during reaction is unlikely, all three of the possible forms of IX must exist, lee.,

XII

XI11

XIV

Thus, XI1 forms the cyclobutane derivative, and XI11 and XIV form the cyclohexene derivatives 111 and I respectively, when cyclization occurs as shown. The unusual formation of the cyclobutane derivative in high yield in this thermal reaction might arise because intermediate XI1 derives greater stability than XI11 or XIV from the chlorine susbtitution. (Chlorine substitution on the terminal carbon atom of an " allylic " radical might be expected to increase the stability of the radical, whereas substitution on the central carbon atom would have no such stabilizing effect.) If three such transition states participate in chloroprene cycloadditions, the kinetic

View Article Online

48 5


evidence above implies that the free energies of formation of these states must be similar. (c) SOLVENT EFFECTS. In an attempt to distinguish between one-centre- and twocentre-attack mechanisms, Little 2 2 studied the reaction of butadiene with a-cyano vinyl acetate, which gives the products 4-cyano-4-acetyl-cyclohexene 1 and l-cyano-lacetyl-2-vinyl-cyclobutane. The ratio of the product yields was almost unaffected by changes in solvent. (The solvents examined were benzene, toluene, nitromethane and acetonitrile). On the basis that two transition states of significantly different structure would not be expected to show the same solvent effect, Little concluded that his results support a two-step (i.e. one-centre-attack) mechanism. In contrast, Stewart23 studied the reaction of 4-methyl-pentadiene-1 : 3 with tetracyanoethylene : CN

f C H 3

CN

+ >( CN

CN

-

'"')=,H

= "3

, %'"

CH3

CHj

+

CN

CN

'CN

and in this case the product ratio was strongly solvent-dependent. Stewart therefore concluded that the adducts are formed by independent reactions, which could allow two-centre attacks to give different transition states. With chloroprene, the product ratio is independent of solvent within experimental error. This result is consistent with the postulation of one-centre-attacks to give transition states which are structurally similar and which differ only in chlorine substitution.

(d) EFFECT OF LIGHT. The formation of the 4R and 6R dimers from chloroprene is a purely thermal process ; the rates were unaffected by ambient light and even by u.-v. irradiation. We now consider the extent to which this observation can be related to theoretical predictions of the dimerization reaction. Hoffmann and Woodward predicted the steric course of cycloaddition reactions by a method involving correlation diagrams for the molecular orbitals involved, classifying the levels with respect to the symmetry elements of the transition state. They conclude that the thermal formation of cyclobutane derivatives by a one-step (i.e., two-centre-attack) process is improbable, and suggest that the formation of such 4R products in a Diels-Alder reaction must result either from photochemical activation or from a two-step (one-centre-attack) process. The latter, by analogy with other 1 : 2 cycloaddition reactions,24 is assumed to involve an intermediate having diradical-like characteristics. These ideas are entirely consistent with the general mechanism proposed in $ (a) and (b) above. C O M P A R I S O N OF C H L O R O P R E N E D I M E R I Z A T I O N S W I T H T H O S E OF B U T A D I E N E A N D O T H E R SYSTEMS

Benson has attempted to deduce the mechanism of the Diels-Alder reactions of butadiene from a consideration of kinetic data on the following reactions : (i) The thermal dimerization of butadiene to give vinyl cyclohexene. (ii) The thermal decomposition of cyclooctadiene to give butadiene and vinyl cyclohexene. (iii) The thermal decomposition of vinyl cyclohexene to give butadiene. (iv) The solution pyrolysis of trans- 1,2-divinyl-cyclobutane to give butadiene, vinylcyclohexene, and cyclo-octadiene.

View Article Online

486


The scheme which he proposes may be summarised as

Thus, a common diradical intermediate is postulated for the 4R, 6R, and 8R dimerizations of butadiene. The standard enthalpy of formation and the standard entropy of formation of this diradical were calculated by the method of group additivity; this method was also applied for all other components except butadiene. These thermochemical data were used to calculate the equilibrium constants which were involved in demonstrating the validity of the scheme. Superficially, Benson’s conclusions add weight to our mechanistic proposals for chloroprene 4R and 6R dimerizations ; in both systems an octadiene, 1-7-diyl, 3-6diradical is assumed as the structure of an intermediate formed by one-centre attack. Transition state

React ants

rn

Transition state

Reactants

Product (b)

FIG.1 1 .-Schematic free-energy diagrams for butadiene reactions (see text) : (a), based on Benson’s diradical hypothesis ; (b), conventional mechanism, involving only the transition state.

However, we are not convinced that this scheme has been established (rather than proposed) for the butadiene system. A rather simplified illustration indicates the fallacy of some of Benson’s arguments. Considering a Benson free energy diagram incorporating the hypothetical diradical intermediate as shown in fig. 1l(a), his calculation of the rate constant for the reverse reaction implicitly involves entropy and

View Article Online

B I L L I N G H A M , E B D O N , LEHRLE, M A R K H A M A N D R O B B

487

+ +

heat summations p (q r ) . On the other hand, the same summations [ ( p+q) + r] are involved if no diradical intermediate (fig. 1l(b) is invoked, and an identical rate constant is obtained. Thus, his satisfactory prediction of such a reverse reaction rate constant is not evidence for the participation of the diradical intermediate. Only for one of the reactions (cyclooctadiene+vinyl cyclohexene) is there convincing evidence for the diradical intermediate ; in this case the predicted energy level difference between the diradical and the reactant is almost identical with the experimental activation energy. Unfortunately, this is an isomerization reaction, and is not necessarily related to the butadiene dimerization process. From a study of the dimerization of styrene, Mayo 2 5 has concluded that the products of the simplest thermal processes correspond to both reactions :

Thus styrene and chloroprene differ from butadiene in dimerization behaviour, in that the former monomers form a 4R product thermally, whilst butadiene does not. The explanation of the styrene behaviour may be the same as that offered earlier for chloroprene, i.e., that the substituent groups (in this case phenyl) confer stability on the free radical sites of an intermediate diradical. Studies of the addition of l,l-dichloro-2,2-difluoro-ethylene to various dienes to produce substituted cyclobutanes 24 support the view that diradical intermediates are involved in the formation of these 4R adducts. The retention of stereochemistry characteristic of many Diels-Alder processes is not found in these systems; this indicates that free rotation is possible in the transition state. The evidence presented in the present paper therefore suggests that the unusual behaviour of chloroprene in thermally dimerizing to give both cyclobutane- and cyclohexene-type products may involve a mechanism of the following type. The three possible types of one-centre attack ( 1 , l ; 1,4 ; 4,4) all occur to give intermediates which are probably diradical in character and which possess similar free energy. The final cyclization occurs to give cyclohexene-type products except for the 1,l intermediate, where the chloro-substitution influences the cyclization to give the substituted cyclobutane. The formation of the cyclooctadiene derivative, (especially at high temperature) probably occurs by a Cope rearrangement of the dichloro divinyl cyclobutane ; since this rearrangement does not occur quantitatively even at pyrolysis-GLC temperatures, it is possible that both cis- and trans-dichloro divinyl cyclobutane are initially formed, and only the former isomerizes to the cyclooctadiene compound. We thank the Science Research Council for the award of a maintenance grant to N. C . B., the Distillers (BP) Co. Ltd., for the award of a maintenance grant to J. R. E., and British Industrial Plastics, Ltd., for the award of a research scholarship to J. L. M.

View Article Online

488

CHLOROPRENE D I M E R I Z A T I O N

A. C. Cope and W. J. Bailey, J . Amer. Chem. SOC.,1948,70,2305. A. C. Cope and W. R. Schmitz, J. Amer. Chem. Soc., 1950,72, 3056. A. L. Klebanskii and M. M. Denisova, Zhur. Obschei Khim., 1947,17,4. I. N . Nazarov and I. A. Kuznetsova, Zhur. Obschei Khim., 1960, 30, 134. R. Hoffman and R. B. Woodward, J. Amer. Chem. SOC.,1965, 87,2045. S. W. Benson, J. Chem. Physics, 1967, 46,4920. ' P. A. Leeming, R. S. Lehrle and J. C. Robb, S.C.I. Monograph (Society o f Chemical Industry, London, 1963), no. 29, p. 203. * N. C. Billingham, P. A. Leeming, R. S.Lehrle and J. C. Robb, J. Polymer Sci., C, in press. N. C. Billingham, J. R. Ebdon, R. S. Lehrle, J. L. Markham, and J. C. Robb, in preparation. l o N. C. Billingham, P. A. Leeming, R. S. Lehrle and J. C. Robb, Nature, 1967, 213,494. S. D. Lessene and H. L. Lochter, Ind. Eng. Chem. ( A d ) , 1938, 10, 450. l 2 N. C. Billingham, Thesis (University of Birmingham, 1967). I 3 A. C. Cope and W. R. Schmitz, J. Amer. Chem. SOC.,1950,72, 3056. F. Hrabak and J. Webr, Makromol. Chem., 1967,104,275. 'l4 A. Wassermann, Diels-Alder Reactions, (Elsevier, Amsterdam, 1965). l 6 G. S. Hammond and W. M. Hardham, J. Amer. Chem. Soc., 1963, 85, 63. l7 G. S. Hammond, N. J. Turro and A. Fisher, J. Amer. Chern. SOC.,1961, 83, 4674. '* E. Vogel, Angew. Chem., 1963, 2, 1. l 9 E. Vogel, Annulen, 1958, 615, 1. 2 o G. S. Hammond and C. D. DeBoer, J. Amer. Chem. SOC.,1964,86,899. 2 1 D. J. Trecker and J. P. Henry, J. Amer. Chem. SOC.,1964, 86,903. 2 2 J. C. Little, J. Amer. Chern. Soc., 1965, 87, 4022. 23 C. A. Stewart Jr., J. Amer. Chern. Soc., 1962, 84, 117, 24 P. D. Bartlett, L. K. Montgomery and B. Seidel, J. Amer. Chem. Soc., 1964, 86,616, 622, 628. 2 5 F. R. Mayo, Amer. Chem. SOC. Polymer Preprints, 1961,2,55 ; J. Amer. Chem. Soc., 1968,90, 1289.