Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The pyrolysis of diallyl and reactions of allyl radicals with l-butene. 1958

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata

Download

Media
UBC_1958_A6_7 R8 P9.pdf
UBC_1958_A6_7 R8 P9.pdf
UBC_1958_A6_7 R8 P9.pdf [ 3.17MB ]
Metadata
JSON: 1.0062213.json
JSON-LD: 1.0062213+ld.json
RDF/XML (Pretty): 1.0062213.xml
RDF/JSON: 1.0062213+rdf.json
Turtle: 1.0062213+rdf-turtle.txt
N-Triples: 1.0062213+rdf-ntriples.txt
Citation
1.0062213.ris

Full Text

T H E P Y R O L Y S I S O F D I A L L Y L A N D R E A C T I O N S O F A L L Y L R A D I C A L S W I T H 1 - B U T E N E by DALIBOR J . RDZICKA cand. mag. (University of Oslo) 1956 A THESIS SUBMITTED IN PARTIAL FIIEJFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE i n the Department of Chemistry We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA June, 1958 A B S T R A C T A study has been made of the pyrolysis of 1,5-hexadiene ( d i a l l y l ) and of the reactions of the a l l y l radicals thus produced with 1-butene. The pyrolysis of 1,5-hexadiene was found to be f i r s t order and appears to proceed mainly by free radical reactions i n the temperature range 460 - 5209 C. However, molecular rearrangement i n the formation of cyclo- hexene and benzene are possible. The over-all activation energy for the 8.5 —1 process was found to be 31.3 kcal/mole for an A factor of 10 * min"* . The main products among the gaseous compounds were methane, ethane, ethylene, propylene, and 1-butene. The main l i q u i d products were cyclopentene, cyclo- pentadiene, 1-hexene and benzene. Evidence was obtained for hydrogen abstraction by a l l y l and addition of the a l l y l radical to o l e f i n i c double bonds. Activation energies for the formation of methane, ethane, propane, ethylene, propylene and 1-butene were determined. Addition of approximately 5 i° by volume of 1,5-hexadiene was found to increase the rate of the pyrolysis of 1-butene about 6 - 7 times at 506° C. The formation of C^-cyclic products and methane was found to be proportional 1/2 to ( d i a l l y l ) . This provides kinetic evidence that a l l y l radicals are involved directly i n the formation of these products i n the sensitized decomposition and also that the primary step i n the pyrolysis of d i a l l y l involves the formation of two a l l y l radicals. Mechanism for the pyrolysis of d i a l l y l and for the reactions of a l l y l with 1-butene are discussed. A brief study has also been made of the pyrolysis of 1-butene i n the presence of acetaldehyde. The thermal decomposition of 1-butene sensitized by addition of 25 ia by volume of acetaldehyde was carried out at 477° C. - i i i - The decomposition of acetaldehyde was found to be inhibited by the presence of 1-butene by a factor of more than two. 1-butene, which practically does not decompose at the above temperature by itself, de- composes appreciably in the presence of acetaldehyde. In presenting this thesis i n p a r t i a l fulfilment of the requirements for an advanced degree at the University of B r i t i s h Columbia, I agree that the Library shall make i t freely available for reference and study. I further agree that permission for extensive copying of t h i s thesis f o r scholarly purposes may be granted by the Head of my Department or by his representative. I t i s understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of Chemistry The University of B r i t i s h Columbia Vancouver 8, Canada June 27, 1958 - i v - A C K N O W L E D G E M E N T S This investigation was carried out under the direction of Dr. W.A. Bryce to whom the author i s greatly indebted. Thanks are due to Mr. S.A. Ryce for the use of his gas chromato- graphic apparatus as well as for many valuable discussions. Mr. E.W.C. Clarke carried out the mass spectrometric analyses which was gratefully appreciated. The author i s also indebted to the Defence Research Board for financial assistance during the course of th i s work and to the B r i t i s h Columbia Sugar Refining Company Limited for a graduate scholarship (1957 - 1958). - V - T A B L E OF C O H I E H S Page DiTRODUCTION 1 GENERAL 1 HISTORICAL 2 EXPERIMENTAL 4 PYROLYSIS APPARATUS 4 METHOD OP PYROLYSIS 6 ANALYSES OP THE REACTION PRODUCTS BY GAS CHROMATOGRAPHY 7 Apparatus 7 Separation of the products 9 Product Identification 10 Quantitative determination of the products 12 Qualitative and quantitative determination of hydrogen 13 Analytical results 14 RESULTS 17 I. PYROLYSIS OP 1,5-HEXADIENE 17 Order of the reaction 17 Pressure increase i n reaction system 17 Rate and over-all energy of activation for the pyrolysis of 1,5-Hexadiene 17 Rates of formation and over-all activation energies for the l i g h t hydrocarbons 19 Temperature and time dependence i n the formation of products 26 I I . PYROLYSIS OP 1-BUTENE IN THE PRESENCE OP DIALLYL 32 Introduction 32 - v i - Page Results for the decomposition of 1-butene sensitized by d ia l ly l 52 III. PYROLYSIS OP n-BDTANE IN THE PRESENCE OP 1,5-HEXADIENE 57 If. PYROLYSIS OP 1-BUTENE IN THE PRESENCE OP ACETALDEHYDE 59 DISCUSSION 45 PRECISION OP THE QUANTITATIVE DETERMINATIONS OP GAS CHROMATOGRAPHY 43 UNCERTAINTY IN THE VALUES OP THE KINETIC RATE CONSTANTS 44 THE OVER-ALL ACTIVATION ENERGY POR THE DECOMPOSITION OP 1,5-HEXADIENE 46 MECHANISM OP THE PYROLYSIS OP 1,5-HEXADIENE 43 General 48 Reactions of the a l l y l radical 49 (1) Hydrogen abstraction by a l l y l 49 (2) Addition of a l l y l to double bond 51 (5) Combination of a l l y l with other radicals 55 The fate of the 1,5-hexadienyl radical 54 (1) Rearrangement 54 (2) Decomposition 55 (5) Combination with other radicals 55 Formation and reactions of other radicals 55 Mechanism for the formation of the pyrolysis products 57 MECHANISM OP THE THERMAL DECOMPOSITION OP 1-BUTENE SENSITIZED BY DIALLYL 60 (1) Addition to the double bond 60 (2) H-abstraction 61 DECOMPOSITION OP n-BUTANE SENSITIZED BY DIALLYL 62 - v i i - Page BIBLIOGRAPHY 64 APPENDIX MASS SPECTROMETRY IDENTIFICATION OF SOME OF THE PRODUCTS 66 - v i i i - TABLES Page I. Products of 1,5-Hexadiene Pyrolysis (Mole at Various Temperatures " 15 II . Rate Constants for the Over-all Decomposition of 1,5-Hexadiene 20 III. Rate Constants for the Overfall Decomposition of 1,5-Hexadiene for Different Reaction Times 20 IV. Rate Constants (min - 1) for the Formation of the Light Products 23 V. Energies of Activation and Frequency Factors for the Formation of the Light Products 23 VI. Decomposition Products of 1,5-Hexadiene Pyrolysis as a Function of Time at 501° C. 31 VII. - Composition of the Reaction Mixture for Sensitized Decomposition of 1-Butene with 1,5-Hexadiene 35 VIII. Main Products of Pyrolysis of n-Butane and Sensitized Pyrolysis of n-Butane hy Addition of Appr. 5 i° by Volume of 1,5-Hexadiene at 506° C. Reaction Time 5 Minutes. 38 IX. Percentage of Decomposition of Acetaldehyde and 1-Butene at 477° C. 40 - ix- FIGURES l a . Pyrolysis Apparatus l b . Circuit Diagram of Furnace Heaters 2a. Gas Chromatography Apparatus 2b. Charcoal Trap System 2c. U-Tube Trap 3. Gas Chromatographic Separation of the Pyrolysis Products 4. Pressure Change vs. Time for 1,5-Hexadiene 5. Arrhenius Plot for Decomposition of 1,5-Hexadiene 6. Arrhenius Plots for the Light Saturated Products in the Pyrolysis of 1,5-Hexadiene 7. Arrhenius Plots for the Light Unsaturated Products in the Pyrolysis of 1,5-Hexadiene 8. Light Products of the Pyrolysis of 1,5-Hexadiene vs. Temperature. Reaction Time 5 Minutes 9a and 9b. Hydrogen and the Heavy Products of the Pyrolysis of 1,5-Hexadiene vs. Temperature. Reaction Time 5 Minutes. 10. Products of the Pyrolysis of 1,5-Hexadiene vs. Time at 501» C. 11. Heavy Products of the Pyrolysis of 1,5-Hexadiene vs. Time at 501« C. 12. Pressure Change vs. Time at 506° C. Ini t ia l Pressure 200 mm. 15. Products of Sensitized Decomposition of 1-Butene vs. 14. Extent of Decomposition at 477° C. Ini t ia l Pressure 200 mm. I. Acetaldehyde + Helium (l:jj) II. Acetali + 1-Butene (l:5) Page 5 5 8 8 8 11 18 21 24 25 27 28 29 50 33 36 I I N T R O D U C T I O N -1- I N T R O D P C T I O N GENERAL Previous work on the thermal decomposition of 1-butene (4,12,19) provided very good evidence that the primary step in the process involves formation of methyl and a l l y l radicals. By studying the decomposition of 1-butene sensitized by methyl radicals from mercury dimethyl (lO) i t was further established that the methyl radicals play an important role i n the mechanism of the subsequent decomposition. It i s to be expected that the a l l y l radical formed in the primary step would also participate in the mechanism of the decomposition. L i t t l e i s known about the reactions of the a l l y l radical in hydrocarbon systems and therefore a study was undertaken of the fate of a l l y l in the pyrolysis of 1-butene by sensitizing the decomposition with the a l l y l radical. Mercury d ia l ly l and other metal-allyl compounds would be suitable sources of a l l y l radicals for the investigations. However, they are not available, nor are they mentioned in the literature. It was therefore necessary to use a hydrocarbon as a source of a l l y l . Lossing and co-workers (12) have observed a l l y l radicals to be the main radical products of the pyrolysis of 1,5-hexadiene (diallyl) in the temperature range 690 -890° C in a mass-spectrometer with a flow reactor. The decomposition of 1,5-hexadiene in a static system at temperatures around 500° C should thus most probably involve generation of a l l y l radicals i n the primary step. It was therefore chosen as the source of a l l y l radicals for the present study. -2- A preliminary study of the pyrolysis of d i a l ly l i t se l f was under- taken to gain some insight into the mechanism of i t s decomposition before using i t as a source of a l l y l radicals in reactions with 1-butene. HISTORICAL Very l i t t l e is known about reactions of a l l y l radicals. Mooney and Ludlam (16) studied the pyrolysis of a l l y l bromide but radical mechanisms were not known at that time. This pyrolysis has been studied recently by Szwarc, Gosh and Sehon (23»24) using the toluene carrier technique. They ascribe the overall activation energy of 47»5 kcal to the dissociation of the al lyl-Br bond in the primary step. Szwarz (22) has also obtained excellent kinetic evidence for the formation of a l l y l and hydrogen i n the pyrolysis of propylene i n a flow system at 680-870° C. Taylor and Smith (25) mentioned that the resonance characteristics of the a l l y l radical result in a weak al lyl-H bond in propylene. Hence in hydrocarbon systems the resonance stabil i ty of the a l l y l radical (2,5) would be expected to have a long l i f e and to disappear by reactions such as methyl + a l l y l — ^ 1-butene or by combination with i t se l f to give d i - a l l y l . According to the authors the reaction: a l l y l + C 5 H 6 . CgH^ may be involved in the polymerization of propylene sensitized by the photolysis of mercury dimethyl, and in the mercury-photosensitized re- actions of propylene. Hydrogen abstraction by a l l y l has been demonstrated by several workers: Molera and Stubbs (15) have postulated that hydrogen abstraction by a l l y l takes place from isobutene i n the pyrolysis of isobutene. -3- Yery recently McNesby and Gordon (14) have studied the decomposition of cyclopentane in presence of acetone and found: at 381° very l i t t l e hydrogen abstraction by a l l y l takes place, a l l y l combines with CH^ to give 1-butene. at 453° 1-butene is s t i l l present i n the system, but hydrogen abstraction by a l l y l both from cyclopentane and from CD-COCD, 3 3 takes place* at 500° extensive abstraction of hydrogen by a l l y l takes place but also small amount of 1-butene is formed. Some work has also been done on the reactions of a l l y l i c radicals i n solution ( l ) . The information to date on the reactivity of a l l y l radicals in hydro- carbon systems can be summarized as follows: (1) the resonance stabi l i ty of about 17 kcal/mole (average of values given by 2 and 5) leads to appreciable life-time of the a l l y l radical so that combination with other radicals may take place. (2) abstraction of hydrogen at temperatures above about 450° C takes place. (3) addition of a l l y l to olefinic double bonds may take place. E X P E R I M E N T A L -4- E X P E R I M E N T A L PYROLYSIS APPARATUS The pyrolysis of 1,5-hexadiene was carried out in an all-glass static system (f ig. l a ) . V is a quartz vessel of a volume of about 300 ml placed in an electric furnace (p), A^ to A^ are reservoirs for the gases to be pyrolysed, P i s a sampling pipette with a metal-teflon tap to avoid adsorption i n stopcock grease. M^ and Mg are manometers. The furnace was heated by means of three heating elements connected as shown in f i g . l b . The potentials across the three heating elements e l ' e 2 ' ey w e r e adjusted so that temperature gradient along the reaction vessel did not exceed 1°C. Temperature measurements were made by means of two chromel-alumel thermocouples placed i n position a and b. A rough calibration curve "applied voltage vs. furnace temperature" was constructed and used for approximate temperature setting. Pine temperature adjustments were obtained by switching the resistance r on or off. The heat capacity of the system was increased by f i l l i n g the space between reaction vessel and furnace with sand. The purpose of this was to keep the temperature constant for sufficiently long periods of time to carry out a reaction even when str ict steady state between the furnace and the surroundings was not reached. This technique proved satisfactory. It was possible to keep the furnace temperature constant during the pyrolysis experiments to within 19 C at 500" C. In order to prevent condensation in the manometer M ând the adjacent capillaries, this part of the system was heated electrical ly to about 50 - -5- to vacuum Fig , la Pyrolysis Apparatus Ampm. Pig. lb Circuit Diagram of Furnace Heaters -6- 60B C when necessary. METHOD OF PYROLYSIS When pyrolysis experiments are to be carried, out in a static system there is the problem of admitting the reactant with a certain i n i t i a l pressure i n the reaction vessel. At the same time the starting point of the reaction should be reasonably well defined. In order to satisfy both requirements, the volume of the system was increased by attachment of vessel A^. A pre-expansion volume was thus created. A calibration curve was constructed giving the dependence between the pressure in the pre-expansion volume (read on manometer M )̂ and the pressure in the whole system (which is equal to the i n i t i a l pressure in the reaction vessel) at different temperatures. Before each experiment the system was pumped down to "black vacuum" which was tested with a Tesla c o i l . From the calibration curve the pressure in the pre-expansion volume corresponding to the desired i n i t i a l pressure i n the reaction vessel was read. The pre-expansion volume was f i l l e d with the gas or mixture of gases to be pyrolysed. In the case of a mixture 5 minutes was considered to be sufficient for the gases to mix completely. By opening the stopcock S^ the reactant was admitted into the reaction vessel. The moment when the mercury column in manometer M^ settled at a certain value was taken as the beginning of the reaction and the i n i t i a l pressure was read. During the course of the reaction the pressure i n the reaction vessel was read at 1 minute intervals. At the end of the reaction the stopcock Ŝ  was opened (S^ was closed) such that the products expanded freely into the evacuated sampling pipette. -7- ANALYSSS OF THE REACTIOH PRODUCTS BY GAS CHROMATOGRAPHY Apparatus The separation of the products of the pyrolysis of 1,5-hexadiene and of the a l l y l sensitized pyrolysis of 1-butene was effected by gas chromatography. A gas chromatograph constructed i n the laboratory was available (f ig . 2a). With the exception of Hg analyses where nitrogen was employed, helium was used as the carrier gas i n a l l analyses. A self-compensating thermal conductivity ce l l (18) was the detector. Several gas chromatographic columns with different packings were available. The unbalance signal from the thermal conductivity ce l l mounted in a bridge circuit was recorded on a chart of a Speedomax recording potentiometer. The existing gas chromatograph had to be modified for the present analysis: a) a second column furnace was built and attached to the apparatus so that two chromatographic columns, one for light hydrocarbons and the other for heavier hydrocarbons, could now be connected in paral lel . This enabled a quick switch from one column to the other by merely turning the stopcocks S^ and S^ (f ig. 2a). b) i t was found that the amount of sample admitted into the bypass by free expansion from the sampling pipette was insufficient. A Toepler pump was therefore attached to the admission system. The efficiency of the pump was such that in the f i r s t pumping operation the pressure i n the admission volume was increased more than four times. c) Moisture was shown to be not the only impurity in the helium carrier gas, and therefore the CaClg purifying tower alone was not sufficient. Two charcoal traps were f itted to the apparatus as indicated in f i g . 2b. -8- u a> § s — © - p 2 W o Columns with Heaters Toepler Pump Pig. 2a Gas Chromatography Apparatus C7 ^ 5 Fig. 2b Charcoal Trap System Fig. 2c U-Tube Trap -9- While one was being used as a purifier, the other was regenerated by- evacuating and heating to about 300° C. Separation of the products Previous experience i n this laboratory (9) had shown that two columns were necessary to separate the components of a mixture of pro- ducts with such widely varying boiling points as that obtained from the decomposition of 1-butene. The light products, i . e . methane, ethane, etc. , up to 1-butene had been successfully resolved on an alumina column ( i . e . an adsorption-elution column). Products of higher boiling points than 1-butene were separated on a T r i C*esyl Phosphate (TCP) column ( i . e . a partition elution column). Both columns had been operated with r is ing column temperature to decrease the elution times. In the present work, however, i t appeared desirable, with a view to quantitative measurements, to use one column only. While the TCP column had been found satisfactory to separate the products of boiling point higher than 1-butene (9), the lighter products were not satis- factorily resolved on this column. An attempt was therefore made to separate a l l the products of the butene decomposition on a TCP column 12 feet long ( i .e . twice the length of the column previously used for this kind of work i n this laboratory). Fire brick was used as a support for the TCP to give a lower flow resistance (6) than that obtained with Celite . The separation of the products was better on this column of double length. A l l the light products were i n fact separated, except ethane and ethylene. The column was lengthened further to 18 feet ( i .e . three -10- times the original length) but even this was not sufficient to separate ethane and ethylene. In the end i t was found that a 6 foot alumina column was sufficient to separate a l l the products of the pyrolysis of 1,5-hexadiene i f i t was gradually heated to high enough temperature. It was found important to purge the column from high boiling compounds previous to each run. A standard purging method was to heat the column to 210° C for about 30 minutes with the carrier gas flowing. The separation of the products of the hexadiene pyrolysis are shown i n the chromatogram (f ig. 3)« The light products are clearly separated. Although two pairs of the products following 1-butene are not completely resolved, the separation was sufficient for identification of the products and for their quantitative estimate. Product Identification A l l the light products were identified by comparison of present analysis charts with those previously obtained in this laboratory. An indication of the identity of the higher products was obtained by injection of a pure sample of a compound suspected to be present, together with the reaction pro- ducts and thus obtajjiing re-inforcement of a particular peak. Re-inforcement of a particular peak i s , however, merely a necessary but not sufficient con- dition to prove the presence of a compound. Positive identification was effected with a mass spectrometer. The method adopted was to trap a particular peak. The mixture of the products was admitted onto the column as usual. The stopcock in f i g . 2a was turned so that the gas stream was diverted from the flow meter directly into the atmosphere. A U-tube with two stopcocks (f ig . 2c) was cooled i n 0 O 5 J vj 10 12 14 min. 22 80 125 150 170 200"C Fig. 3 Gas Chromatographic Separation of the Pyrolysis Products -12- l iquid nitrogen and kept ready to be connected to the outlet from the thermal conductivity ce l l * As soon as the peak of the compound to be trapped appeared on the chart, the U-tube was attached to the outlet and the gas stream allowed to pass through. The trapped sample of the com- pound concerned was then analyzed with a mass-spectrometer as described i n the Appendix. Quantitative determination of the products It has been shown that the peak areas in gas chromatograms are approximately proportional to the amount of the substance put through the column. (6) This relationship was made use of i n the present quantitative determinations of the pyrolysis products, although some modifications had to be made because the column was not operated at constant temperature. Also since internal standards for a l l the reaction products were not avail- able, approximations had to be used in some cases to determine the products quant itat ively. The procedure used was as follows: the pure compounds which were available were mixed i n the following proportions: Compound: Percentage: Benzene 11.9 1,5-hexadiene 13.9 cyclopentene 6.28 Butene-1 8.28 Propylene 48.0 Ethane 12.6 The above mixture was run through the gas chromatographic column -13- under exactly the same conditions as the mixture of reaction products and the peak areas thus obtained were related to the amount of the particular compound. The compounds for which no internal standards were available had to be determined quantitatively by an approximation. RaisiHg; the temperature causes the flow-rate of the carrier gas to decrease which i n turn influences the peak areas. Therefore the variation of flow-rate of the carrier gas with column temperature (or time, since the temperature rise as a function of time was well reproducible) was determined. The thermal conductivity of the compound to be determined was assumed to be equal to the thermal conductivity of the internal standard with the closest retention volume. Assuming that the peak areas were inversely proportional to flow rate of the carrier gas (6), correction was then made for the variation of the flow-rate between the peak of the internal standard and the peak of the compound to be determined. The peak areas were measured with a plsjiimeter. In general two correction factors were applied in the quantitative determinations: correction for the variation in the ce l l current and in the ce l l temperature. It was found that a ce l l current of around 200 ma. •ef the peak area increased by approximately 1.7 # per ma. current increase. The peak area decreased by approximately 0.6 io when the ce l l temperature was increased by 1 ° C. Qualitative and quantitative determination of hydrogen Although no hydrogen peak was obtained when analyzing for the products of the pyrolysis of 1,5-hexadiene using alumina column and He as carrier gas i t was suspected to be present because of the presence of certain other -14- compounds. An activated charcoal column was used to separate hydrogen from the other products. Injection of a mixture of hydrogen and methane indicated that these two gases are easily separated by the charcoal column at room temperature. Analyzing for hydrogen with helium as carrier gas, the sensitivity of the apparatus is greatly reduced as the thermal con- ductivities of hydrogen and helium differ by a factor of 2 only. Hence nitrogen was employed instead of helium. The analytical method was further slightly modified; because of the low heat conductivity of nitrogen (as compared to helium), less heat i s removed from the detection filaments in the ce l l than when helium is used as carrier gas. The temperature of the filaments consequently increases. Since the sensitivity of the detector increases with temperature of the filaments, stabil ity of the base l ine produced by the recorder was affected. The ce l l current had to be decreased from 200 ma. to 160 ma. The latter value was found sufficiently low to maintain the base-line stabi l i ty. Analytical Results The composition of the reaction mixture from the pyrolysis of 1,5- hexadiene for 5 minutes at various temperatures is given i n Table I. A l l i n i t i a l pressures (P. ..) were 70 mm. except for 506e C where P. . , was 100 mm. l i n t n u t The figures show percentage of sample withdrawn from the reaction vessel. -15- Table I Products of 1,5-Hexadiene Pyrolysis (Mole jo) at Various Temperatures 5 minutes, P. . , — 70 mm ' in i t « G. Products 460 470.5 480 490 501 506 510 521 Hydrogen 1.0 2.0 3-3 4.5 6.3 6.5 7.0 7.8 methane 1.4 1.5 2.6 4.6 4.1 6.3 5.4 7.7 ethane 0.9 0.8 1.5 2.2 2.0 3.4 2.3 2.7 ethylene 4.7 4.2 6.7 9.2 9.2 9.2 8.4 10.3 propane 0.11 0.17 0.28 0.8 0.6 1.3 1.1 1.3 propylene 18.2 16.6 26.8 35.3 35.6 36.7 31.0 35.3 1-butene 3-2 2.9 4.4 5.7 5.3 5-3 4.5 5.1 cyclopentene 1.7 1.3 1.3 2.3 0.9 0.8 0.85 1.0 cyclopentadiene 3-2 2.6 4.5 5.4 3.9 3.1 2.6 2.5 1-hexene 1.7 . 1«5 2.4 2.7 2.2 1.6 1.9 2.2 cyclohexene & 1.4 1.4 1.6 2.1 0.7 1.0 0.6 0.7 methylcyclopentene 1,5-hexadiene 43.5 36.2 16.8 11.2 5.3 4.2 3-2 3.2 benzene 2.4 2.9 4.6 7.7 6.4 8.7 6.9 8.8 Total mole i» 73-4 74.1 76.8 93.7 82.5 89.1 75.8 88.6 From the above table i t can be seen that the analysis has accounted for approximately 70 - 90 % of the reaction products. There may be several reasons for this: (l) Uncertainty in the quantitative gas chroma- tographic determinations (discussed in a later chapter). (2) The sensi- t iv i t ies for those products for which no internal standards were available -16- (approximately one half of the products) had to be estimated from the values for similar compounds. (3) Some loss might possibly also be accounted for by assuming that some of the heavier products are so strongly adsorbed to the column that desorption does not take place. From Table I i t can be seen—at least qualitatively—that the "total mole$", i . e . products accounted for, increases with increased temperature. It has been mentioned earlier that the characteristics of the pressure- change vs. time curves for different temperatures ( f ig . 4) indicate that more of the heavier products are formed at low than at high temperatures. If this i s the case the loss should be higher at lower temperatures than at higher temperatures. Although some doubt exists regarding the absolute values of a few of the concentrations given in Table I, values for the compounds for which internal standards were available should be reasonably accurate. For the other compounds the relative concentrations at the different tempera- tures should be reliable and are of some considerable interest. On the average about 85$ of the pyrolysis products have been accounted for. This should be sufficient to be able to make reasonable assumptions about the principal features of the mechanism of 1,5-hexadiene decomposition. R E S U L T S -17- R E S U L T S I. THE PYROLYSIS OF 1.5-HEXADIENE Order of the reaction The differential method was applied to determine the order of the pyrolysis of 1,5-Hexadiene. The reaction was f i r s t order but rose to higher values as extensive decomposition occurred. Pressure increase in reaction system In f i g . 4 the relative pressure change in the reaction vessel as a function of time is given for temperatures between 460° and 5219 C. The pressure dependence curves show a characteristic shape. At lower temper- atures the curvature is convex to the time coordinate while at higher temperatures the curvature is concave to the time coordinate. This behaviour should indicate that more products of lower molecular weight than that of 1,5-hexadiene are formed at higher temperatures while relatively more products with higher molecular weight than that of hexadiene are formed at lower temperatures. The pressure change and the distribution of the products w i l l be discussed later. An attempt was made to make an estimation of the overall activation energy from the i n i t i a l pressure change in the above mentioned temperature range. The relative pressure change after both l/2 minute, 1 minute and 2 minutes of reaction was attempted, used for an Arrhenius plot but un- successfully. Rate and overall energy of activation for the pyrolysis of 1.5-hexadiene As was mentioned above, the kinetic order of the hexadiene pyrolysis Fig. 4 Pressure Change vs. Time for 1,5-Hexadiene -19- i s approximately f i r s t . F irst order rate constants were therefore calculated for reaction time 5 minutes at different temperatures and are given in table II. F ig . 5 i s the Arrhenius plot for the pyrolysis of 1,5-hexadiene. The activation energy obtained is 31'3 kcal/mole and the frequency factor 8.5 —1 found to be 10 * min . As a check on the constancy of the rate constants with time several experiments were performed, the results of which are l i s ted i n table III. Using the values given i n the table an activation energy of 32.6 kcal/mole was obtained for the decomposition of 1,5-hexadiene. This is approximately the same value as the one obtained using the rate constants for reaction time of 5 minutes. Since the values at 1 and 3 minutes reaction time are considered more unreliable, both because of the high uncertainty i n the reaction time and i n the determination of the percentage of decomposition, only the values obtained for the reaction time of 5 minutes were used for the Arrhenius plot. Rates of formation and overall activation energies for the ligfot hydrocarbons In order to calculate the kinetic rate constants for the formation of the light products from methane to 1-butene i t was assumed that their formation is dependent upon the f i r s t power of (hexadiene). It has been previously shown (9) that the above dependence exists i n the pyrolysis of 1-butene. As the products of the pyrolysis of 1,5-hexadiene are similar to those obtained i n the pyrolysis of 1-butene, i t was supposed that the mechanisms are very similar and the above assumption should be just i f ied. The mathematical expression for the rate constants for the formation -20- Table II Rate Constants for the Over-all Decomposition of 1,5-Hexadiene Reaction time 5 minutes. P. . . = 70 mm. Temp. l /T fo decomp. k (min"*'') -In k »C. «K. x 10~5 p f 460 733 1.365 50.5 0.705 0.141 1.960 470.5 743-5 1.345 61.7 0.960 0.192 1.73 480 753 1.328 69.5 1.19 0.238 1.44 490 763 1.311 79.0 1.56 0.312 1.17 501 774 1.291 89.6 2.27 0.454 0.790 510 783 1.278 92.6 2.60 0.520 0.653 Table III Rate Constants for the Over-all Decomposition of 1,5-Hexadiene for Different Reaction Times Temp. 9C. 1 min. k (min - 1) 3 min. 5 min. k - l o g 1 0 l 510 0.539 0.543 0.520 0.534 +0.273 501 0.554 0.545 0.454 0.517 +0.287 490 0.307 0.275 0.312 0.298 +0.526 -21- 5 Pig. 5 -22- of the products was derived as follows: p = (product) k = rate const, for formation of p. P P p^ = (hexadiene) k^ = " " " decomposition of hexadiene Pho= ^ h e x a d i e n e ) i n i t i a l d p p = k p h where p h = p k o e ' V dt dt P ^ G P J P dP~ = I K P h o e _ k h -k. t e T i dt P p = Jj> P h 0 (1 " e"V) * h No approximation can be used since k^t i s neither small nor large. Prom the above expression the rate constant for the formation of a product can be obtained: P, \o 1 - e - V By means of the above expression the rate constants were calculated and are given in Table 17. Pigs. 6 and 7 are the Arrhenius plots for the formation of the saturated and unsaturated light products respectively. Table V gives the energies of activation and the frequency factors obtained from the above graphs. -25- Table IV Rate Constants (min - 1) for the Formation of the Light Products 460" 480<» 4900 5010 5100 5210 Methane 4.24 * x 10"° 10.6 x 10~5 19.7 , x 10"° 27.0 x 10" 44.0 0 x 10" , 65.9 . 0 x 10~ Ethane 3.05 6.58 9.28 15.0 18.5 22.7 Propane 0.336 1.17 3-38 4.0 9.15 11.4 Ethylene 14.5 27.7 39-5 60.8 68.1 87.8 Propylene 55.6 111.2 150.0 235.0 257.0 501 1-Butene 9.65 18.1 24.1 34.8 56.6 45.2 Table V Energies of Activation and Frequency Factors for the Formation of the Light Products Product E (kcal/mole) £1 A (min""') Methane 50.1 i o 1 2 ' 6 Ethane 59.1 I O 9 ' 1 Propane 65.4 I O 1 6 ' 1 Ethylene 52.8 I O 8 ' 9 Propylene 56.0 i o 8 ' 5 1-Butene 55.9 i o 8 * 0 -24- -2.5001 1.250 Fig. 7 1.300 1.350 x 10"3 Arrhenius Plots for the light unsaturated Products in the Pyrolysis of 1,5-Hexadiene l/T -26- Temperature and time dependence in the formation of products Table I [from quantitative determination of the products] was used to construct the graphs in figures 8, 9a, and 9b giving the concentrations of the different products at different temperatures after a reaction time of 5 minutes. It i s seen that the concentrations of the saturated pro- ducts increases with temperature while the concentrations of the unsaturated products increase to a certain point and then gradually decrease. This fact can be explained by decomposition (presumably attack by free radicals present i n the system) of the unsaturated products formed. More attention wi l l be paid to this fact i n a later chapter. Similarly the dependence of the concentration of the products at 501° C on the reaction time was plotted in figa. 10 andv11 using the values given in table VI. Also i n this case there is an indication that the unsaturated products, both acyclic and cycl ic , undergo decomposition i n later stages of the reaction. -27- 40 - Mole fo 30 20 10 O Methane % Propylene A Ethane A 1,5-Hexadiene • Propane • Ethylene V 1-Butene 460 470 480 490 500 510 520 C Temperature Fig.8 Light Products of the Pyrolysis of 1,5-Hexadiene vs. Temperature Reaction Time 5 Minutes -28- Mole <fo 10 A Benzene O Hydrogen • Cyclohexene & Methylcyclopentene -1 1 L 460 470 480 490 500 510 520 C Pig. 9a Temperature Mole fc 10 O Cyclopentene # Cyclopentadiene _ A 1-Hexene 460 470 480 490 500 510 520 C Fig. 9b Temperature Fig. 9a and Fig. 9b Hydrogen and the heavy Products of the Pyrolysis of 1,5-Hexadiene vs. Temperature Reaction Time 5 Minutes -29- -30- -31- Table VI Decomposition products of 1,5-Hexadiene pyrolysis as a function of time at 501° C. Product 1 minute 3 minutes 5 minutes 10 minutes Methane 1.2 $> 3-2 1» 4.1 % 6.0 i» Ethane 0.40 1.1 2.0 2.5 Ethylene 4.2 7.3 9.2 9.2 Propane 0.0 0.26 0.60 1.0 Propylene 15.5 28.9 35.6 33-0 1-Butene 2.5 4-4 5.3 4.9 Syclopentene 2.0 1.6 0.87 0.75 Cyclopentadiene 2.1 3-1 3-9 2.6 1-Hexene 1.4 2.1 2.2 1.8 Cyclohexene & methylcyclopentene 0.93 1.1 0.68 0.76 1,5-Hexadiene 35.9 10.7 5.3 3-2 Benzene 1.4 5.3 6.4 6.9 -32- II. PYROLYSIS OF 1-BUTENE IN THE PRESENCE OF DIALLYL Introduction It has been shown (4) that the primary split i n the pyrolysis of 1-butene in the temperature range round 500° C occurs i n the following way: CH2 = CH - CH2 - CH 5 > CH 2 = CH - CHg. + ^CHj [l] It has also been shown that the addition of methyl radicals increases the rate of pyrolysis of 1-butene, increasing the amount of the light products. From that i t was. deduced that methyl radicals play an important part even in the unsensitized thermal decomposition of 1-butene and that i t s pyrolysis most probably proceeds by free radical chain reactions. In the present work the fate of the a l l y l radical in the pyrolysis of 1-butene was taken up. The concentration of a l l y l radicals was increased by addition of 1,5-hexadiene, which on decomposition yields a l ly l s i n the primary step. The condition that a sensitizer should decompose much more readily than the substrate i s satisfied in the case of 1,5-hexadiene, the difference of activation energies of 1-butene and 1,5-hexadiene being almost 35 kcal/mole. The product distribution for varying amounts of the sensitizer was determined by gas chromatography. Results for the decomposition of 1-butene sensitized by d ia l ly l A preliminary run was made by decomposing 1-butene by i t se l f at 5069 c and the pressure change in the system studied. Then approximately 5 i<> by volume of 1,5-hexadiene was mixed with 1-butene and pyrolysis was carried out under the same conditions (total pressure: 200 mm). F ig . 12 gives the pressure increase with time i n both cases. The slopes of - 33 - AP (mm) 30 slope = 3.0 mm/min Pig. 12 Pressure Change vs. Time at 506°C. I n i t i a l Pressure 200 mm -34- tangents drawn to these curves should, at least i n the early stages of the reaction, give an indication of the relative rates. Thus the ratio of the rates "sensitized to unsensitized" at the reaction time of 1 minute was found to be 6.5. This indicates that increase i n concentration of a l l y l does increase the decomposition of 1-butene. A series of pyrolyses with different concentrations of the sensitizer was carried out. The purpose of this was to relate the formation of the products to the concentration of a l l y l , which i n turn should be proportional l /2 to (hexadiene) ' i f the primary spl i t i n the 1,5-hexadiene pyrolysis i s symmetrical. Table VII gives the concentrations of the products for the different concentrations of the sensitizer. In f i g . 13 the dependence of the formation of methane, cyclopentene, cyclopentadiene, cyclohexene and methylcyclopentene on the concentration of d i a l ly l i s given for reaction time of 5 min. It i s seen that the rates of formation of the above products are roughly proportional to (d i a l ly l ) 3 7 ' 2 . - 3 5 - Table VII Composition of the Reaction Mixture for Sensitized "Decomposition of 1-Butene with 1.5-Hexadiene 5 min. at 506a <j. Product 0 i» 3-1 $ 5.5 i» 6.7 io 9.2 i> 11.4 Methane 1.9 4.4 4.4 5.0 5.5 5.7 Ethane 0.7 2.0 2.4 2.4 2.6 2.8 Ethylene 1.2 2.8 5.5 4.2 4.2 5.2 Propylene 2.9 6.3 9.0 9.0 9.9 11.5 1-Butene 52.2 51.4 55-6 51.1 52.6 29.8 Cyclopentene 0.7 0.9 1.5 1.1 1.4 1.2 Cyclopentadiene 1.0 1.8 2.0 2.0 2.2 2.4 1-Hexene 0.0 0.5 0.7 0.5 0.9 1.0 Cyclohexene & 0.0 0.5 0.4 0.4 0.5 0.6 Me-cyclopentene Benzene 0.0 0.6 0.9 0.7 0.9 1.5 PRODUCTS OF SENSITIZED DECOMPOSITION OF 1-BUTENE VS. {% DIALLYL) Fig . 13 -37- III. PYROLYSIS OF n-BUTANE IN THE PRESENCE OF 1.5-HEXADIENE In order to compare the reactivity of a l l y l with an olefin and a paraffin hydrocarbon, the sensitized decomposition of n-butane was studied in one experiment. n-Butane was decomposed by i t se l f and i n presence of appr. 5 i° by volume 1,5-hexadiene at 506° C. The products of the pyrolysis of n-butane had previously been determined (2l). The pyrolysis of n-butane by i t se l f at 506" C gave methane, ethane, ethylene and propylene as the principal products for a reaction time of 5 minutes (Hg was not analyzed for) . The sensitization of the pyrolysis of n-butane with 1,5-hexadiene increases the amounts of products present when n-butane i s decomposed by i t se l f and in addition 1-butene and one compound with higher molecular weight than 1-butene are formed. The retention volume of the above compound appeared to be equal to the retention volume of 1-pentene and i t i s most probable that the compound concerned actually i s 1-pentene. This may be expected since both a l l y l and ethyl radicals are present i n the system and upon combination would yield 1-pentene. Table VIII gives the products of the pyrolysis of n-butane and sensitized pyrolysis of n-butane by 1,5-hexadiene. From the ratio sensitized/non-sensitized i t can be seen that the formation of unsaturated products is favoured by the increase of con- centration of a l l y l . -58- Table VIII by Addition of Appr. 5 $ by Volume of 1,5-Hexadiene at 506° C. Reaction Time 5 Minutes. P. . . = 200 m i t mm. Non-sensitized * Sensitized Sensitized Non-sensitized Methane 2.0 4.9 2.5 Ethane 0.9 2.0 2.2 Ethylene 0.9 3-4 3-8 Propylene ? * ? * 5-5 1-Butene negligible 1.7 1-Pentene 0.85 Benzene negligible * Propylene and n-butane are not completely resolved on the alumina column so that the absolute values could not be obtained. Since the propylene peaks were extremely sharp, the ratio of the peak heights was taken as a measure of the re1. amount of propylene in the two runs. - 3 9 - IV. PYROLYSIS OF 1-BUTENE IN THE PRESENCE OF ACETALDEHYDE It has been shown (4) that methyl radicals play an important role in thermal decomposition of 1-butene and that methyl radicals generated by pyrolysis of mercury dimethyl sensitize the decomposition of 1-butene. In order to show the effect of methyl radicals obtained from a source other than mercury dimethyl on 1-butene, and in order to get acquainted with the general method of pyrolysis of gases, pyrolysis of 1-butene i n presence of acetaldehyde was carried out. It i s considered as well established that the thermal decomposition of acetaldehyde is a free radical chain process (11,17)• Methyl radicals are formed i n this reaction and have been detected by several workers ( l l ,17) . It would be therefore expected that reactions between these and 1-butene would take place in analogy with the work mentioned above. From a pressure vs. temperature plot i t was found that 1-butene starts to decompose at about 480° C At 477° C there i s essentially no decomposition of 1-butene while acetaldehyde was found to decompose appreciably at 440 - 450° C. The temperature of 477° C was therefore chosen as a suitable temperature and two series of experiments were performed: I. acetaldehyde and helium i n molar ratio 1 :3 II. acetaldehyde and 1-butene i n molar ratio 1 :3* The total i n i t i a l pressure was 200 mm. in both cases. The degree of decomposition of acetaldehyde was determined by taking the total amount of methane and carbon monoxide (20) as a measure of the decomposition. Percentage of 1-butene decomposed was estimated by gas chromatography from the amount of 1-butene which remained in the reaction -40- vessel after each reaction was completed. The analytical results of the above pyrolyses are given in Table IX. Table IX Percentage of Decomposition of Acetaldehyde and 1-Butene at 477° C. 1 minute 3 minutes 5 minutes S e r i e s I Acetaldehyde 8.2 11.8 - 24 S e r i e s I I Acetaldehyde 3.2 4.8 7.9 1-Butene 8 11 19 The above experimental results are plotted in f i g . 14. It can be seen that while the i n i t i a l rate of decomposition of acetaldehyde i n presence of helium is 6.0 $ min"*''", i t i s only 2.8 $ min 1 when 1-butene i s present. Thus acetaldehyde decomposes more than twice as fast i n the former case. Suppression of the decomposition of acetaldehyde in later stages of the reaction i n presence of 1-butene seems to be s t i l l more pro- nounced. 1-Butene, which practically does not decompose at the temperature of the experiments by i t se l f , decomposes quite appreciably when acetaldehyde i s present. Since the thermal decomposition of acetaldehyde i s known to be a free radical chain process, the decrease in rate of i t s decomposition in presence of 1-butene must be therefore caused by decrease i n the steady state con- centration of the free radicals which are the chain carriers. Thus -41- Decomp. (*) 30 20 L 10 slope = min~* Acetaldehyde(I) A ^ _ . "~ "l-Butene(ll) slope = 2.8#tnin* cetaldehyde (II) 5 min. Time Fig. 14 Extent of Decomposition at 477 C. I n i t i a l Pressure 200 mm I. Acetaldehyde + Helium ( l : 3) I I . Acetaldehyde + 1-Butene ( l : 3) -42- reactions of the free radicals formed must have taken place with 1-butene. Also, i f 1-butene was inert to attack by methyl or/and other radicals, more decomposition of acetaldehyde would rather be expected in the presence of 1-butene than of helium. Molecules of 1-butene (being poly- atomic molecules with a high number of degrees of freedom) would be expected to enable a far better energy transfer during the pyrolysis. The nature of the products formed when 1-butene is attacked by the above free radicals was not determined in these preliminary experiments. D I S C U S S I O N -43- D I S C I S S I O N PRECISION OF TBE QUANTITATIVE DETERMINATIONS BY GAS CHROMATOGRAPHY There are several factors influencing the reproducibility of quanti- tative gas chromatographic determinations (6). The main ones have already been mentioned when the quantitative determination of the products was described: flow-rate of the carrier gas, column temperature (one of the factors determining the flow-rate), thermal conductivity,, ce l l current and ce l l temperature. In addition to these variations in room temperature could produce variation in the amount of the gaseous sample admitted into the by-pass of the admission system. There is also uncertainty i n deter- mining the peak area with a planimeter. The following are the estimated or assumed maxima of uncertainties in the above parameters: 1) uncertainty in flow-rate (including the column temperature) - 3 fo 2) uncertainty in ce l l current (after correction was applied) - 0.8 fo 3) uncertainty in ce l l temperature (after correction was applied) negligible 4) uncertainty in room temperature negligible 5) uncertainty in planimeter measurements - 2 $ - 5.8 <?o~6 <fo In addition to the above uncertainties there might be a number of factors leading to systematic errors which might have increased the maximum of the total error. -44- UNCERTAINTY IN THE VALUES OF THE KINETIC HATE CONSTANTS The mathematical expression for the f i r s t order rate constant i s : reactant, p pressure of reactant at time t . If a i s the degree ofifccom- The kinetic constant i s i n this case dependent upon two variables the uncertainty in which wi l l determine the total uncertainty of the value of the rate constant. The maximum uncertainty in the degree of decomposition, a, i s , in the best case, equal to the uncertainty i n the quantitative gas chromato- graphic determinations. In cases where l i t t l e decomposition took place the peak areas being subtracted from each other are of the same order of magnitude and consequently their difference wi l l be of a higher degree of uncertainty than the peak areas themselves. Since normally two to four parallel analysis-runs were made and the results averaged, i t should be reasonable to take the uncertainty in a equal to the uncertainty i n the peak areas, or - 6 °?o. The reaction time was 5 minutes i n a l l the runs the data from which were used to calculate the rate constants. The diff iculty i n estimating the starting- and end-point of the reaction was found to lead to a maximum uncertainty in the reaction, time of i 5 The variation in the rate constant with t and a i s given by: [2] where t is. the reaction time (min in our case), pQ i n i t i a l pressure of -45- and the maximum relative uncertainty by: lAti . 1 ^ 1-a l*vl + - ^ " K l n + T |Ao max JAklmax = - _t 1 k 1 , 1 t  l n T ^ = + (1- ) 1*! W + '""'max [4] Putting: a = -|- I A 1 1 6 | A a | max = 2 X 100 t = 5 (min) U t l m a x = 5 x 100 " * * one gets: ( A k l ^ = + ? ^ k The above considerations were made on the assumption that there was no variation i n the furnace temperature during a run. As was mentioned i n the paragraph "Pyrolysis Apparatus", the furnace temperature (assumed to be equal to the temperature at which the reaction takes place) was kept constant to within - 1« C. Considering the variation of temperature one obtains the following: „ - a k = A e RT E - a E RT and l Ak k = E a 1 IA TI C5] ' 1 1 — — 1 1 max max R T2 -46- Putting: E a 32 000 cal/mole R 2 cal/mole-deg T 773° K 1» K i t i s obtained: Ak max Considering uncertainty i n both t , a, and T, the max. uncertainty in k i s then - 10 f>. Consequently,. most probably no more than two figures i n the rate constants obtained are significant. THE OVER-ALL ACTIVATION ENERGY FOR THE DECOMPOSITION OF 1.5-BEXADIENE It was mentioned in a previous section that the attempt to determine the over-all activation energy for the pyrolysis of 1,5-hexadiene from the i n i t i a l pressure changes at different temperatures was unsuccessful. Although this method of determining the over-all activation energy had successfully been applied in the case of 1-butene (9) the negative result i n the present work should not be surprising. A large number of products i s formed in the pyrolysis and there i s a considerable variation with temperature of the distribution with respect to their molecular weights (see Table I and Figs. 8, 9a and 9b). Consequently, the pressure change produced i n the pyrolyses at different temperatures cannot be assumed as being proportional to the extent of the reaction in this system. Furthermore, the i n i t i a l pressure of the reactant i n the reaction vessel was rather low (approx. 70 mm) so that the i n i t i a l pressure changes involved were correspondingly small and the values obtained therefore uncertain. -47- The pressure change can be used as a measure of the extent of a reaction taking place in the gas-phase only i f the distribution of the products i s more or less constant with respect to temperature and i f the pressures involved can be measured with sufficient accuracy. The determination of the over-all activation energy from the gas chromatographic analyses yielded the value 31 «3 kcal/mole. The dis- sociation energy for the central bond in 1,5-hexadiene had been estimated (22) to be around 42 kcal/mole. Taking the heat of formation of the a l l y l radical as 32 .3 kcal/mole ( 1 3 ) and the heat of formation of d i a l ly l as 20.6 kcal/mole (calculated from the heat of hydrogenation, -60.5 kcal/mole ( 3 ) , and the heat of formation of n^hexane, -39.96 kcal/mole (8)) i t can be calculated that the heat of the reaction: CH 2 = CH - CH 2 - CH 2 - CH = CH 2 —> 2 CHg = CH-CHg. [6] i s 44 kcal/mole. The value obtained for the over-all activation energy is considerably lower than both of the above values. This fact indicates that reactions other than reaction [6] must take place, such as hydrogen abstraction by a l l y l radical and a l l y l addition to double bonds in the system. Although the activation energies for these two processes are not known, i t i s certain that they are far below the bond dissociation energy for the central bond in d i a l l y l . Thus the over-all value of 31*3 kcal/mole should not be surprising. Conversely, the fact that the over-all activation energy found i s con- siderably lower than the heat of reaction [6] could be used as an evidence that free radical processes take place. It can be shown (7, p. 232) that for a mechanism involving a single -48- chain carrier with second-order termination, as follows init iat ion =» R + propagation R + . . . . » R + termination R + R 5» i f the activation energy for termination i s non-zero or i f chain termination is second order (as above) the apparent (over-all) activation energy may be significantly less than the activation energy for the ini t ia t ion step. MECHANISM OF THE PYROLYSIS OF 1.5-HEXADIENE General Three possibilities exist for the mechanism of pyrolysis of 1,5- hexadiene: (a) molecular mechanism (b) free radical mechanism (c) combined molecular and free radical mechanism. It was seen that a large number of both aliphatic and cyclic products i s formed in the pyrolysis. Since such a large spectrum of products could not be expected i f the mechanism was purely molecular, possibility (a) should thus be ruled out. It has been shown (l2) that at high temperatures (690 - 890° C) 1,5-hexadiene decomposes by splitting the central bond: CH2 = CH - CH2 - CH2 - CH = CH 2 > 2 CH2 = CH - CE^ [6] The results obtained in the present work also show that split of the central bond must have taken place. If the primary step in the pyrolysis at present conditions involved exclusively reaction [6], a pure free radical chain mechanism could be postulated. -49- On the other hand, cyclohexene, benzene and hydrogen have been found among the reaction products. As i t i s shown in the following chapter, the formation of cyclohexene and benzene can be explained on the basis of a free radical mechanism. But the presence of these products might equally well indicate that the following molecular rearrangement followed by de- hydrogenation also takes place to some extent: At this stage i t can therefore be concluded that possibil ity (c) i s most l ike ly . Reactions of the a l l y l radical (l) Hydrogen abstraction by a l l y l Since the decomposition of 1,5-hexadiene probably involves formation of two a l l y l radicals i n the primary step, the a l l y l radical must play an important part i n the over-all reaction mechanism. Because of the resonance energy of 17 kcal/mole (2,5) the a l l y l radical would be expected to be rather unreactive in thermal reactions. However, i t can be shown from thermochemical data (26) that the weakest bond in a l l y l (a C - H bond) has a strength of about 68 kcal. The life-time of a l l y l should thus be long enough to build up an appreciable steady state con- centration and permit various reactions involving a l l y l to take place. McNesby and Gordon have recently investigated the thermal decomposition CH, •CH, -50- of cyclopentane-dg and cyclopentane-acetone-dg mixtures (14) and have shown that reactions of a l l y l radicals are involved. According to the authors appreciable H-abstraction, both from cyclopentane and from CDjCOCD^ takes place at 453° and that the degree of H-abstraction i s s t i l l higher at 5000 c. The main product of the pyrolysis of 1,5-hexadiene i s propylene. This suggests that H-abstraction by a l l y l takes place readily in the present system. It would be expected to take place mainly by the reaction: CH2=CH-CH2. + CH2=CH-CH2-CH2-Cft=CH2 >-CH2=CH-CH5 + CH2=sCH-CH-CH2-CH=CH2 since a resonance stabilized radical i s formed. It has the following two contributing forms: CH2=CH-CH-CH2-CH=CH2 * > .CH2-CH=CH-CH2-CH=CH2 Hydrogen abstraction by a l l y l would probably also take place from the pro- ducts of the decomposition which, upon hydrogen abstraction, would yield resonance stabilized radicals. The following products would be expected to undergo such reactions: 1-butene cyclopentene cyclopentadiene 1-hexene cyclohexene methyl-cyclopentene Abstraction from propylene would also take place, but i t would only produce back propylene and a l l y l . The side reactions mentioned would be negligible i n the early stages of the decomposition, but may be quite extensive in the later stages when -51- the concentrations of the products concerned are appreciable. Evidence for attack of the above products by a l l y l (and possibly other) radicals i s indicated in figs. 10 and 11 where the concentration vs. time graphs for a l l the products l i s ted above show a maximum. There i s also a drop i n concentrations of these products with increased temperature at constant reaction time (see figs. 8, 9a and 9b). In the case of a l l y l radical, which in general reacts slower than the non- stabilized alkyl radicals, the extent of hydrogen abstraction would be expected to be quite temperature-dependent. Drop i n concentrations of certain products from a certain temperature on may also be a good indication of increased hydrogen abstraction by a l l y l radical. (2) Addition of a l l y l to double bond Addition of a l l y l radical to the double bond of olefins has not pre- viously been reported. It appears, however, that such an addition must have taken place i n the present system. The formation of cyclopentene, cyclopentadiene and methyl-cyclopentene i s di f f icult to explain i n any other way. In the present system there are two principal ways of addition to 1,5-hexadiene possible: (a) addition to the terminal C-atom i n d i a l l y l : \ CH, 2 CH-CH2-CH2-CH=CH2 CH CH, 2 CE CH, CH-CH2-CH2-CB>*CH2 2 .CH -52- This large radical would l ike ly be too unreactive to abstract hydrogen. It seems rather more probable that i t s fate is the following: ,CH„ CH, \ .CH- CH-CH0-CH0-C&=CH0 , 2 2 2 CH„ -RH -* Cfl2 CH-CH2-CH2-CHaCH2 CĤ .CH, CH. CH: 3H r CH + .CH2-CH2-CH=£H2 / + RH - R + H CH 2 CH-CH2. + . c H ^ j j ^ CH—CH +RH - R + RH - R 1-butene propylene (b) addition to the non-terminal double bond in d i a l l y l : CH2=CH-CH2. + CH2=CH-CH2-CH2-CHaCH2 CH2-CH2-CH=CH2 CH2-CH2-CH=CH2 ,CH CH, / C H 2 CH=—CH This radical is identical with the one obtained above. Addition of a l l y l to both double bonds of 1,5-hexadiene would yield products impossible to detect by the present method of analysis. Since the concentration of propylene is appreciable from the early stages of the reaction, addition of a l l y l radical to propylene has to be considered: CH2=CH-CH2. + CH2aCH-CH5 > CBg .CH-CHj CH= CH2 CH, 0H2 C I K H , ,CH 3 t!H=CH Addition to the non-terminal carbon atom would lead to the same products: CH2=*CH-CH2. + CH2=CH-CH5 • . C H ^ CH. CH-CH. 2 I 3 CH CH 2 .CH. same products < CHg CH-CIL^ .CH CH 2 (3) Combination of a l l y l with other radicals Apart from hydrogen abstraction and addition to the double bond the fate of the a l l y l radical may be combination with other radicals present. As wi l l be indicated in a subsequent chapter, hydrogen atoms areassumed to be present i n the system. Thus combination of an a l l y l with a hydrogen atom would yield a propylene molecule. Combination with a methyl would give a 1-butene molecule and recombination of two a l ly l s would produce the parent, 1,5-hexadiene. -54- The fate of the 1.5-hexadienyl radical Since both the a l l y l radical and unreacted 1,5-hexadiene are present i n quite high concentrations in the beginning of the pyrolysis, the abstraction of hydrogen from hexadiene by a l l y l must be quite extensive. The rate of formation of 1,5-hexadienyl radical must consequently be rather high. It seems that the hexadienyl radical formed may mainly disappear in three different ways: (l) Rearrangement: It may appear surprising that no cyclohexadiene, which would be expected to1 be the intermediate i n the dehydrogenation process, has been found among the reaction products. The reason may be that the radical of resonance stabilization. Hydrogen abstraction from cyclohexadiene would thus readily take place because of the high resonance energy of the radical formed. Abstraction of hydrogen from this radical would be far more pro- bable than hydrogen abstraction by, the radical . I 2 CH CH 2 obtained upon hydrogen abstraction from cyclohexadiene has a high degree - 5 5 - (2) Decomposition There are several ways i n which the 1,5-hexadienyl radical could be thought to decompose. It appears, however, that such decompositions would also necessarily lead to compounds l ike butadiene and allene which have not been found among the. products. The resonance energy of the radical i s probably sufficiently high that the life-time i s appreciable and the radical rather undergoes re-arrangement as mentioned above. (3) Combination with other radicals The third possibility for the disappearance of the 1,5-hexadienyl radical may be combination with other radicals to yield products which may be the ones that were suspected to suffer irreversible adsorption to the alumina column. Since the steady state concentration of the 1,5-hexadienyl radical i s presumably quite high, the total amount of such addition products might have been appreciable. Formation and reactions of other radicals Hydrogen atoms: formation: CH^H-CHg > a l l y l + H (22) R > M + H where M is an unsaturated molecule and R may be cyclopentenyl or cyclohexenyl. reactions: unsaturated product + H —? saturated or partially saturated product R + H RH CH2=CH2 + 2H > C ^ g CH2=CH-CH5 + 2H * CH 5-CH 2-CH 5 1,5-hexadiene + 2H 1-hexene -56- Methyl: formation: CH2=CH-CHj • CH2=CH. + CH^ 0"cs3 -*rv - + - c h 3 CH- CH-CH2 > f ^ l + CH, ^ / 3 L J 3 • CH—CH2 reactions: GHj + RH —> CH 4 + R CH, + CH2-—>C-H,. 3 3 2 o CHj + a l l y l > 1-butene Possibly addition to the double bonds of 1,5-hexadiene and some of the products giving compounds which have been adsorbed irreversibly to the alumina column. Vinyl : formation: CH2=CH-CH^ - CH^aCH. + .CH^ CH2=aCH-CH2-CH2. * CH2=CH. + CH^CHg reactions: CH2=CH. + RH * CH2=CH2 + R possibly also addition to the double bonds. Butenyl: formation: CH„ CH !H„ CH-CH„ CH„-CH=CH« ^ CHg JE + .CH2-CH2-CH=CH2 CH=CH CH=CH reactions: CHgaCH-CHg-CHg. + RH ^ 1-butene + R possibly CH2=CH-CH2-CH2. CH2=CH. + CH2=CH2 -57- The cyclic radicals: The cyclopentenyl and cyclohexenyl radicals probably expel H-atoms to yield unsaturated compounds. H-abstraction by the radicals also takes place to some extent to produce cyclopentene and cyclohexene respectively. Mechanism for the formation of the pyrolysis products Summarizing the discussion of a l l the above mentioned elementary pro- cesses the mechanism of the pyrolysis of 1,5-hexadiene could be represented by the following scheme: A. Molecular rearrangement dehydrogenation' B. Free radical reactions Formation of: (l) methane: .CK"5 + RH > CH 4 + R GH_=CH + RH * CH0=CH_ + R CH5 + CHj- >- CgHg CHgsCSHg + 2H 5- C 2 H 6 CH_sCH-CH„. + RH * CH0=CH-CH, + R . 2 2 e. 5 CH.=CH-CH, + 2H — * CH,-CH0-CH, 2 3 3 2 3 (2) ethylene: (3) ethane: (4) propylene: (5) propane: (6) 1-butene: eH2=CH-CH2. + .QH5 — CH2=GH-CH2-CH2. + RH •> 1-butene * 1-butene (7) cyclopentene: a l l y l + d i a l ly l -> addition radical - 58 - (8) cyclopentadiene: + R *<0' + m (9) 1-hexene: (io) cyclohexene: + H CH2=CH-CH2-CH2-CH=CH2 + 2H • CH2=CH-CH-CHg- CH 2-CH 5 . CH-CaaCH-CH2-CH=CH2 RH ( l l ) Me-cyclopentene: (12) hexadiene: (13) benzene: a l l y l + d i a l ly l » addition radical RH 2 a l l y l > hexadiene it I + H 2R CH 2 . a l l y l (14) hydrogen: H + H -> H, H + RH -?H 2 + R The postulated mechanism of the pyrolysis of 1,5-hexadiene was mainly based upon the qualitative analyses of the products and upon their concen- trations at different reaction times at 5010 C. Some of the mechanisms for the formation of some of the products are rather speculative. However, a large variety of products is formed in the pyrolysis and the system is thus very complex and di f f icul t to deal with. Nevertheless, the postulated mechanism should give an over-all picture of the reactions which may be expected i n the system. It would be desirable to isolate some of the elenehtary reactions assumed to take place in the system and obtain definite proof of their occurrence. -59- The most interesting feature of the mechanism is probably the addition of a l l y l radical to olefinic double bonds. Formation of f ive- membered cyclic products results. Since the formation of these products could not easily be explained in any other way, they should be sufficient evidence for addition reactions of the a l l y l radical with the 1,5-hexadiene. Thermal decompositions taken to a much lower conversion than was the case in the present study might have simplified the mechanism of the reaction. However, such low-conversion pyrolyses would have involved much shorter reaction times in the temperature range studied and the re- producibility with the present system would consequently have been extremely poor. The dif f iculty caused by the uncertainty in the reaction times could be overcome by decreasing the temperature of the pyrolyses and use of longer reaction times. However, in both cases there remains the problem of determining the percentage of conversion. Comparing peak areas on the chromatogram which are of the same order of magnitude leads necessarily to very serious errors and is thus not usable. It i s realized that i t would have been desirable for the sake of com- parison to have carried out pyrolyses at the different temperatures to approximately the same degree of conversion rather than use of constant reaction time. It would be possible to predict the reaction time at a certain temperature for a certain degree of conversion i f the reaction time at a different temperature for the same degree of conversion was known. The following i s the method of calculation. First order process i s assumed. -60- If a ss degree of conversion, o a 1 then k = l . In _1_ (Cx 1 [3] t 1-a t -Ba RT But we have also: k = A e If t^ is the reaction time at T^ and tg i s the reaction time at T^ for a ss constant i t i s obtained: g a It̂  _ t^ _ RT 2 = e k 5 t JBa RT o r : -Ea / l • a 1̂ _ 1 j t 2 = t x e R T l T2 [8] MECHANISM OF THE THERMAL DECOMPOSITION OF 1-BUTENE SENSITIZED BY DIALLYL As was mentioned i n a previous chapter a preliminary experiment clearly showed that addition of a small amount of d i a l ly l to 1-butene accelerates the pyrolysis of the latter . The more detailed investigation involving variation in the d i a l l y l concentration shows that formation of a l l the pro- ducts of 1-butene pyrolysis is accelerated. There are two reactions possible between the a l l y l radical and 1-butene: (l) Addition to the double bond: CH„=CH-CH« + CH„-CH-CH„-CH, CH, -CH-CH--CH-. \ //E* N C H cyclopentene cyclopentadiene ethane ethylene hydrogen CH 2 CH, CH-CH2-CH5 CH„ CH -61- or: CH2=CH-CH2. + CH2=aCH-CH2-CH5 • .CH 2 CH-CH"2-CH5 CH CH ^ / 2 CH cyclopentene cyclopentadiene ethane ethylene hydrogen C H 0 CH-CH--CH, I I 5 C H 5 ca 9 ^ / 2 CH (2) H-abstraction to produce the resonance stabilized butenyl radical: CH2=aCH-CH2. + CH2=CH-CH2-CH5 -> CH2=CH-CH5 + CH0=CH-CH-CH2 2 • 3 • CH 2̂ CH—CH—CH^ The butenyl radical may also add to the double bond of 1-butene: CH -CH-CH -CH, » CH,5 GH-CH0-CH, I 2 . 2 3 j 2 j 2 3 CH 0 CH-CH, CH. CH-CH, \2 / 3 ^ / 5 CH CH * methylcyelopentene cyclopentene cyclopentadiene ethane ethylene methane hydrogen -62- or: CH- CH-CH0-CH_ CH0 CH-CH^-CH, • c. 1 d y \ d. c •) CH,-CH CH. CH^CH CH CH CH •v same products as above If the primary step in the d ia l ly l decomposition is mainly d i a l l y l > 2 a l l y l and i f re-combination of two a l l y l radicals i s also taken into consideration, the following is obtained: . ( a l l v l ) 2 = K (diallyl) or (allyl) cc ( d i a l l y l ) 1 , 7 2 This means that the rate of formation of those products which result from attack of 1-butene by the a l l y l radical should be proportional to l/2 (diallyl) . In f i g . 13 i t can beseen that the formation of methane, cyclopentadiene, cyclopentene, methylcyclopentene (and eventually but not necessarily cyclohexene) follows the above dependence within experimental error. It can therefore be concluded that the above postulated mechanism for the reactions of a l l y l with 1-butene is correct. Furthermore, this should be an evidence that the primary step i n the 1,5-hexadiene decom- position i s mainly formation of two a l l y l radicals. DECOMPOSITION OF n-BUTANE SENSITIZED BY DIALLYL Earlier work has shown that i t i s possible to sensitize the thermal -63- decomposition of n-butane by addition of small amounts of azomethane or ethylene oxide. It appears that the a l l y l radical i s reactive enough at 506° C to sensitize the pyrolysis of 1-butene. It must be borne i n mind, however, that in the case of 1-butene the hydrogen abstraction by a l l y l i s eased by the formation of a resonance stabilized radical . No stabilized radical could be expected in the case of n-butane. Prom the results obtained there does not seem to be any doubt that the a l l y l radical i s capable of sensitizing the decomposition of n-butane. The rate of formation of the saturated products (methane and ethane) increased by a factor of 2 - 2.5 and the rates of formation of the un- saturated products increased by a factor of 3*5 - 4» It was not intended to postulate any mechanism for this sensitized decomposition, but i t can at least be concluded that hydrogen abstraction takes place: CH_=CH-CH0. + CH_CH0CB0CH, » CH0=CH-CH, + CH2CH0CH CH, [9] d d d d 2 3 d 2 3 2. 3 The activation energy for the decomposition of the radical formed i s probably lower than that for the decomposition of n-butane. Therefore an over-all acceleration in the pyrolysis process results. The formation of 1-butene and 1-pentene i n the sensitized decomposition can be accounted for by: CH2=CH-CH2. + .CH 2 > CH2=CH-CH2-CH5 CH2=CH-CH2. + .CH2CH5 > CHgaCH-CHg-CHg-CHj It can be therefore concluded that the reactivity of the a l l y l radical at 506° C i s sufficient to abstract hydrogen from secondary carbon atom in n-butane. More generally i t might be concluded that the a l l y l radical is capable of abstracting hydrogen even from compounds which upon abstraction do not yield stabilized radicals. B I B L I O G R A P H Y -64- B I B L I O G B A P H Y 1. Bateman, L . , Gee, G . , Morris, A .L . and Watson, W.F., Disc. Faraday S o c , 10, 250 (1951) 2. Bolland, J . L . , and Gee, Geoffrey, Trans. Farad. S o c , 42, 244 (1946) 3. Branch, G.E.K. , and Calvin, Melvin, The Theory of Org. Chem., Prentice-Hall Inc., New York, 1947, p. 275 4. Bryce, W.A., and Kebarle, Paul, Trans. Farad. Soc (in press) 5. Coulson, C.A. , Proc. Roy. Soc., A 164. 383 6. Dimbat, M. , Porter, P .E . , Stross, F . H . , Anal. Chem., 28, 290 (1956) 7. Frost, A . A . , and Pearson, R.G., Kinetics and Mechanism, John Wiley & sons, New York, 1953, p. 231 8. Handbook of Chem. and Physics, Chem. Rubber Publishing Co., 1956, p. 1772 9. Kebarle, Paul, Thesis, Brit ish Columbia 1957 10. Kebarle, Paul, and Bryce, W.A., Can. J . Chem., 35, 576 (1957) 11. Letorfc, M. , Thesis, Paris, 1937 12. Lossing, F .P . , Ingold, K .U. , and Henderson, I.H.S., J . Chem. Phys. 22, 621 (1954) 13. McDowell, C.A. , Lossing, F .P . , Henderson, I.H.S., and Farmer, J.B. Can. J . Chem., 34, 345 (1956) 14. McNesby, J.R., and Gordon, A .S . , J . Amer. Chem. S o c , 79, 4593 (1957) 15. Molera, M.J . , and Stubbs, F . J . , J . Chem. S o c , 381 (1952) 16. Mooney, R.B., and Ludlam, E .B . , Proc Roy. S o c , Edinburgh 49, (1929), 160 17* Rice, F .O. , Johnston, W.R., and Evering, B .L . , J . Amer. Chem. S o c , 54, 3529 (1932) 18. Ryce, S.A., Kebarle, Paul, And Bryce, W.A., Anal. Chem., 29, 1386 (1957) 19. Sehon, A.K. , and Szwarc, M. , Proc. Roy. Soc. (London), A 202, 263 (1950) -65- 20* Steacie, E.W.R., Atomic and Free Radical Reactions, Reinhold Publishing Corporation, New York 1954, Vol I. , p. 206 21. Steacie, E.W.R., and Puddington, I .E . , Can.J. Research, B 16, 176 (1938) 22. Szwarc, M. , J . Chem. Phys., 17, 284 (1949) 23. Szwarc, M. , and Ghosh, B.N. , J . Chem. Phys., 17, 744 (1949) 24. Szwarc, M. , Ghosh, B.N., and Sehon, A . H . , J . Chem. Phys., 18, 1142 (1950) 25. Taylor, H.S., and Smith, J .O. , J r . , J . Chem. Phys., 8, 543 (1940) 26. Trotman-Dickenson, A . F . , Gas Kinetics, Butterworths Scientific Publications, London 1955, p. 156 A P P E N D I X r66- A P P E N D I X MASS SPECTROMETRIC IDENTIPICATIOH OF SOME OF THE PRODUCTS As mentioned previously, i t was necessary to identify some of the products formed in the pyrolysis of 1,5-Jiexadiene by means of a mass spectrometer. Definite proofs have not been obtained i n a l l cases, but very good evidence in a l l cases. The procedures used are described in each individual case. Cyclopentene; sample of pure cyclopentene was run through the mass spectro- meter and i t s spectrum compared with the unknown. Definite proof was obtained* Cyclopentadiene: pure substance was not available. From the value for the retention volume on alumina column i t could be deduced that a C^ hydrocarbon was present. The trapped sample was run at low electron energies and the mass number of the parent peak was thus obtained to be 66* The empirical formula i s thus C^H^. Since -ynic products are highly improbable, i t must be concluded that cyclopentadiene is present. 1-Hexene: pure sample was not available. Presence of impurities made i t diff icult to compare the spectrum with that of 1-hexene published by the American Petroleum Institute. Mass spectrometric analyses at low electron energies showed that the parent peak had a mass number of 84* Retention volume on alumina column indicated that a Cg hydrocarbon was present. From the cyclic compounds cyclohexane i s the only one which might be con- sidered. But i t s retention volume was different from the retention volume - 6 7 - of the unknown. Thus only a straight-chain hydrocarbon can be considered. Conclusion: 1-hexene. Cyclohexene and methyl-cyclopentene: these two products were not resolved on the alumina column. Mass spectrometric analyses carried out at low electron energies showed that there was only one parent peak and that was of mass number 82. Retention volume indicated a Cg hydrocarbon. Thus CgH^Q. Comparison with the spectrum of pure cyclohexene showed that cyclohexene might be present but that also some other hydrocarbon CgR^ was present. Methyl-cyclopentene seemed to be the most probable.

Cite

Citation Scheme:

    

Usage Statistics

Country Views Downloads
China 22 27
United States 5 0
India 3 0
Japan 2 0
France 2 0
Norway 1 0
City Views Downloads
Beijing 22 0
Unknown 6 0
Sunnyvale 2 0
Tokyo 2 0
Mountain View 1 0
Ullensaker 1 0
Ashburn 1 0

{[{ mDataHeader[type] }]} {[{ month[type] }]} {[{ tData[type] }]}

Share

Share to:

Comment

Related Items