1774
JOHN
Vol. G8
A. KINGAND FREEMAN H. MCMILLAN
ing point determination with authentic 2,4-methylenexylitol caused no depression of this value. The yield was 2.8 g. (889$). The series of reactions constitutes a definitive proof that monomethylene-gluco-gulo-heptitolis 3,5-methylene-gluco-Qulo-heptitol. Although the expected dialdehydo oxidation product could not be crysfrom 3,5-methylene-gluco-guZo-heptitol tallized, i t was isolated in a separate experiment in the Form of its di( phenylhydrazone). The latter compoui~d crystallized from 30 parts of alcohol as small, light yellou plates which melted with decomposition a t 188-190'. Anal. Calcd. for Cl8HzOOlS4:C, 63.51; H, 5.92; 1, 16.46. Found: C , 63 23; H , 6.10; Ir;, 16.56
Summary Proof is presented that the monobenzylidene-
gluco-gulo-heptitol (m. p. 21s') which Emil Fischer discovered is the 3,5-benzylidene acetal. The condensation of gluco-gulo-heptitol with formaldehyde also gives a high yield of a monoacetal which proves to be 3,5-monomethylene-glucocub-heptitol. These 3,5-monoacetals are those t o he expected from generalizations relating the configuration of polyhydric alcohols and the structure of their derived benzylidene and methylene acetals. Crystalline 2,4-benzylidene-xylitol and a number of its derivatives have been described. RECEIVED MAY9, 1946
RETHESDA, R~ARYLAND
'CONTRIBUTIOS FROM THE STERLING-\VIS1 IIROP R E S E a R C H [NSTIrUTE ]
The Preparation of Some Monoalkyl- and Symmetrical Dialkylethylenediamines BY JOHN A. KINGAND FKEEXAKH. I t was hoped that a conveiiient preparation of monoalkyleth.ylenedinminesl~lacould be realized by making use of the fairly readily available 2methylimidaz,oline. Following the observation of Ladenburg2 that 2-methylimidazoline methylated easily, and the finding by Aspinal13 that 2-substituted. imidazolines could be hydrolyzed to ethylenediamine without great difficulty, i t seemed reasonable to assume that the preparation of 1alkyl-2-methylimidazolines and their hydrolysis t o monoalkylethylenediamines should afford a satisfactory ;source of such diamines. In the present work only moderate success was attained by this method. Commercial (70y0 j ethylenediamine was converted in essentially quantitative yield to the symmetrical 'diacetyl derivative4 with acetic anhydride. By a procedure based on Chitwood and Reid's5 modification of Ladenburg's original preparation of the material, symmetrical diacetylethylenediamine was pyrolyzed over magnesium powd.er to givc 2-methylimidazolitie in yields ranging from 86 to 945,;. This represents some improvement over previoitsly reported yields for- this reaction.6 (1) A literature review of the preparation of monoalkylethylenedi JOURNAL. 63, 852 (1941). amines bas been given by Aspinall, THIS Bloom, Breslow and Hauser, i b i d . , 67, 539 (1945), have recently reported the preparation of isoamylethylenediarnine in 2OY0 yield hy sodium and alcohol reduction of isoamylaminuacetonitrile: while still more recently 1,iuskt.r and Evans, ibid., 67, 1581 (19451, have reported the preparation of higher monoalkylethylenediamines in 93 t u 9S% yield by direct alkylation. The present work was com,iIeted before the rrppearance of the last paper cited. ' la) Pearson, Jones and Cope, ibid., 68, 1225 (1946), have very re.. r.cntly described the preparation of cyclohexylethylenediamine and i~opropylethytenediamine. (2) I,adenburg,*Ber., 27, 2957 (1894). (3) Aspinall, J . Org. Chem., 6 , 895 (1941). ( I ) Hofmann, Ear., 21, 2332 (1888). 5 ) Chitwood and Reid, THIS JOURNAL, 67, 2424 (1935). 6) Kyrides, U. S. Patent 2,392,326, prepared this and other lower alkyl 2-substituted imidazolines in substantially lower yields by cstlciiirn oxide cyclization of monoacylethylenediamines.
'N H
"
--" i
N 'CHI
The alkylation reaction proved to be much more com:,lex than had been anticipated; in addition to the desired product (C) there was T X H --nw -1 -SR LNiCH3 + RX
A
+
1
x?CHrHX f ,I$ "CH3 1 B
--9
C
formed the quaternary amnioiiium salt (Dj, as well as the hydrohalide salt (B) of the starting material (A). In an effort to prevent the formation of the hydrohalide of the initial imidazoline, which removed the starting material from participation in any addition reaction, some of the reactions were run in the presence of potassium carbonate to neutralize the halogen acid as soon ;is it was formed. However, only quaternary products were formed. 'l'he results obtained are summarized tn Table 1. Our alkylation data are in general agreement with the very recently available results of Kyrides6 who obtained 30 to 40yo yields (based on halide used) of 1-higher alkyl-2-lower alkyl imidazolines by alkylation of 2-substituted imidazolines with only one-half mole of alkyl halide in an hydrocarbon solvent. He did not hydrolyze the imidazolines to ethylenediamines or report any investigation of higher alkylation products.
XONOALXYLAND SYMMETRICAL DIALKYLETHYLENEDIAMINES 1775
Scpt., 1946
TABLE I" h4oles Alkyl halide of used halide ChHjCHiCl 4 it-GHoBr 2 .I-CIHYRT 2 v-CdIsBr I C Ha1 2 ,i-CaHiBr I Ii-CEHIIBr 1
Yield Yield Yield (%).of (%) of (%) of mimono- dialkyldazoline alkyl- ethylhydro- ethylene- eneSolvent halide diamine diamine CalI6 46 8.3b 6.2b CRH~ 31 23 19 HiO, KzCOa. EtOH 0 0 34 I h O . KzCOI, EtOH 0 0 38b CeHs 34e Hz0, EtOH 23 11.6 H20. EtOH 23 12
All yields are based on imidazoline used. As monoAs alkyl halide quaternary of the diacetyl derivative. alkylimidazoline. 0
The formation of the various products in the alkylation can be readily explained by consideration of some theoretical aspects of the reaction. The relative basicities of A and C are determined by the relative stabilities of the ions produced by Lidding a proton to each. I n the case of A this ion has two absolutely equivalent resonance forms and the resonance energy is maximal, while in
I
1
N
]-I+
It
SH--+HN
-'Y
I
CH3
- 1
NH-HN
CHa
I +
f&y CHs
the case of C: the forms are only approximately cquivalent; therefore A is a stronger base than C ( t h eions formed from X and C in which the proton !
I
iI+
N\/NR-+HX
I CH,
+I
-a/,
I
NRt--,HN CHa
, +-,\,
+
XK CJ&
atids to the already substituted nitrogen atom do not make major contributions to the resonance). However, the situation is somewhat different when an alkyl group is added to A and C, to form a quaternary ammonium ion. The addition of
CHI
CH*
CH3
alkyl halide K X to X forms an ion in which the resonance structures are nori-equivalent, :In
already present in the 1-position. In other words, there is no exact parallelism between basicity and quaternizability in these compounds. We take pleasure in acknowledging the suggestions of Dr. Elmer J. Lawson regarding the alkylation mechanism herein presented. Experimenta17sx 2-Methylimidazo1ine.--In
a ~~O-CC round-bottomed ., flask was placed a mixture of sym-diacetylethylenediamine' (258 g,, 1.79 moles) arid magnesium powder (21.ti g., 0.90 mole). The flask was fitted with a downward condenser and was heated in a metal-bath a t 31C)--315°. The 2-methyliIiiidazoline distilled as a nearly colorless liquid which crystallized in the receiver. The distillate was crystallized from dry benzene (250 cc.) t o give 112 g. (WC( yield) of material, m. p. 85'. This melting point is considerably lower than that of a redistilled sample which melted a t 100-103 '. While the redistilled material conipared favorably with that of Ladenburg2 who reported a m . p. of 105", the once recrystallized material was whit?, nicely crystalline and of satisfactory purity for the alkylation reactions. Alkylation with Benzyl Chloride.-Benzyl chloride (126 g . , 1 .OO mole) was dropped into a solution of 2-inethylimidazoline (42.0 g., 0.50 mole) in dry benzene (150 cc.). The solutioii was refluxed two hours then all the solvent was removed untlcr vacuum. The residue was taken up iii a I;niall amount of dry ethaiiol, the solution was hrdted to 1)oilingarid ethyl acetate was added t o incipient cloudiness. After the solution was chilled there was ohlaiiictl S(i.0 g. (f?~$) of 5-rnethyliiiiidazolirie hydrochloride, ni. p. Isis-150 . After two recrystallizations from ethanol-ethyl acetate the tnaterial nielted a t 171'. .ltta/. Calcd. for CIH8S..HC'I: S ,33.21. Fouiitl: S , 33 .A:;. Thc filtrate front the hydrochloride was takeii t o dryness utitier vacuum aut1 the residue waq refluxed for three llours with concentrated hydrochloric acid (400 cc.). The hydrolysis mixture was taken to dryness under varuuni and the residue was made strongly alkaline with 355 i, aqueoits caustic soda (200 cc.). The resultant oil was separated and dried over anhydrous potassium carhonat c Fractionation gave 16 g. (8.3%)of material, 1). 1). l i i 0 167" (14 mni.), and 1 i . 5 g. (6.2(%) of material, b. p . 347-25io(14mm.).Q The lower boiling fraction gave a picrate rneltiiig at 165'. A r i d . Calcd. lor CIIHIGS20~C6J13X30,: S , 10,CZ. Found: X, 16.46. The lower boiling fraction (1.0 g.) was refluxed for one hour with 10% aqueous caustic soda (20 cc.). The product was extracted with ether, the solvent was removed, and the residue was converted into a dipicrate, 111. p. 216 10 AnaE. Calcd. for C , I H ~ ~ S ~ ~ C ~ H ~ N, X ; ~18.41. O,: FoUlld: s, l8.:
