Chemical Reactions in Electrical Discharges

20 Competition of Ethylene and Propane for

Downloaded by UNIV OF CALIFORNIA SAN DIEGO on July 23, 2016 | http://pubs.acs.org Publication Date: June 1, 1969 | doi: 10.1021/ba-1969-0080.ch020

"Active" Nitrogen P. T E R E N C E H I N D E and N O R M A N N . L I C H T I N Boston University, Boston, Mass., 02215

Ethylene and propane reacted individually and together with glow discharge-generated "active" nitrogen at autoge nous temperature over a wide range of reactant ratios. Consumption data do not correspond to simple competition of ethylene and propane for N( S) but can be explained by assuming additional attack on each substrate of reactive species generated in the primary reaction. Experiments with mixtures of C-labeled ethylene and ordinary propane show that, with N( S) in molar excess, ethylene is generated at least in part from propane. It is concluded that the spe cific rate of primary attack of N( S) on ethylene is more than five times the value for propane. The value of this rate ratio estimated from literature data for the individual compounds is of the order of ten. 4

14

4

4

^ p h e chemical reactions of "active" nitrogen, the luminous, reactive gas produced when molecular nitrogen is passed through an electric dis charge at low pressures, have been studied for several decades and the results of these studies have been summarized in a number of reviews, the most recent of which is that of Brocklehurst and Jennings (2). The principal chemically reactive species is widely believed to be Ν ( S) but mechanistic elucidation of the reactions of organic substrates has been frustrated by deep seated molecular disruption and wide diversity of products. For several years work in this laboratory has been devoted to the gas phase reactions of "active" nitrogen with propylene (8, 11,' 12) and with conjugated dienes (3, 4, 13). Experiments involving propylene have employed the three C-labeled isomers to identify (11, 12) the molecular origins of six products, H C N , C H , C H , C H , C H , and C H C N and to explore (8) the resynthesis and molecular rearrangement of propylene under reaction conditions. Experiments employing conjuA

4

14

2

2

2

4

2

6

3

3

250 Blaustein; Chemical Reactions in Electrical Discharges Advances in Chemistry; American Chemical Society: Washington, DC, 1969.

8

20.

HiNDE AND LiCHTiN

Table I.

251

"Active" Nitrogen

Yields of Monomeric" Products from 1,3 -Butadiene as a Function of Reactant Concentration 6

10 (N ),M 10 (N) ,M (C H ) /(N) Linear Flow Rate, m.sec." 4

1.6 8.1 6.7 7.5

2

7

o

4

6

0

0

1

0.5 27. 5. 2.3

1.6 8. 5. 7.5

1.5 8.1 1.0 5.7

1.6 8. 0.8 7.5

1.5 8.1 0.25 5.7


Relative Yields, % Pyrrole cis-Crotononitrile frans-Crotononitrile cis-1 -Cyanobutadiene-1,3 trans-1 -Cyanobutadiene-1,3 3-Cyanobutene cis-Cr.CN trans-C CN ri

a b

23 2 10 11 14 12 4 8

28 3 15 9 10 7 3 6

26 2 10 11 12 12 4 9

22 2 9 16 20 12 3 5

24 2 8 10 13 10 4 8

29 3 13 3 21 3 1 2

84

81

86

89

79

75

A monomeric product incorporates one complete butadiene residue in its structure. Data (3) are relative areas of gas chromatographic peaks.

gated dienes (3, 4, 13) have determined yields of as many as 17 (3) products as a function of reaction parameters and have provided data amenable to kinetic analysis. Examples of data from these prior studies are presented in Tables I—III. Speculative mechanisms have been pro posed (3, 8, 10) on the basis of the resulting information—e.g., data for the reaction of 1,3-butadiene can be rationalized by a mechanism the bare elements of which are summarized in Equations 1-7 (3). This scheme, in common with others we have proposed (8,10), assumes primary attack C H + Ν -> .C H N* species 4

G

C H N* 4

6

4

(1)

6

HCN + NH + C H (radicals) + C H (stable molecules) + p

x

q

y

pyrrole + CN + Η + etc.

(2)

C H + Η or CN - » products

(3)

C H + Ν -> HCN + NH + etc.

(4)

NH + N H - ^ N

2

(5)

NH + N - ^ N

+ H

p

p

q

q

2

+ H

2

(6)

C H + R' + R - » adducts 4

(?)

6

solely by N( S) on the organic substrate. Herron has, however, con cluded (6,7) from his mass spectrometric studies of the rate of consump tion of olefins upon reaction with "active" nitrogen that hydrogen atoms, whether formed as a consequence of primary attack by Ν ( S) or present adventitiously, compete with N( S) both with respect to primary attack 4

4

4

Blaustein; Chemical Reactions in Electrical Discharges Advances in Chemistry; American Chemical Society: Washington, DC, 1969.

252

CHEMICAL REACTIONS IN ELECTRICAL DISCHARGES

on the substrate and with respect to secondary attack on reactive species Ν + C H 2

-> H C N + C H

4

(8)

3

Ν + C H —> H C N + 2H

(9)

8

H + C H -*C H 2

4

2

H + C H

5

—» 2 C H

Ν + C H

5

-> N H + C H

2

2

(11)

3

2

N + NH-*N


(10)

5

2

(12)

4

+ H

(13)

formed in the initial attack. Equations 8-14 summarize the mechanism which Herron has proposed (6,7) for the reaction of ethylene. Table II. Position of in Propylene

Molar Activities of Products from C-Labeled Propylenes Relative to Those of Reactant Propylenes" 14

HCN

CH 2

CH

6

2

CH

h

2

2

CH

CH

CH CN

3

8

s

e

3

(C H ) /(N) =10 s

e

o

o

C-l C-2 C-3

0.29 0.32 0.39

0.77 0.15 0.95

0.66 0.78 0.34

0.47 0.86 0.39

0.90 0.82 1.00

1.02 1.01 0.98

0.45 0.99 0.53

Sum

1.00

1.87

1.78

1.72

2.72

3.01

1.97

(C H ) /(N) =1.0 3

e

o

o

C-l C-2 C-3

0.30 0.25 0.39

0.69 0.15 1.06

0.74 1.00 0.38

0.47 0.88 0.56

0.98 1.00 1.02

1.00 1.03 1.02

0.46 0.95 0.48

Sum

0.94

1.90

2.12

1.91

3.00

3.05

1.89

0.47 0.71 0.43"

1.05 0.89 0.91"

0.92 1.06' 0.96

0.48 0.85 0.41"

1.61

2.85

2.94

1.74

(C H ) /(N) 3

C-l C-2 C-3

0.33 0.26 0.32

Sum

0.91

b

6

0

0.70 0.02 0.87

0.59 0.68 0.36

1.59

1.63

b

o=

0.17 c

c

" From Ref. 12 unless otherwise indicated. "Data from Ref. 11. Data from Ref. 8. c

The extensive literature of "active" nitrogen chemistry does not appear to record any study of competition between organic substrates. Such an investigation is capable of revealing whether the reactive species produced by attack on one substrate are capable of inducing attack on the second. Such information has some relevance to the role of attack


20.

HiNDE AND LiCHTiN

"Active" Nitrogen

253

on a single substrate by reactive transients produced in primary or later events. Such an investigation also offers the possibility of evaluating the relative reactivity of structurally different substrates with respect to primary attack by N ( S ) . This paper reports an investigation of the consumption of hydrocarbons in the competition between an olefin, ethylene, and a paraffin, propane. In addition, C-labeled ethylene has been employed to provide information on resynthesis of reactants from reactive species produced in the reaction and, incidentally, on the molecu lar origin of one of the reaction products. Evidence has been presented (10,12) that, under conditions like those employed in this work, reaction of propylene is completed in the gas phase before trapping. It is assumed that this was also the case in the present work. 4


14

Table III.

(C H ) /(N) 9

e

0

0

Changes in Activity of C-Labeled Propylenes upon Reaction with Active Nitrogen" 14

A(Total Molar Activity) %

1/6 1/12

-7.7 -7.0

1/6 1/12

+5.9 +7.6

1/6 1/12

-3.8 -6.9

Δ(Atomic Activity at C-l) % Q*H - 1 -

1 4

C H„ - 2 -

1 4

6

3

C -15.3 -22 C

C.,H

+7.0 +9.5 (i

- 3-

1 4

C + 12.5 —

" Ref. 8.

Experimental The unpoisoned flow system, 2450 MHz. microwave generator, photometer, in-line gas chromatograph and most operating procedures were the same as those described previously (11) except that provision was made for introducing a steady flow of the organic substrates without interrupting the discharge. "Active" nitrogen was generated by electrode less glow discharge supported by 2450 MHz. microwaves at a total pres sure of 3 it 1 torr. Transport rates were 230 cm. sec." , 150/A mole sec." and 1.3 ± 0.2μ mole sec. for linear flow, N flow and N ( S ) flow, respectively. The latter was measured by nitric oxide "emission" titration (5). Hydrocarbon substrate was introduced countercurrent (11) into the "active" nitrogen stream at autogenous temperature (approx. 50 °C.) 20 cm. (0.087 sec.) downstream from the glow discharge. The gas stream passed through liquid nitrogen traps 50 cm. (0.22 sec.) downstream from the substrate inlet. Amounts of ethylene and propane recovered in the traps were determined by gas-chromatography on a 4 ft. X 4 mm. i.d. 1

1

2

1

4


254


silica gel column employing helium as carrier gas and thermal conductivity as the monitor. Area factors were determined periodically with knowns. In experiments involving C-labeled ethylene the molar radioactivities of products relative to that of reactant ethylene were determined by proportional counting of C 0 obtained by combustion in 0 over CuO at 450°C. of products which had been purified by gas-chromatography over silica gel, as has been described previously (8). Bernstein-Ballentein proportional counter tubes, filled with P-10 gas at 1 atm. were used with a Tracerlab P30 amplifier and SC72 scaler. Nitrogen was Matheson's "Pre-Purified" grade further purified by passage over copper wire at 500°C. C-labeled ethylene, helium used as carrier in gas chromatography, nitric oxide which was used in titration of "active" nitrogen and was purified by bulb to bulb distillation, P-10 gas used to fill counter tubes, and oxygen used in the combustion of products for assay of radioactivity were also Matheson products. Isotopically normal ethylene and propane were Phillips research grade. 14


2

2

14

Data Table IV summarizes percent consumption of ethylene consequent upon reaction of the pure substrate and three of its mixtures with propane with six different ratios of N ( S ) . Table V summarizes analogous data for propane. The data for mixtures given in the two Tables are derived from the same sets of experiments. Table VI summarizes apparent relative specific rates of consumption, "fcr H /fcc..>Hs" These ratios were calculated by means of Equation 14, the expression which would be appropriate 4

2

"*C H /V,HH" = 2

4

^

4

^

(

1 4

)

if the relative rates of consumption depended entirely on the bimolecular reaction of each of the substrates with the same reagent. Table VII summarizes molar radioactivities of recovered reactants or product ethane relative to that of reactant ethylene for the reaction of equimolar mixtures of C-labeled ethylene and ordinary propane. 14

Discussion The systematic variation of the ratio "fc /fc(.>HS" over the range of concentration parameters summarized in Table VI demonstrates that relative consumption of competing hydrocarbons is not determined simply by the relative rates of attack of N( S) on ethylene and propane since such determination would lead to constancy of the ratio. A similar conclusion can be drawn from the data of Tables IV and V by comparison of the consumption of a given substrate at a fixed concentration upon reaction with a fixed proportion of N( S) in the presence and absence of the competing substrate. Such matched points are designated in (2ll4

4

4


20.

HiNDE AND LiCHTiN

Table IV.

Percent Consumption of Ethylene"

(Total Hydrocarbon)

n 3 3 4 2 3 2

c

4 2 1 2/3 1/2 1/6

ft

Average Percent Consumption

0


255

"Active" Nitrogen

19 24* 39 45 78 98° +

x

3.0 18 27 45 59 87

1.0 28* 37 47

n 3 3 4 1 3

(

0.33

n 10 10 16 c

+

x

51 56 73° 90 84

—

10 7

93 93

—

rT

—

6 9 2 7 3

" Flow rate of N( S) was 1.3 ± 0.2μ mole sec." throughout. Pairs of numbers bearing identical superscripts are from experiments in which the ratio (C2H ) /(N)o was constant. Number of independent determinations averaged to give the tabulated figure. 4

1

b

4

0

c

Table V .

Percent Consumption of Propane"

(Total Hydrocarbon) Wo Pure

Average Percent Consumption (C H>) /(C H )

0

C' H {

4 2 1 2/3 1/2 1/6

8

29* 35 +

2

rT 1 2

—

41 48°

x

b

2 2

3.0 25 32 34 36° 39

rT 3 3 4 1 3

1.0 27* 32 32 + x

0

3

n 10 10 16 c

—

39 45

—

10 7

8

0

0.33

n

c

—

22 30 32 37 44

6 9 2 7 3

"Flow rate of N( S) was 1.3 ± 0.2/x mole sec. throughout. Pairs of numbers bearing identical superscripts are from experiments in which the ratio ( C 3 H ) / ( N ) was constant. Number of independent determinations averaged to give the tabulated figure. 4

-1

b

6

0

0

c

Table VI.

Apparent Relative Reactivities (C H ) /(C.,H ) 2

(N)

Jf

0

8

0

0

4 2 1 2/3 1/2 1/6

3.0

1.0

0.33

0.69 0.83 1.5 2.0 4.1 —

1.1 1.2 1.7 — 5.4 4.5

— 2.9 2.3 3.5 5.2 3.1

Tables IV and V by identical superscripts. If the simple model of the competition from which Equation 14 is derived were correct and, since the degrees of consumption of the pure substrates are similar, addition


256


of the second substrate should always substantially suppress the con sumption of the first and should have the largest effect when substrate is in excess over reagent. In fact, with (reactant hydrocarbon) J(N)„ ^ 1 such suppression is absent or negligible. It becomes significant* only when N( S) is in excess. The complexity of the reaction is also indicated by the shallow dependence of percent consumption of pure hydrocarbon on the ratio, (pure hydrocarbon) J(Ν) , particularly with propane for which a twelvefold decrease in the ratio is associated with increase in percent consumption by a factor of only 1.65. Such behavior is consistent with destruction of the substrate by both attacking reagent and reactive intermediates arising from the primary attack if, as the proportion of N( S) is raised, these intermediates are increasingly consumed by the reagent before they can attack the substrate. This feature is present with ethylene to a smaller extent. 4


0

4

Except with the largest excess of N ( S ) , the value of "fcc^Hé/kcaiis" (Table VI) increases with decrease in the ratio ( C H ) / ( C 3 H ) at constant values of (total hydrocarbon) /(Ν) . This systematic change is consistent with occurrence of competition between the two substrates with respect to their destruction by reactive intermediates arising from primary attack of the reagent. More specifically, (cf. Tables IV and V) ethylene appears to be more sensitive to consumption by intermediates arising from propane than vice versa. Thus, at a constant value of the ratio (Total Hydrocarbon) /(Ν) , the percent consumption of ethylene increases more steeply with decrease in the ratio ( C H ) / ( C H ) than the percent consumption of propane decreases. Presumably, the effect of attack by reactive intermediates can be reduced or eliminated by using sufficiently large excesses of Ν ( S) that the intermediates react with the latter rather than with hydrocarbon. This analysis suggests that the values of "fcc2H Ac3H " obtained with the largest excess of N( S) most closely approximate the true value of this ratio for primary attack and indicates that this true value is at least 5. An additional difficulty in estimating the correct value of Α ^ Η Α Ο Η becomes apparent upon con sidering the data of Table VII. These data establish that, over a twelve fold range of reagent ratios, synthesis of propane in the reaction mixture from fragments originating at least in part in ethylene is not detectable. However, production of ethylene either by degradation of propane or synthesis at least in part from fragments originating in propane occurs to a detectable extent with a sixfold excess of N( S) and equal initial concentrations of the hydrocarbons since the relative molar activity of recovered ethylene is significantly lower than that of reactant ethylene. In the absence of excess N( S) such synthesis was not detected, pre sumably because of the relatively large amount of unreacted ethylene. The relative molar activity data are supported by the data of Table IV 4

2

G

0

4

0

8

G

G

2

4

0

3

8

4

4

4

8

2

4

3

8

4

4


0

0

20.

HiNDE AND LiCHTiN

257

"Active" Nitrogen

Table VII. Relative Molar Activities of Recovered Reactants and Product Ethane" b

(Total Hydrocarbon),,

Relative Molar Activity

(Wo 2 1 1/2 1/6

0.98 0.98

c

7 4

0.01 "

—

—

0.8 0.5

—

0.89

—

2

1

0.01" 0.01

1 1

1 1

d

" Compared to unreacted ethylene; hydrocarbon reactants equimolar. Flow rate of N( S) was 1.3 ± 0.2μ mole sec. throughout. The number of independent experiments. Indistinguishable from the value found for unreacted propane. 6


n

n'

4

-1

r

d

which show that, with hydrocarbons equimolar, percent consumption of ethylene does not change as (Total Hydrocarbon) J (Ν) decreases from 1/2 to 1/6. With excess propane, the apparent consumption of ethylene actually diminishes over the latter change in (Total Hydrocarbon) J(Ν) . Apparently synthesis of ethylene in whole or in part from propane is even more important when propane is in excess. Such abnormal trends are not found in the consumption data for propane. The values of " & Ο Η Α Ο Η " obtained with excess N( S) are, accordingly, reduced by replacement of consumed ethylene to an extent which increases with decrease in both the ratio (Total Hydrocarbon)J(Ν) and the ratio ( C H ) / ( C H H ) . These reductions are apparent in the data of Table VI for a sixfold excess of N ( S ) . Reduction with decrease in ( C H ) / ( C H ) is also barely detectable with a twofold excess of N ( S ) . The synthesis of ethylene during reaction appears to make it impos sible to obtain better than a minimum estimate of the value of fcc H /fcc3H for primary attack. Excess N( S) must be used to suppress induced attack on hydrocarbon by reactive transients but with excess N( S) synthesis of ethylene becomes important. It is conceivable, but by no means certain, that with excesses of N( S) greater than those used in this work, synthesis of ethylene might be prevented by destruction of transient precursors. The importance of transient intermediates derived from both of the competitors is further suggested by the relative molar activity of ethane produced in the reaction (cf. Table VII). With a sixfold excess of N( S) this is derived equally from ethylene and from propane. With a twofold excess it derives to a greater extent from ethylene. Extensive data of this sort could be instructive as to the nature and reactions of such transients. The above discussion tacitly assumes that N( S) is the only com ponent of "active" nitrogen which is significantly involved. That this is so for ethylene is widely accepted. The work of Jones and Winkler (9) 0

0

2

3

4

4

8

0

2

4

0

8

0

4

3

8

2

4

0

4

0

2

4

S

4

4

4

4

4


258


suggests that it is probably also the case for propane. The latter study provides a value of 1.0 X 10 M^sec." for the rate constant of C H at 50°C. and an activation energy of 5.5 kcal. mole" . Since nitrogen atom concentrations were determined from H C N yields produced by reaction with ethylene, this constant should be adjusted downward 30 to 50%. Since the constant was evaluated from experiments in which propane was in molar excess over N( S) it is probably also significantly high because it does not correct for the consequences of attack by reactive intermediates arising from initial attack by N( S). However, the quantity actually measured was the rate of formation of H C N which was assumed to equal three times the rate of consumption of propane. To the extent that significant yields of other products are formed, this procedure will give erroneously low values of the specific rate. Herron has estimated from his elegant mass spectrometric study (7) of the reaction of N( S) with ethylene that the rate constant for primary attack at 70°C. in the (virtually temperature independent) reaction of N( S) with ethylene is 1.0 ± 0.5 χ 10 M^sec." . A value of the ratio of specific rates of primary attack by N( S) on ethylene and propane, respectively, in the vicinity of ten is thus indicated by earlier work. Agreement with the present result is well within the range of the mutual uncertainties. Herron's work further establishes that hydrogen atoms produced in the reaction of N( S) with ethylene compete with the reagent for the substrate and, in fact, suggests the model upon which interpretation of the present data is based. (!

1

S

8

1

4


4

4

4

7

1

4

4

Literature Cited (1) Bernstein, W., Ballentein, R., Rev. Sci. Inst. 21, 158 (1950). (2) Brocklehurst, B., Jennings, K. R., "Progress in Reaction Kinetics," G. Porter, Ed., Vol. 4, pp. 1-36, Pergamon Press, New York, 1967. (3) Fujino, Α., Lundsted, S., Lichtin, Ν. N., J. Am. Chem. Soc. 88, 775 (1966). (4) Hanafusa, T., Lichtin, Ν. N., Can. J. Chem. 44, 1230 (1966). (5) Harteck, P., Reeves, R. R., Mannella, G., J. Chem. Phys. 29, 608 (1958). (6) Herron, J. T., J. Phys. Chem. 69, 2736 (1965). (7) Ibid., 70, 2803 (1966). (8) Hinde, P. T, Titani, Y., Lichtin, Ν. N., J. Am. Chem. Soc. 89, 1411 (1967). (9) Jones, W. E., Winkler, C. Α., Can. J. Chem. 42, 1948 (1964). (10) Lichtin, Ν. N., "The Chemistry of Ionization and Excitation," p. 181, G. R. A. Johnson, G. Scholes, Eds., Taylor and Francis, Ltd., London, 1967. (11) Shinozaki, Y., Shaw, R., Lichtin, Ν. N., J. Am. Chem. Soc. 86, 341 (1964). (12) Titani, Y., Lichtin, Ν. N., J. Phys. Chem. 72, 526 (1968). (13) Tsukamoto, A. Lichtin, Ν. N., J. Am. Chem. Soc. 84, 1601 (1962).

RECEIVED April 25, 1967. Paper VI in the series, "Reactions of Active Nitrogen with Organic Substrates." Work was supported in part by a grant from the Graduate School of Boston University.


Chemical Reactions in Electrical Discharges

Recommend Documents