The Effect of Halogen Atoms on the Reactivity of Other Halogen Atoms

The Degradation Of Carboxylic Acid Salts By Means Of Halogen - The Hunsdiecker Reaction. Robert G. Johnson and Robert K. Ingham. Chemical Reviews ...
4 downloads 0 Views 514KB Size
3886

JACK

HINE,C. H. THOMAS AND S.J. EHRENSON

[ C O S T R I B U T I O T FROM THE SCHOOL O F

CHEMISTRY O F THE GEORGIAINSTITUTE O F

Val.

TT

TECHXOLOGY]

The Effect of Halogen Atoms on the Reactivity of Other Halogen Atoms in the Same Molecule. V. The SNZ Reactivity of Methylene Halides' BY JACK HINE, CYRUSH. THOMAS AND STANTON J. EHRENSON RECEIVED SOVEMBER 2, 1954 The effect of the various halogens as a-substituents, on reactivity by the S Nmechanism, ~ has been studied by determining the rate constants for the reactions of FCHZBr, C1CH2Br!CHZBrz, CzHsBr,CHzClzand ICH2C1with sodium iodide in acetone and also for the reaction of each of these halides, in addition to CH212. CHBIand CHqBr, with sodium methoxide in methanol ~ of the halides studied varies with the nature of the a-substituent as follows: H > F > I t is found that the S N reactivity C1 > I , Br.

Introduction There are a number of reports which show that t.he ease of displacement of a halogen atom in an alkyl halide by the bimolecular nucleophilic substitution (SN2)' mechanism may be decreased by the replacement of an a-hydrogen atom by another halogen atom. Petrenko-Kritschenko and coworkers have studied the reactivity of a number of organic polyhalides toward several reagents. They have determined the extent of reaction after a given time under a given set of conditions but, in general, have not determined the kinetic form of the reaction nor its rate constant. Products were isolated in very few cases and many of the reactions mere heterogeneous. I n a number of cases, however, reactions were carried out in a homogeneous medium and many of these probably proceeded by the SNZ mechanism. Some of the nucleophilic reagents were potassium hydroxide, sodium ethoxide, piperidine and potassium thiocyanate. Most of the following pairs of compounds were tested toward several of these reagents and in each case relative reactivities were found as shown: CH3X > CH2X2(X = C1, Br, I) ; XCHzCOzH> XzCHCOzH (X = C1, B r ) ; C6H5CH2C1> C6H5CHC12; CH3CHzX > CHaCHX2 (X = C1, Br, I). The reactivities of haloforms and carbon tetrahalides have been omitted since these probably do not react by the S N mechanism.* ~ Backer and van Melsj have reported several instances in which a ,a-dihalocarboxylic acid anions are less reactive toward potassium sulfite than are the corresponding monohalides. Toward the same nucleophilic reagent, potassium chlorobromoacetate was found to be more reactive than potassium dibromoacetate, showing that a-bromine atoms decrease S N reactivity ~ more than a-chlorine atoms in this case.5 Davies, Evans and Hulbert found that toward trimethylamine in 90% acetone, methylene chloride is only about one-fifth as reactive as ethyl chloride but more than twice as reactive as ethylene

chloride (all a t 55') and methylene bromide is slightly less reactive than ethylene bromide and only about one-tenth as reactive as trimethylene bromide (all a t 35°).6 McKay reported that methylene iodide is about one-tenth as reactive as ethyl iodide toward radioactive iodide ions in ethanol a t 60°.7 Except for the work of Backer and van Mels on chlorobromoacetic acid and dibromoacetic acid no data appear to exist permitting a direct comparison of the relative effects of the various halogens, as a-substituents, on S N reactivity ~ (although some indirect comparisons may be made). I n fact, n o kinetic study a t all appears to have been made on any compound with an a-fluorine as a substituent. Since the halogen substituents, being monatomic, are relatively simple (e.g., many of the complexities due to internal rotation and its hindrance are absent) and since they have received considerable study in physical organic chemistry, both as substituents and otherwise, we felt it desirable to carr;: out such an investigation. We have studied the effect of a-halogen substituents by investigating methylene halides since the sN2 reactivity of these compounds is greater than that of most available geminal dihalides and since competing olefin-forming elimination reactions and reactions by the S N 1 (carbonium ion) mechanism2 are very unlikely. We have studied the reactions with iodide ion in acetone, a reagent whose reactivity toward a wide variety of alkyl halides has been examined. To get some idea of the extent to which the relative reactivities of the methylene halides may vary with the nucleophilic reagent and solvent. used, we have also studied the reaction with sodium methoxide in methanol. Experimental

Apparatus.-The me,isurernent and control of tcmper:rture was carried out as described previouslL-.8 Materials.-All of the organic halides used, except incthyl bromide, were fractionated under nitrogen (CHgIz anti ( l i F o r p a r t 11. of this series see J . Hine a n d U'. H. Brader, J r . , BrCHZI a t reduced pressure) before use. Mathesoil T H I S J O ~ . R N A I . , 77,361 (1955) C.P. methyl bromide was used from the cylinder without ( 2 ) For t h e significance of t h e t e r m s Sx2 a n d Sr;l see C . K . Ingold, further purification. Chloroiodomethane and bromoiodo"Structure a n d Mechanism i n Organic Chemistry," Cornell Univermethane were prepared from methylene chloride and brosity Press, I t h a c a . K.\ 7 . , 1.957, Chap. 1-11, mide by the action of sodium iodide in acetone. Fluoro6'4) P. Petrenko-Kritschenko. D . T a l m u d , R . T a l m u d , 'A' B u t m y bramomethane was prepared from silver fluoroacetate ani1 d e - K a t z m a n a n d A . Gandelman, d. p h y s i k . C h c m . , 116, 313 (1925); bromine by the method of H a ~ z e l d i n e . ~The other halidei 1'. Petrenko-Kritschenko and V . Opotsky, Ber., 59B, 2131 (1926); were the best grades available commercially. € 3 . Petrenko-Kritschenko, rl Rawikowitsch, V. Opotsky, E. P u t j a t a Acetone was dried over calcium oxide and potassium pcrand II Diakowa, ibid , 61B,845 (1928); P.Petrenko-Kritschenko, V. (Opotsky, A I . Diakowa a n d A . Losowog, ibid., 62B,581 (1929). ( 4 ) J . H i n e , THISJ O U R N A L , 7 2 , 2438 ( 1 9 5 0 ) ; J . Hine a n d A. 31. P o w e l l , J r , i b i , l , 76, 2488 (19.54) 1 J . Hacker anti \V. 11. v a n h l r l s , Re'. I u o o ~ h i i i i . , 49, 177

(6) W. C . Davies, E . B . E v a n s a n d F. L. Hulbert. J . Chcnz .So< 412 (1939). 17) H A . C TvlcKay, T H I SJ O U R K A L , 66,702 (1943) (8) I . Hine and W. H . Brader, J r . , ibid., 7 6 , 3964 (1!133) (9) K.N. Haszeldine, J . C h e m . Soc., 4259 (1952).

,

manganate by the method of Conant and Kirner'O although it was known t h a t the material obtained thus is about 0.15 M in water." The drying method is evidently reasonably reproducible since we were able t o obtain the same rate constants with different batches of solvent. Furthermore, since other workers have also used acetone containing about this much water, our data may b e compared with theirs better than if we had used more completely anhydrous material, which is also rather difficult t o work with because of its hygroscopic character. Methanol was dried with magnesium b y the method described by Fieser.12 The sodium iodide and potassium iodate were rea,gent grade chemicals vacuum dried a t 100". Carbonate-free sodium hydroxide solutions were used in titrations. Kinetic Runs.-Runs near the boiling point of the solvent or halide being used were made by sealing a thin-walled ampoule containing a weighed amount of the organic halide, a few glass beads, and a known volume of a standard solution of the nucleophilic reagent into a small glass tube.

TABLE I REACTIONOF CICH2Br WITH KaI IN ACETONEAT 20.3" [ClCHzBrIp = 0.286 X , [NaIjo = 0.0385 M . Time, sec.

0

3,84') 10,560 15,180

[KaIIt

(1.0385 ,0356 ,0312 ,0281

3887

THES N 2 REACTIVITY O F METHYLENE HALIDES

July 20, 1955

lo%, 1. mole-' sec.-I

7.17 7.06 7.40

Time, sec

22,680 34,260 81,600

[NaIIt

10jk, I . mole-] see.-]

0.0238

7.63 ,0196 7.16 ,0076 7.47 Av. 7.32 f 0.19

All of the iodide-containing reaction mixtures were protected from the light by the use of opaque or "low-actinic" reaction vessels and/or a constant temperature water-bath t o which a large amount of black ink had been added. Most of these reawtions were also carried out under nitrogen.

Results For the reactions with sodium iodide in acetone, rate constants were calculated from the integrated second-order rate equation k = - 2.303

-

b(a

- X)

a(l;-) where a = [RXIo, b = [NaIIo, x = A[RX]t, and t = time (sec.). In these reactions only the replacement of the first halogen atom was studied, the organic halide being used in excess. The calculated rate constants showed no marked trend as the reaction proceeded, except in the cases of ethyl bromide and bromoiodomethane. The data from a typical run may be seen in Table I. The average values (and average deviations) of the rate constants for all except the two compounds mentioned are listed in Table 11. The rate constants for methylene bromide and chloride have been divided by two to obtain the rate constants per bromine or chlorine. I n addition to our data, those of Evans and Hat(a

b ) log

TABLEI1 KINETICCONSTANTS FOR REACTIONS WITH IODIDE I O NIN ACETOXE lO5k(l. mole-1 sec.-l) AH *, Halide

20.3'

50'

36'

A S ?=, e.u.

kcal.

*

* *

CHaBr" 22,9006 276" 15.9 1.0 -7.2 3 CHsCH?Br 120 f 20d 14 f Zdm" 1700 f 200d 16.4 2.0 -16.0 f 6 FCH?Br 64 zk 10 1350 f 18 1 8 . 8 += 1 . 2 -9.1 f 4 ClCHzBr 7.32 =I= 0.19 47.3 f 0 . 5 218 f 5 20.9 =I= 0 . 5 -6.2 f 2 BrCH2Br' 2.03 f 0 . 0 8 16.0 ?c 0 . 4 69 i~ 1 21.7 3z 1 . 2 -6.0 f 4 ICH2Br 5 f 0.6d 100 f 20d 18.4 & 2 . 5 -15.Fi =!= 7 ClCHZC1' 0,842 f 0.02' 0 . 2 1 1 f 0.003 29.0 += 1 . 0 +5.0 3 ICH2Cl 0.155 f 0.005 a From data of ref. 15c. At 20.0'. At -19.6'. Estimated by extrapolation to zero time. e At 0.0'. The observed rate constants have been divided by two t o get the rate constants per bromine (or chlorine) shown. At 60.0'.

*

Q

rhis tube was allowed t o reach thermal equilibrium in the constant temperature bath and was then shaken t o break the ampoule and start the reaction. The reaction with iodide in acetone was stopped by breaking the sealed tube into ice-cold hydrochloric acid, and the unreacted iodide ion was titrated with potassium iodate.l0J3 I n this titration t h e iodide is first oxidized t o iodine and then t o iodine chloride. T h e end-point is the disappearance of iodine from the few milliliters of carbon tetrachloride used as a n indicator. Since Senior, Hetrick and Miller have reported that acetone interferes with the end-point when chloroform is used as an i n d i ~ a t o r , 'we ~ have made tests and have shown that this is not the case when carbon tetrachloride is used. The reaction with sodium methoxide in methanol was stopped b y breaking the tube into a known volume (excess) of standard hydrochloric acid and back-titrating t o the rosolic acid end-point, except in the case of fluorobromomethane where the reaction mixture was added t o 5 ml. of methanol a t -80" and the unreacted sodium methoxide titrated t o the brom phenol blue end-point with methanolic p-toluenesulfonic acid. Reactions a t lower temperatures were carried out in volumetric flasks, samples being withdrawn by pipet a t various times. (10) J. B. C o n a n t and W. R . Kirner, THISJ O U R N A L , 46, 232 (1924). (11) A . R. Olson, L. D. F r a s h i e r a n d F. J. S p i e t h , J . P h y s . Chcm., 66, 860 (1951). (12) L. F. Fieser, "Experiments i n Organic Chemistry," P a r t 11, 2nd E d . , D. C . H e a t h a n d C o . , Boston, Mass., 1941, p. 359. (13) L. W. Andrews, THISJ O U R N A L , 25, 736 (1903). (14) R I, Senior, R. R . Hetrick a n d J . G. Miller, i b i d . , 66, 1987 (1944).

mannlj on methyl bromide has been included for the purpose of comparison. Also listed are heats and entropies of activation calculated from the absolute rate equation16

For ethyl bromide, rate constants calculated from equation 1 fell sharply as the reaction proceeded. Dostrovsky and Hughes made a similar observation for the reaction with both sodium and potassium iodide a t 64' and attributed the fall to reversibility. l7 They obtained rate constants for the reaction with lithium iodide a t 0 and 20°, but the data for the individual points are not given. The lithium iodide reaction, in which the metal bromide formed does not precipitate, was said to be only about 10% complete a t equilibrium a t 64" under the conditions used. The fall in our rate constants appears to be partly, but not entirely, due to reversibility. We have calculated rate (15) A. G . E v a n s a n d S. D . H a m a n n , T u a n s . F a v a d a y S O L . ,47 25 (1951). (16) S. Glasstone, K . J . Laidler a n d H. Eyrinp, " T h e T h e o r y o f R a t e Processes," McGraw-Hill Book Co.,Inc , New York, K ,Y . , 1941, I ) . 14.

(17) I. Dostrovsky a n d E. D. Hughes, J . Chcin. S o c . , 161 (1946).

constants by use of an equation derived on the assuniption that the reaction was reversible but homogeneous, and another equation which assumed t h a t the sodium bromide precipitated very early in the reaction with its concentration in solution reiiiaining constant thereafter. The fall i n these rate constants, specially those calculated horn the second equation (whose assumptions are probably nearer the truth), was less than in those calculated from equation 1, but it was still considerable. 1 i - e have checked the purity of our reagents arid used both sodium aiid potassiuni iodide. Runs have also beeii tnade independently by thrcc diiferent investigators," but we still have fourid no explanation for our results. The rate constants listed for ethyl bromide in Table I1 were obtained by extrapolating our data t o zero time. The values a t 0 and 20" are highcr than those of Dostrovsky and Hughes, perhaps because of the greater tendency of lithium iodide to exist as an ion-pair. Our rate constants may be estrapolated to 64' to !.ieltl a value in reasonable agreement with the data thcsc workers obtained using sodiuin and potas,siiiiii iodide. The rate coiistants ohtaiiieci in the reactioii oi hromoiodoniethane with sodium iodide in acetonc also fell as the reaction proceeded. Because of the greater sensitivity of this compound to air and light and the possibility of the reaction of the sodium bromide formed with both, the organic reactant and product, this case may be more complicated than that of ethyl bromide. Therefore, without studying the reaction in great detail, we obtained rate c ~ t i stants by extrapolation to zero time. For these rate constants estimated by extrapolation and for the heats and entropies of activation for all of the reactions we lial-c attempted to coiisider all relevant factors in :in estimate of the de\riation suilficient t o bring the reliability of our data to the S5cG prohahility kl-rl. These de\-iations are listed ill Tables I1 a i i d TV. - . Siiice :ilkyl hrniiiidcs l i a i ~hcen fouiid to hr from .LIO to I>On fitiics as rt:ic(ive toward indirle inn i n

ciable concentratlon of it ever accumulates, eqitation 2 is applicable.

wlierc (1 =: [CH,X2ln,b = [ K a O l l c j , , , ; i t i ( I x = A[CH2X?ILso that h is cxpressed in 1. (iriolcs of CH?X?) see.-' Equation 2 was used for all of tlic methylene halides except Auorohroiiionieth~tiic. Data on chloroiodoinethaiie, the incniber of this zroup for which aceuinttlation of the iiiteriiictliate woiild : i ~ ) p c atn r hc inost likely, arc shon-it i n 'I'ablc I JI. .1_ .4RI.b. I I I K I : A C I I OO~F CICI121

ivI.rii

SaOhle

IS

nIc(-)li

Icc.

7,740 1 1 ,940

23 , 40il ::: , 5.Kr i

.';7,140 (IC), 2111 S I , tYi

IClCllrlI. 11

'

[NAOAlCl,

-1 ,so

1.2% 1 .,12

Oi:3!I"

1,27

1m?

1,2;;

.I\, "

~ S C 3 , O N V ] (., I~.";i7,S.

I'

'

I . :ii1 1.