NORMANI€. CROMWELL, Q. T. WILESAND 0. C. SCHROEDEK
2432
Vol. 0 4
employing ethyl phenylacetate with isoquinaldaldehyde; recrystallized from ethyl alcohol as white needles, m. P. 134.5-135.5'; yield, 45%. Anal. Calcd. for C U I H ~ ~ ~ UC, O ~74.72; : H, 5.88; N, 4.36. Found: C, 74.46; H, 5.61; N, 4.47.
nium dioxide to yield isoquinaldaldehyde and 3-methyl-6,7-methylenedioxy-isoquinaldaldehyde, respectively. 2 . Isoquinaldaldehyde has been found to undergo condensation reactions with nitromethSummary ane, acetophenone, phenylacetonitrile and ethyl 1. 1-Methylisoquinoline and 1,3-dimethy1-6,7- phenylacetate. methylenedioxyisoquinoline are oxidized by sele- NEWYoRK, N RECEIVEDJIJL.Y 17, 1912
[ C O N ? RIBUTION FROM THE
AVERYLABORATORY OF CHEMISTRY OF THE UNIVERSITY
OF
NEBRASKA I
Amino Ketones. I. Synthesis of Amino Alcohols and 1,3-Diamino Compounds from @-AminoKetones BY NORMANH. CROMWELL, Q. T. WILES' A N D 0. C. SCHKOEDER?
Although many previous investigations3 have been concerned with the addition of various amines t o cu,P-msaturated ketones, few studies have been made with the resulting p-amino ketones. It seemed of interest to attempt the preparation of certain amino alcohols and the corresponding I ,&&amino compounds of possible pharmacological value from such p-amino ketones by the application of certain known reactions. The present investigation deals with the addition of certain amines to bcnzalacetone and ben/alacetophenonc and the conversion of the products to amino alcohols and 1J-diamino compounds. I t has been found that both morpholine and piperidine add readily to benzalacetoiie t o give, respectively, p-morpholinobenzylacetone (I) and b-piperidinobenzylacetone (11), isolated as the hydrochlorides. The preparation of &amino ketones using high boiling, water soluble amines is best accomplished in water insoluble solvents. This allows the removal of the excess reactant amine by water washing. Conversely, the preparation of 8-amino ketones from water insoluble amines such as aniline is easiest to manipulate in a water soluble solvent such as alcohol. The oximes (111) and (IV) of the amino ketones Shell Development Co , Emeryvrlle, Calif (2) Present address E I du Pont de Nemours and Co , Charleston, W Va (3) (a) Tarnbor and Wildr, Ber 81, 333 (1898), (b) Smith and .4dkins 1 HIS JOURNAL, Bo, 407 (1938), ( c ) Georgi and bchwyzer J prakt Chein 86, 273 (1912), (d) Kohn .ad Ivlorgenstern, M o n a r s h , 84, 773 (190J), 28, 479 (1907), (e) Pollard and Stewart, THIG JC)IIRNAI.. 68, 1980 (1%3f)), (f) 69, 200h 2703 (1937), ( 8 ) Macowkt m d %lbery, I prukt ( l w i n 187, 1 1 1 fLq?L) (h) Tom\ tn(1 Ktrncr r
1 , 3 - D I A M I N O COMPOUNDS
+CeHsCH-CH2COCH3
2433
forms seemed to be present here, only the higher melting ones were isolated. Both of these amino ketoximes were soluble in dilute ?N hydrochloric acid and in dilute potassium hydroxide solutions. These two substances were unstable to heat, especially in acid solutions. / OH Attempts to reduce these amino ketoximes (111) and (IV) (IX) and (X) to the diamines (V) and (VI) using catalytic N a i CzH50H methods were not successful. Using fifteen CaH6-CH-CH2-CH-CHa CeHsCOCli hundred pounds of hydrogen with Raney C~H~CH-CH~-CH-CHS I \I nickel a t 40°,ethanol solutions of (111) and z " I 1 /N 0-COCsH6 (IV) gave only 4-phenyl-2-aminobutane, iso(VI and (VI) >N lated as its hydrochloride, m. p. 142",' and . JCeHsCOCl (XI) and (XII) n morpholine and piperidine, respectively. ReC s H r CH-CHs-CH-CHa odd (no.), -N< = -N 0 sults of the same nature were obtained using I pressures of fifty pounds of hydrogen and \' HN-COCsHb Ranev nickel catalvst. In one exDeriment a even no.),-^< = N -/N(VII) and (VIII) little concd. ammohium hydroxide-was added The benzoates (XI) and (XII), isolated as the to the reduction mixture but still only these decomposition hydrochlorides, were prepared from the corre- products could be isolated. 4-Phenyl-4-morpholino-2-aminobutane and 4-Phenyl-4sponding am no alcohols. piperidino-Z-aminobutane.-The corresponding amino keThe studies of these reactions and products are toxime (10 g.) was dissolved in 80 ml. of ethanol and heated being continued and extended. under reflux. Over a period of two hours 12 g. of sodium was added a piece at a time. Enough ethanol (80 cc.) was Experimental6 added from time t o time t o dissolve the sodium. The p-Morpholino- and p-Rperidino-benzy1acetones.mixture was cooled and 12 ml. of water added to destroy Benzalacetone (10 g., 0.068 mole) dissolved in petroleum the sodium ethoxide. The solution was cooled to 0' and ether (35 ml., b. p. S8-lOO0) was refluxed for fourteen neutralized with dilute hydrochloric acid. The salt was hours with an excess (0.10 mole) of the corresponding filtered off and the alcohol removed by vacuum distillation. amine. The mixtures were then allowed to stand in the To the thick residue, strong sodium hydroxide (50%) was ice chest for two days. Ether was added to dissolve any added to precipitate the free base as an oil which was reprecipitated oil and the mixtures completely extracted with moved by extraction with ether. The ether solution was several portions of water t o remove excess amine. The washed well with saturated salt solution, dried and evapodry ether-petroleum ether solutions were then treated with rated. dry hydrogen chloride to precipitate the white hydroThe products were distilled under vacuum with an effichlorides (I) and (11). The hydrochloride (I) was re- cient pump. I n each of these preparations some decomcrystallized from methanol-dry ether mixtures while the position took place to give 4-phenyl-2-aminobutane. The hydrochloride (11) was recrystallized from ethanol-dry formation of the diamine (VI) was also accompanied by the ether solutions. Both of these hydrochlorides decomposed formation of a very high boiling, glassy-like product which to give benzalacetone when water solutions of them were was soluble in dilute hydrochloric acid but was not identiwarmed. fied. Amino Ketoximes, (111)and (IV).-To a cooled solution Both of the water-clear diamines (V) and (VI) were only ot potassium hydroxide (48 g., 0.84 mole) in methanol (200 slightly soluble in water but were readily soluble in dilute to 300 ml.) hydroxylamine hydrochloride (14.4 g., 0.20 hydrochloric acid. mole) dissolved in water (30 ml.) was added. To this Benzamides.-The benzamides (VII) and (VIII) were solution the corresponding P-amino ketone hydrochloride prepared from the diamines (V) and (VI) by treating (0.044 mole) in methanol (25 ml.) was added. The reac- cooled ether solutions of them with one equivalent of tion mixtures were allowed to stand at room temperature benzoyl chloride. The precipitated hydrochlorides were for two days. The precipitated potassium chloride was hydrolyzed with dilute sodium bicarbonate solutions to then removed and most of the methyl alcohol evaporated give the benzamides. These products were recrystallized in vacuo. The remaining water solutions were cooled in from dilute alcohol solutions. Both of these compounds an ice-bath and slowly neutralized with dilute hydrochloric were soluble in dilute hydrochloric acid solutions. acid. The oily solid which separated was extracted in each 4-Phenyl-4-morpholinobutanol-2and 4-Phenyl-4-pipericase with 50 ml. of ether. The impure products were obdinobutanol-2.-Attempts to reduce the amino ketone tained by evaporation of these solutions. hydrochlorides (I) and (11) with hydrogen and noble metal The products were recrystallized from petroleum ether catalysts gave only very low yields of the desired amino (b. p. 35-40°)-ether solutions. Although two isomeric alcohols. Decomposition, with the loss of morpholine or piperidine, respectively, occurred; the p-piperidinoberizyl(6) Micro Dumas analyses for nitrogen and semi-micro analyses for carbon and hydrogen by the Analytical Laboratory, Department acetone was the least stable.
CsH&-CH=CHCOCH3
f
NH
I
1
I
-
of Chemistry, University of N-brash, under the supervision of H. Armin Pagel.
(7) Harries and Osa, Be*,, 8S, 2997 (lOOa),
XORMAN H. CROMWELL, Q. T. WILESAND 0. C. SCHROEDER
2434
Vol. 64
TABLE I ANALYTICALAKD PHYSICAL DATAFOR AMINO KETONES AND DERIVATIVES 7
Compound
Cpd. no.
Yield,
----CalculatedC H
?&
hf. p . . OC.
67
1.71
79 71
6 37
63 50 57
152 107 178
ti7.71 78.51
8 12 7 14
37 5 p < 1,6>
Summary 1. Two new P-aminobenzylacetones have been prepared and procedures for preparing oximes of @-aminoketones discussed. 2. General methods for preparing ,&amino alcohols and 1,3-diamino compounds from a,@-
[CONTRIBUTION FROM
THE
2435
unsaturated ketones have been investigated. 3. TWOnew amino alcohols with their corresponding benzoates and two new 1,3-diamines with their corresponding benzamides have been preParedLINCOLN, NEBRASKA
RECEIVED MARCH 27, 1942
DIVISION OF CHEMISTRY. NATIONAL INSTITUTE OF HEALTH, U. S. PUBLIC HEALTH SERVICE]
Studies on D-Galactosan < 1,s>p < 1,6> BY RAYMOND M. HANNAND C. S.HUDSON
The hexose anhydride, D-galactosan < 1,5> 8< 1,6>, was first synthesized by Micheel' by the
sis with strong acid, formed the known crystalline 2-methyl-~-ga~actopyanose~; the monomethylaction of hot aqueous barium hydroxide upon monoacetone-D-galactosan, upon selective mild 2,3,4,6 - tetraacetyl - D - galactopyranosido - acid hydrolysis and subsequent methylation, trimethyl-ammonium bromide; the over-all yield formed a crystalline trimethyl-D-galactosan, of the anhydride, based on the crystalline aceto- which, upon complete acid hydrolysis, yielded bromo-D-galactose employed, was 38%. Micheel crystalline 2,3,4-trimethyl-galactopyranosemonoalso reported that he obtained a small amount of hydrate. The ring structure of the parent galacthe same anhydride by the pyrolysis of p-D- tosan is therefore and the 8-congalactose a t temperatures of 270 to 360' and a figuration is assigned to the 1,6 ring, as first propressure of 3 millimeters.2 The < 1,5 > < 1,6> posed by Micheel because the space formula instructure was assigned by Micheel on the 'basis dicates a high probability for this configuration. that acetobromo-D-galactose contains the 1,5 The monoacetone derivative is accordingly 3,4ring and that the 1,6 ring is probably more stable isopropylidene-D-galactosan< 1,5> ,6 < 1,6>, as than the 1,3 and 1,4 rings; the originally inferred by Micheel. Recently we6 have found that the pyrolysis of structure was also the only probable one containing two adjacent hydroxyl groups in the cis- a-lactose monohydrate, under the experimental position, which seemed necessary in order to conditions previously used for the preparation of from vegetable account for the ready formation of a monoacetone D-mannosan < 1,5> /3 < 1,6 > [ a ] 2 0 D -73.3' in ivory, yields a distillate containing both levocompound (m. p. 151-152'; chloroform). These inferences of Micheel were glucosan and D-galactosan < 1,5> fl< 1,6 >. The proved to be correct by the work of McCreath anhydrides are readily separable through the fact and Smith3; they isolated as a by-product in the that the galactosan, but not the glucosan, conpreparation of 1,2 :3,4-diacetone galactose a mono- denses with acetone; average yields of 12.3 g. of acetone galactosan agreeing in physical properties levoglucosan and 20.7 g. of 3,4-isopropylidene-~with the one described by Micheel, and by methyl- galactosan < 1,5> @ < 1,6 > have been obtained by ation they converted it t o a sirupy monomethyl- the pyrolysis of a 200-g. charge of lactose monomonoacetone-D-galactosan, which, upon hydroly- hydrate; economy in price of starting material, simplicity and speed of experimental procedures, (1) Micheel, Bcr., 61, 687 (1929). (2) Ordinary galactose is the u-form; we have shown (THIS and the relatively high yields of the desired prodJOURNAL, 6S, 2241 (1941)) that its pyrolysis gives good yields of Ducts, thus combine to make this procedure of galactosanj3. We find that the yield of levoglucosan (D-glUCOSanj3) is independent of the form of anhydrous pyrolysis an excellent method for obtaining an glucose (u or p) that is pyrolyzed. The melting of such u- or pabundant supply of these two sugar anhydrides forms establishes equilibrium between them, as may be inferred from an old record by C. Tanret (Bull. SOL. chim., [3] 18, 734 (1895)). We a t relatively low cost. In the case of the galachave repeated his experiment under more precise control. The tosan, such a result is of special importance; the glassy melts which were obtained by heating samples of pure u- and 8-D-glucose at 170' for fifteen minutes were cooled and then dissolved galactosan and its acetone derivative are now inin water at 20"; the specific rotations after seven minutes were essenexpensive and readily accessible substances, suittially alike ( + 5 3 and +50), representing the equilibrium rotation of glucose. (3) McCreath and Smith,J . Chcm. SOC.,387 (1939).
(4) Oldham and Bell, THIS JOURNAL, 60, 323 (1938). BS, 1484 (1941).
(5) Hann and Hudson, i b i d . ,
