Rearrangements of α-Halogenated Ethers. I. 2,2,3,3-Tetrachloro-p

R. K. Summerbell and. Daniel R. Berger. Vol. 70 water, dilute sodium carbonate solution and saturated so- dium chloride solution before being filtered...
0 downloads 0 Views 416KB Size
water, dilute sodium carbonate solution and saturated sodium chloride solution before being filtered through Drierite and theii concentrated t o yield 3.79 g. (97yo)of light )-ellow solid, m.p. 69-74'. After the first recrystallization froin methanol tlie yield was 2.79 g. (71yo)of yellow iieedles, 1n.p. 71.5-74O. T h e iiifrarecl spectrum indicated t1i:it both the cyano atid nitrci groups were present. T h e analytical sample, 1n.p. 72-75', was prepared h>- threc su1)sequeiit crystallizations from inetlianol. AmzZ. Calctl. for CI4Hlo02Sz:C, 70.6; I I , 4 . 2 ; S,11.8. Found: C, 70.6; H,4.4; Iu, 11.8. 2 4 2 '-Nitropheny1)-phenylacetonitrile (X) with Sodium Hydroxide.-In duplicate runs 0.205 g. (0.87 nimole) o f S iu 0.55 ml. of methanol was rnixed with 0.20 g. (5 innioles)

[COSTRIBUTIOS FROM

THE

of sodium hydroxide. Twenty-four seconds after the beginning of t h e addition of t h e base a yellow precipitate began separating; t h e temperature had risen from 33 t u 38' in about 1 inin. hfter about 8 inin. from the time of iiiixiiig t h e precipitate was filtered and waslied. T h e reactioii misture had been kept below room temperature with iccwtter about 5 of tlie 8 miii. T h e yieltl was O.1GG g. (Si',;) of ycllow crystals, 111.p. 218.>220.5° with 110 depressioii i i i in.1n.p. with authentic XII. In another run duplicating t h e above except t h a t tlie rcaction mixture was refluxed for 5 Inin., a yelliiw precipitate formed instantaneously but it dissolved aiitl no pure XI1 could be isolated by the above procedure. ~IORGASTOUX, IY.I-A.

CHEMICAL LABORATORIES O F NORTHWESTERS USIVERSITI']

Rearrangements of a-Halogenated Ethers. I. 2,2,3,3-Tetrachloro-p-dioxane BY R.K. SURI~IERBELL AND D.INIELR.BERGER' RECEIVEDJULY 25, 1957 T h e compound previously assigned the structure 2,S,3,3-tetrachloro-p-dioxane (111) has been shown t o be (2-chloroethoxy)-dichloroacetyl chloride (IV). I11 has been synthesized, its structure proved, and t h e formation of I V from 111 b y a thermal rearrangement has been demonstrated

Discussion Some ten years ago the high temperature chlorination of tvans-2,3-dichloro-p-dioxane2 was described,3 and the liquid product, obtained in 55%; yield, WRS assigned the structure 2,2,:3,3-tetrachloro-p-dioxane. This assignment seemed justified as solid derivatives of both ethylene glycol a1.1~1 oxalic acid were isolated from the hydrolrsate ; no other tetrachlorodioxane could gi.i-e these derivatives. \Ye recently have had occasion to prepare this compound, and a routine examination of its infrared spectrum revealed a strong absorption a t 5.55 p , entirely inconsistent with its formulation as tetrachlorodioxane. I n order t o resolve this inconsistency, the tetrachlorodioxane was prepared by an alternate route. Addition of chlorine t o 2-chloro-p-dioxene gave 2,2,3-trichloro-$-dioxane (I). Removal of hydrogen chloride from I b y means of sodium hydroxide gave 2,3-dichloro-p-dioxene (II), which upon addition of chlorine gave the desired 2,?,3,3-tetrachloro-p-dioxane (111). However, I11 is entirely different from the previously described liquid, I\-. IT1 is a stable white solid melting a t 140°,whereas IT: is a water-white mobile liquid boiling a t 20.5'. Cornpourid IV reacts "violently" with water and alcohols ; 111 can be recrystallized successfully froni aqueous ethanol a n d is stable to a. humid a t mospliere. Furthermore, 11I does riot absorb in the region of 5-6 p . The structural relationship between 1I1 and IV was established as follows: Both gave correct analyses for CdH40?C14,and after hydrolysis both gave derivatives of oxalic acid and ethylene glycol. \Then 111 was heated under reflux in tetrachloroethane solution, with the periodic withdrawal of (1) Northwestern University Fellow, 1955-1~5(1: .4llied Chemical and Dye Corp. Fellow, 195fi-1957. ( 2 ) See R. K. Summerbell and H. E.Lunk, THISJOURXAI.. 7 9 , 4802 (1937), for t h e structures of cis- and Irails-~,3-dichlorti-p-dioxane. ( 3 ) R . K . Summerhell, R . 12. TJmhoefer a n d G. R . T.appin, i h i d . , 69, 1332 (19471.

aliquots for spectral examination, the infrared spectrum of I11 was observed t o change slowly to t h a t of IV, the apparently first-order reaction requiring approximately 30 hours for the disappearance of 111. Finally, on heating IT: at 190' without sol\-ent, there was a very slow distillation, the products being ethylene chloride and oxalyl chloride. On the basis of these observations, the liquid IV is not 2,2,3,3-tetrachloro-p-dioxane, but rather must be (2-chloroethoxy)-dichloroacetylchloride. I t is entirely possible t h a t in t h e previously described

I

I1

c1 C l IV

preparation of IV3, I11 is actually formed as an intermediate. Howex-er, the reaction temperature (153') and the time taken for the reaction (3-1 hours) are more than are required to permit essentially complete isomerization of 111 to IIT; should any I11 still be present a t the end of the reaction, i t would be isomerized in the atmospheric distillation used in the purification scheme for IV. Confirmation of this was obtained by heating I11 a t 210' for 30 minutes. On cooling, the liquid did not solidify, and the infrared spectrum of I11 was no longer present, having been replaced b y t h a t of TY. The observation t h a t both I11 and IV could be hydrolyzed to yield derivatives of ethylene glycol requires some comment. Hydrolysis of 111 would be expected to yield t h e glycol, whereas the hydrol\.si? of I\' should instead give ethylene chloro-

Dec. 20, 1057

6505

2,2,3,3-TETRACHLORO-p-DIOXANE

hydrin. I n the previous workj3the glycol was not isolated as such, b u t rather was identified as its dibenzoate prepared b y a Schotten-Baumann reaction. Under the basic conditions employed, the

low temperature (-20') chlorination of isopropyl ether. T h e isolation was achieved b y a distillation at 83-88', during which the type of rearrangement under discussion probably occurred. I n a similar reaction of an a-haloester, a-chloroethyl acetate is reportedg to give acetyl chloride and acetaldehyde upon heating t o 560-5!30°. The yield is SS%, with a conversion of 41%. There are also related reactions in the l i t e r a t ~ r e ' ~ - for l ~ which the authors are led to postulate the formation of a halogenated acetal followed b y a similar type of rearrangement t o give an alkyl halide and ester.

chlorohydrin, which was formed b y the hydrolysis, would be dehydrochlorinated t o give ethylene oxide, which would then react with water to give the glycoL4 It was therefore necessary, in order OR OR t o firmly establish t h e structure of IV, t o isolate I I R'-C-X +R'-C RX ethylene chlorohydrin (or a derivative which could I 11 be formed from ethylene chlorohydrin b u t not from ethylene glycol) from the hydrolysis products of Pyrolysis of 2,3-dibromo-2,3-dichloro-p-dioxane IV. This was accomplished b y neutralizing the hydrolysis solution with ammonium hydroxide, (V) gives a rearrangement analogous t o t h a t of care being taken not t o allow the solution t o be- 111. 2,3-Dichloro-p-dioxene, on addition of brocome basic. T h e chlorohydrin was removed b y mine, gave the solid compound V. On pyrolysis distillation as its azeotrope with water, and the dis- of V a t atmospheric pressure, a liquid, VI, was obtillate was treated with p-anisidine according to the tained. VI has not as yet been completely characmethod of Jacobs and HeidelbergeP to give 2- terized, b u t its infrared spectrum is quite similar t o (p-anisidin0)-ethanol, P - C H ~ O C E H & H C H ~ C H ~ O Ht h, a t of IV, so t h a t its structure is most probably a compound which, under the conditions used, t h a t of ("bromo- /or chloro1ethoxy)-bromochloroacetyl chloride (or bromide). could not be formed from ethylene glycol. The rearrangement of I11 t o give IV proceeds ExperimentalI4 a t a lower temperature than does the rearrange2-Chloro-p-dioxene was prepared by the method of Summent of IV to give ethylene chloride and oxalyl nierbell and Lunk.2 chloride, as is evidenced b y the fact t h a t IV was iso2,2,3-Trichloro-p-dioxane(I)was prepare$ by the method lated from its original preparation. A satisfactory of Lunkl5; b.p. 60-61" (1 mm.), m.p. 20-21 . 2,3-Dichloro-p-dioxene (II).-A mixture of 6.0 g. (0.15 explanation is found in a consideration of the bulk mole) of sodium hydroxide in 5 ml. of water and 20 ml. of of t h e chlorine atoms of 111,and the fact t h a t 111 is catbitol held at 200-205" (bath) was treated dropwise with constrained in a ring system. On both 111 and IV, 19.2 g . (0.10 mole) of I in 20 ml. of carbitol. The slow addithe migrating chlorine atoms may be in a favorable tion took about 0.5 hour, and from time t o time small of water were added to replace that which disposition for the rearrangement; the rearrangement amounts tilled. The distillate was milky and boiled from 110-140". of 111 t o IV, however, gives a much greater relief After the addition of the trichlotodioxane was completed, 20 of strain than does the corresponding rearrange- ml. of water was added in small portions and allowed to disment of IV, and consequently the former requires til. The heavy organic layer of the distillate was taken into a lower energy input. ether, washed with water and dried over magnesium sulfate. Rearrangements of a-haloethers t o give alkyl The filtered solution was freed of ether and the residue halides and carbonyl derivatives are not unknown. was distilled in Z ~ Z ~ O There . was very little forerun, and Over a century ago, MalagutP reported perchloro- practically no residue. The yield of 2,3-dichloro-p-dioxene, ~ w a s 8 g . (52%). ethyl ether to give, on pyrolysis, trichloroacetyl b.p. 65'(2.5mm.), n Z 51.4930, Calcd. for C4H401C12: C, 31.00; H , 2.60. Found: chloride and hexachloroethane. Shostakovsky and C ,Anal. 31.33; H , 2.70. Bogdanova have reported' a-bronioethyl butyl 2,2,3,3-Tetrachloro-p-dioxane (III).-A solution of 20 g. ether to give, on slow distillation, 97yo of butyl (0.129 mole) of I1 in 50 ml. of methylene chloride was stirred bromide and traces of acetaldehyde. Most of the with cooling on the ice-bath. Chlorine mas added slowly oxygen, however, was recovered as water. T h e until the solution turned yellow; the yellow color was removed by "back-titrating'' with a few drops of the dioxene. same bromoether, after three weeks in a sealed The solution was cooled on a Dry Ice-acetone-bath, and tube in diffuse light, gave SO$& of butyl bromide three crops of crystals were obtained, which, after recrystaland again, water. a-Rromoethpl ethyl ether lization from ether, gave 20 g. (0.088 mole, 67%) of the similarly gave, on heating, 8070 of ethyl bromide, white solid. A satisfactory melting point could not be obby recrystallization; however, a sublimed sample with the formation of some acetaldehyde. Hall tained (50" (0.5 mm.)) melted a t 137-140" (sealed tube) The coniand SireP obtained a 20% yield of 1,3-dichloro- pound can be recrystallized from aqueous ethanol with no propanone as the only isolated product from the change in its infrared spectrum. Infrared absorption bands

+

are (s (4) T h e reaction of ethylene chlorohydrin with base t o give ethylene glycol is well described in t h e literature. For references, see G . 0. Curme a n d F. Johnston, "Glycols," A . C . S . Monograph No. 114, Reinhold Publishing Corp., New York, N. Y . , 1952, p 15. ( 5 ) W . A . Jacobs and M . Heidelberger, J . B i d . C h e m . , 21, 421 (1915). ( 6 ) M . J. hlalaguti, A n n . chim. c f p h y s . , [31 16, 5 (1848). (7) hl. F. Shostakovsky and A . V. Bogdanova, Zhiw. Obshchti Khim., 21, 388 (19.51); English translation, J . Gen. Chenz. U . S.S.R . , 21, 429 ( 1 951). ( 8 ) G . R. Hall and I. Sirel, TAISJOURNAL, 74, 8 3 6 (1952).

=

strong; m, medium; w, weak; sh, shoulder; re-

(9) Cons. Elektrochem. Industrie, P B 74539, frames 7431-34.

(IO) S. X'I. RIcElvain and D. Kundiger, THIS JOURNAL,64, 254 (1942). (11) E . N. Marvel1 a n d M. J. Joncich, ibid., 73, 973 (1951). (12) S . R l . McElvain and C . H. Stammer, ibid., 7S2 2154 (1953). (13) J. B. Wright, ibid., 7 7 , 4883 (1955). (14) Microanalyses were performed by Miss H. Beck. Unless otherwise stated, melting points were taken on a Fisher-Johns block and are uncorrected. (15) H . 5 . Lank, P4.D. Thesis, Northwestern University, 1967.

6.i06

R.K. SUMMERBELL

AND

DANIEL R.BERGER

1701.

79

ported as microns in carbon tetrachloride solution): 6.85~7, The above 6 ml., estimated from the refractive index t o 6.91w, 7.86w, 7.98w, 8.82s, 9.00s, 9.38111, 9.60~~1, 10.16s, contain approximately 4 g. (0.05 mole) of ethylene chloro10.65s, 11.80m, 13.9s, 14.2sh. -1 very weak carbon-hyhydrin, was heated under reflux for two hours with 12 g. drogen band appears a t 3.42 p . (0.10 mole) of p-anisidine and 10 ml. of water.'? T h e mixAnal. Calcd. for C4Ha02Cl4: C , 21.27; H , 1.78; mol. ture was made basic with sodium hydroxide and extracted wt., 225.90. Found: C, 21.41; H, 1.75; mol. wt., 231 with ether. After drying over magnesium sulfate, the ether \vas removed on the steam-bath and the residue dis(in benzene). tilled in vacuo. A micro distillation apparatus, having a Hydrolysis of 2,2,3,3-Tetrachloro-p-dioxane (III).-Conicolumn of only 1.5 inches, mas used, and no real attempt a t a pound 111, 1.0 g., and water, 30 g., were heated overnight on the steam-bath. Then 1.1 g. of calcium nitrate was careful separation of p-anisidine from 2-(p-anisidino)-ethaiiol was made. The distillate was taken in three fractions, added, and the mixture was heated for another hour. The precipitate was filtered and dried, and had a n infrared ab- the last of which comprised about 35% of the total. After two recrystallizations of the third fraction from ether a t sorption spectrum (potassium bromide pellet) idelltical with -78", 1.1 g. of compound, 1n.p. 41.0-42.0°, was obtained, that of calcium oxalate prepared from oxalic acid. Sodium hydroxide, 2.0 g., was added t o the filtrate. X identical in the infrared with 2-(p-anisidino)-ethanc,l prepared by the method of Jacobs and Hei~lelberger.~The precipitate appeared, which was removed by filtration and yield was 0.007 mole ( 10TO). discarded. T o the filtrate was then added 0.7 g. of benzoyl A n a l . Calcd. for C9H1302r\T:C , 64.65; H, 7.81. Foulid: chloride, and the solution was heated for 10 minutes on the C, 64.53; H, 7.58. steam-bath. -1fter sitting overnight a t room temperature, the solid which had formed was removed, washed with ethaPyrolysis of (2-Chloroethoxy)-dichloroacetyl Chloride nol, and dried, t o give a compound, m.p. 67-69". It was (IV).-CompoundIT, 15g.,masheatedto 190°0nnsand-bath identical in all respects with authentic ethylene glycol diben(b.p. of IV is 205'). Xvery slow distillation througha4-inch vacuum-jacketed Vigreux column occurred; after 60 hours, zoate. (2-Chloroethoxy)-dichloroacetyl chloride (IV) was pre- 5 g. of distillate had been collected. The infrared spectrum pared by the method of Summerbell, Umhoefer and L a ~ p i t i . ~showed, by comparison with the known compounds, that the distillate consisted entirely of oxalyl chloride and ethylT h e product, n2'D 1.4805, boiled at 205" (760 mm.) and 7678" (6 mm.). Infrared absorntion bands are irenorted as ene chloride. hniline was added dropwise to 0.5 ml. of the distillateuntil microns in carbon disulfide soLtion): 5.555, 7.61\;~, 7.661n, 8.52s, 9.24m, 9.755, 9.90sh, 10.30n1, lt.ZOw, 11.7C,n., 12.4~11, no more vigorous reaction was noted. The precipitate was triturated in ethanol, leaving a solid which melted a t 25213.2s, broad, 14.71~1,15.3m, broad. Isomerization of 2,2,3,3-Tetrachloro-p-dioxane (111) to 253". The infrared spectrum was identical with that of (Ixanilide (m.p. 253') prepared from authentic oxalyl chlo(2-Chloroethoxv'l-dichloroacetvl Chloride iIV).-Comuound ride in a similar manner. 111 was dissolded in redistill& sym-tetrachloroetlian~in a 2,3-Dibromo-2,3-dichloro-p-dioxane (V).-Compound I1 concentration of 54 mg./ml. arid heated under reflux (141.5') (8.3 g., 0.054 mole) was dissolved in 50 nil. of carbon tetrausing a calcium chloride tube t o ensure dryness. Aliquots chloride. To i t was added, with stirring over 0.5 hour, a were withdrawn periodically for spectral analysis. -1r1 solution of 8.65 g. (0.054 mole) of bromine in carbon tetraattempt was made to determine the rate of disappearance of chloride to make 50 ml. An ice-bath vas used t o cool the I11 by using its band a t 10.65 p ; however, the results of reaction. After standing for three hours, the solution was different runs were not consistent and no reproducible rate constant was found. The v ~ r i o u sruns (lid, hmvevcr, d l heated under reflux for 0.5 hour and the solvent removed by distillation from the steam-bath. The residue, on standing follow first order kinetics, and t,he bands due to I11 disappeared after 24-36 hours, being replaced gradually with the overnight, deposited white crystals which were filtered and irashed with cold ethanol to give 6.3 g. (37y0)of 5' melting spectrum of I\'. In refluxing carbon tetrachloride w l u t i m over :t wide range up to 1-10'. Recrystallization from etha( 7 7 O ) , I11 is unchanged after three days. nol gave a 1n.p. of 103-130" (sealed tube). The material Pyrolysis of 0.5 g. of I11 a t 210' for 30 minutes led to can also be purified by sublimation, but it is not stable complete isomerization to IT.'. The infrared spectrum of the enough to give a satisfactory melting point. On standing, liquid product was mostly that of I V , with a small amount of IT decomposes to give fumes with ammonia vapor; it is, ethylene chloride and oxalyl chloride present. I l l was tohowever, fairly stable when stored in a desiccator over sotally absent. dium hydroxide pellets. Hydrolysis of (2-Chloroethoxy)-dichloroacetyl Chloride (IV).-Hydrolysis of I V to yield oxalic acid and ethylene Anal. Calcd. for C4H401Br?Clz: C , 15.26; H, 1.27. glycol dibenzoate already has been d e ~ c r i b e d . ~The isolaFound: C, 15.27; H, 1.24. tion of ethylene chlorohydrin as 2-( p-anisidino)-ethanol Pyrolysis of 2,3-Dibromo-2,3-dichloro-p-dioxane (V).was accomplished as described below. Compound V (2.50 g., 0.00794 mole) was placed in a 5-nil. Compound I V (15 g., 0.067 mole) anti 50 r n l . of flask fitted to a short distillation head. Distillation a t atwater were allowed to stand together overnight; during the mospheric pressure gave 2.02 g. of a red-orange liquid, b.p. first hour the reaction ~vaqexothermic. On cooling, erys180-185". Redistillatiou gave a colorless liquid, b.p. 55' tals, probably oxalic acid, sep:trated and were removed by (0.3 m m . ) , ? z * ~ D1.5167. The infrared absorption spectrum filtration. The solution was then neutralized with ammowas quite siniilar to that o f IT, a strong carbonyl band a t nium hydroxide, care being taken t h a t i t be no tnore basic 5 . 3 p being the most prominent absorption. than pH 7 (the color changes from water-white to yellow as XOTE ADDED I N PROOF.--The chlorination Of p-dioxane to the neutralization point is neared). On cooling in ice, crysgive a ?OYo yield of trichloroacetyl chloride, along with some tals (presumably ammonium oxalate) formed which were fil- hexachloroethane, has been reported recently.'& This result, tered and discarded. The filtrate was distilled through i: obtained in a chlorination a t temperature above 160", quite 10-inch Vigreux column, and the first 30 ml. was collected iu probably involves the formation of halogenated intermediates three 10-ml. fractions, n z 5 1.3535, ~ 1.3430, 1.3393, respecfollowed by rearrangements of the type discussed in the tively. (Ethylene chlorohydrin is reported to form an azeopresent paper. trope, containing 58y0 water. boiling at 97.8'.16) The first EVANSTOS, ILLINOIS two fractions were recombined anti distilled; the first 6 1111. was collected, n Z 5 D1.3591. Data," Advances in Chemistry Series, No. F, American Chemical Society, Washington, D . C . , 1432, p. 5 . ( 1 8 ) "Azeotropic

(17) T h e procedure from this point is modified from t h a t of Jacobs and Heidelberger, ref 5 . (IS) N. B. Lorette J . Orp. C h e m . , 22, 843 (1957).