The product W d b jiltered dud washed with 504; well with Skellysolve B, and dried at 70" overnight t o ivol.) methanol toogive 33.1 g. (61%) of white crystals, remove most of the residual decalin. The base was dis. 104.6-105.5 solved in hot 50% (vol.) ethanol, decolorized with carbon In- C) N-~-(3,4-Diethoxyphenethyl)-3,4-diethoxyphenyl- and allowed t o cool. The solid was iiltered, washed with acetamide.-A mixture of 35.5 g. of 3,44ethoxyphen- dilute alcohol and dried. A second recrystallization gaw ethylamine and 38 g. of 3,4-diethoxyphenylacetic aci: 35.0 g. of a white solid, m. p. 129-130.5°. suspended in 250 ml. of diphenyl ether was heated a t 210 B-Methoxy-7-ethoxy-1-(3 ',4'-dimethoxybenzyl) 4so* 5" for one-half hour. The crude amide was precipi- quinoline Hydrochloride.-Thirty grams of the isoquinotated by adding Skellysolve B t o the cooled reaction mix- line was dissolved in 100 ml. of absolute ethanol and 5 g. ture. Two recrystallimtions from dilute methanol of dry hydrogen chloride was added. The salt was preyielded 49.5 g. (70%) of white crystals, m. p. 106.5-107'. cipitated by the addition of ether. The white crystals All the amides in Table I were treated substantially as in were filtered and recrystallized from 95% ethanol, three the following examples t o obtain the isoquinolines as listed volumes of ether being added t o complete crystallization. in Table 11, and their salts, Table 111. Thirty-two y a m s (9!Yc) of the hydrochloride was oh6-Methoxy-7-ethoxy-1- (3 ',4 '-dimethoxybenzyl) -iso- taincd, m. p . 200-204 . quinoline.-A mixture of 45 g. of W-/3-(3-methoxy-4-ethAcknowledgments.-The authors are indebted oxyphenethyl) -homoveratramide, 200 ml. of thiophene-1 free benzene, and 10 ml. of phosphorus oxychloride was to the following personnel of these laboratories refluxed two and one-half hours. The cooled solution was for their invaluable aid: W, L. Brown, H. L. decomposed with excess dilute sodium hydroxide solution Hunter and W. J. Schenck for microanalyses; and washed with water. The operation from the time of addition of phosphorus oxychloride until beginning of de- K. E. Shipley, IC. K. Chen and R. C. Anderson hydrogenation was carried out in an atmosphere of nitro- for pharmacological evaluation ; and Claudenc, gen. The washed benzene layer was distilled t o removc K. Simrnans for preparative assistance. the last of the water and most of the benzene, at which poim 100 ml. of decalin was added. The distillation of solvent Summary was continued under stirring until the flask content. reached 180". One and one-half grams of 5% palladium Fifteen ethoxy-methoxy homologs of papon carbon suspended in 50 ml. of decalin was added. The dehydrogenation of the dihydroisoquinoline to isoquiiiolitie averine have been prepared and their therapeutic was completed in five hours under total retlnx. The re- value as coronary dilators has been determined. action temperature varied from 184 to 196". The hot A minimum correlation of action and structure has qolution was filtered t o remove the catalyst and Skelly- been found. solve B was eventually added t o complete the precipitation o f the isoquinoline. T h e solid was filtrAred. washed INDIANAPOLIS6, INDIAVA RECEIVEDFEBRUARY 20, 1950 boil
P
.
, ~ O N T R I H T J T I O SPROM rHiE CHEMISTRY DEPARTMEXI OF THE UNIVERSITY OF KANSAS]
Condensation of Acetoacetic Ester with Some Unsymmetrical Epoxides BY KOY
kr
ADAMSL IIND CALVIN
A,
VANDERWERF
N o completely satisfactory explanation of the direction of ring-opening in Sx.:! attack on utisymmetrical epoxides has yet been offered. I t IS becoming increasingly apparentP2however, that steric factors may be mainly responsible for the fact that attack by the nucleophilic agent has generally been found to occur preferentially at the unsubstituted or primary carbon of terminal epoxides. On the other hand, recent evidence" indicates that in the reaction of such epoxides as styrene oxide and 3,4-epoxy-l-butene with certain bases the effect of allylic resonance"' 1x1 lowering the energy of the transition state i n ~ o l v e din nucleophilic attack at the secondari c x b o n atom may outweigh the steric factor.. which favor attack at the primary carbon. In an attempt to shed further light 011 the relationship between steric arid elech-onic factors and the direction of ring opening of unsymmetrical ~ we have studied the epoxides in S N reactions, base catalyzed condensation of acetoacetic ester
with three representative epoxides, propylene oxide, styrene oxide and 3,4-epoxy-l-butene. Choice of acetoacetic ester was dictated by the facts that its reactions with epoxides should certainly be expected to proceed, like those of malonic ester,4 by the S N mechanism, ~ but that steric requirements should be slightly less for acetoacetic than for malonic ester. In addition, the expected products of the condensation reactions themselves should be compounds of considerable chemical interest. Only a few isolated studies of the condensatioii of acetoacetic ester with specific epoxides ha\ c heen reported.s I n each case the product isolated was a substituted a-acetyl-ybutyrolactone, pr(x sumably formed by inner transesterification of the expected 6-hydroxvalkylacetoaceticester.
(1) American Chemical Society Fellow 1046-1947. Geneva College, Beaver Falls, Pennsylvania. ~ 2 )See Brown and Eldred, Tars JOURNAL, 71, 413 (1949) (3) See (a) Bartlett and Ross, sbrd , 70 928 (1048). ib) Swern, liillen and Knight, rbid , TI, I1B2 (1949), ( c ) l'revoy and Brown g ~ . n 71, 1675 t i q m - , r:II.. thrd r i 3 1 ~ 0 19491
(4) Gripby, Hlnd, Chanley end Nestheimer $ b i d , 64, 2606 (1943) ( 5 ) See l'rnobr dud Lelttirrnn BtF ,34, 1971 (1(~01), Kiiunyanti. Cbilintzev and Osetrova. (~0111@6rsnd acaa S C I U K 5 S I 1, d l 1iQ34), CheIiiit\~>,and O x t r o i x J Gen Ciicrn ( I ; 5 5 R ) 7 2371 11937,
Results l'he sole product isolated in the base-catalyzed condensation of the saturated epoxide propylene
Oct., 1950
ACETOACETICESTERWITH
oxide with acetoacetic ester was a-aceto-y-valerolactone, I, indicating that the attack of the anion of acetoacetic ester on the epoxide occurred exclusively a t the primary position. The structure of I was proved by decarboxylation to 5hydroxy-2-hexanone (11)) which was identified as the semicarbazone. The structure of I1 was further confirmed by oxidation to acetonylacetone (111), which was identified as the dioxime and the ?,4-dinitrophenylhydrazone. I was also cleaved in poor vield to y-valerolactone (IV). 0
1 1dok.t 2 HOAc
0
+ CH~-C-CH~-C-OCIH~
--+
--
CHI-CH-CH2
61P0
’O\ CHa-CH-CHz
I
4---
CHz
0
C ‘’
It
5% HCI
0 IV OH
0
/I
CHs-C-CH2-CHz-C-CH3 I11
I1
I
69% 0
I
1
KalCrzO-i, H2SO4
+
11
0 1
C
0
II
CH-C-CHa
16@;
\ / \ /
8
I
0
1. SaOC(CH2) 2 . IICl
/
4369
formed by attack at the secondary carbon, aaceto-P-vinyl-y-butyrolactone (XI), were obtained in what were apparently approximately equal quantities. All attempts to isolate the pure individual lactones by careful fractional distillation were unsuccessful. The mixture was decarboxylated directly to a mixture of the isomeric alcohols 3-hydroxy-6-hepten-2-one(XI1 ) and 5-hydroxy-4-vinyl-2-pentanone (XIII). Recause of the labile nature of these keto-alcohols, no attempt was made t o separate them. In-
11
CHs-CH-CHz
EPOXIDES
S O M E UNSYMMETRICAL
CH~-CH-CH~-CH~-C-CHI I1
61V?
With styrene oxide the product was entirely stead the mixture was reduced, first by a modicy-aceto-y-phenyl- y-butyrolactone (V) formed by fication of the Wolff-Kishner reaction, and then attack of the acetoacetic ester anion at the by catalytic hydrogenation, to give a mixture primary position of the epoxide. The structure from which the isomeric alcohols 3-heptanol of V was established by decarboxylation to 5- (XIV) and 2-ethyl-1-pentanol (XV) were isolated hydroxy-5-phenyl-2-pentanone(VI), which was in approximately equal quantities by fractional oxidized to phenacylacetone (VII). The latter distillation. The structure of XIV was estabwas identified by means of its derivatives with lished by oxidation to 3-heptanone (XVI) which aniline and with p-phenylenediamine. The struc- was identified as the semicarbazone. XV was ture of VI was further confirmed by Wolff-Kishner identified by means of its 3-nitrophthalate, and reduction to 1-phenyl-1-pentanol (VIII) and its structure was confirmed by oxidation to 2oxidation of the latter to valerophenone (IX), ethylpentanoic acid (XVII), which was identified which was identified as the semicarbazone. as the anilide and the p-toluide. In the condensation of acetoacetic ester with Discussion 3,4-epoxy-l-butene, the product resulting from attack at the primary carbon, a-aceto-y-vinyl-yAttack of the acetoacetic ester anion a t the butyrolactone (X), and the isomeric product primary position of propylene oxide is consistent )
0
Cf€Ir-CH-CHz
\/
It
1. KaOEt 2. HOAc
0
I1
+ CH-C-CH2-C-OCzH6
C~HS-CH-CH~
0
I
I
56%
0
II
CII--C--CHs
0
I‘C .’
I/
o v 5% HC1$579:,
OH
QH
I
C6Hr-CH-CHy-CH~CHz-CH3
VI11 KMnO4 in acetone
65%
NH2-NH2, f
KOH
0
I
- CBH6-CH-CH2-CHz-C-CHs
83 %
VI KMnO. in acetone
42%
0
c&-c-cH~-CHZ-C-CH~ I1 VI1
0
It
Ii
is even more completely outweighed,* whereas with alloxide ion3" the steric effects are essentially entirely outweighed. Similarly in the case of 3,4-epoxy-l-butene, the resonance stabilization effect may largely or completely counterbalance the steric cfiects when relatively small bases such as methoxide or alloxide IUIJ u c the attacking agents. Unfortunately, the exact ture of the attacking base in the reduction of epoxides with lithium aluminum hyd r i ~ l is e ~a ~matter of speculation. From the fact that the attack on styrene oxide is dmost entirely, and on 2,4epoxy-1-butene very largel>7, on the primary carbon, however, it would appear that the participating basic i o n i nrc comparable in size to the acetoacetic ester anion. Tt should be emphasiml that, although the steric hindrance explanation here offered seems reasonable and consistent, there is as yet no ennclusive ex-idence for it.
1
XVI
CzHs
with the findings of others6s3"in their studies of S N attack ~ on propylene oxide by other bases. The results of the condensation of acetoacetic ester with styrene oxide and with butadiene oxide, together with those recently reported for other Sx2 reactions involving these epoxides, seem to form a fairly consistent pattern if electronic and steric factors are considered together. Actually the effect of allylic resonance in lowering the energy of the transition state involved in attack at the secondary carbon should be greater in the case of styrene oxide than in the case of 3,4-epoxy-l-butene, With such bulky bases as the anions of malonic ester' and acetoacetic ester, however, steric factors appear to preclude attack at the secondary carbon of styrene oxide. Study of space models suggests that this interpretation is entirely reasonable. Even in the case of the less hindered 3,4-epoxy-l-butene, steric factors still dictate the course of the reaction with malonic ester, but in the reaction with the smaller acetoacetic ester the lesser steric effects are apparently partially overcome by the allylic resonance effect. With a still smaller base such as the phenoxide ion,3dthe steric effect in the case of styrene oxide ~ , 65, 680 (1946) : Rrcvr ( 0 ) Chitwood and Freure, T H IJOURYAI *,14Saddle p.qw ]ire rntcd at the 117th i r i r e r i n ~of t h r 9 m rtc i n held
,it
Ciaicego, April, ' 9 kS r b d 69, I ! f l V & i ,
1 \.ti~ileriTert
Experimental9 Materials.-Propylene oxide, styrene oxide, and 3,4epoxy-1-butene were the purest commercial product available. The last t\\o Yiere distilled through an efficient fractionating column just before use. Absolute ethaiiol was prepared from the commercial prodiwt by the methotl of Manske.10 cr-Aceto-y-valerolactone,I --To 2 1. of dbSOhte ethanol contained in a 5-l., 3-necked flask, cooled in a cold waterbath and equipped with a sealed mechanical stirrer and ji reflux condenser fitted with L: drying tube, 108 g. (4.7 g atoms) of sodium was added in large pieces. The cooled mixture was stirred overnight while the sodium dissolved. The flask was then packed in ice and equipped with 3' calibrated dropping funnel through which 635 ml. (650 g., 5.00 moles) of ethyl acetoacetate was added rapidly. Propylene oxide (350 ml., 290 g . , 5.00 moles), previouslv chilled in a Dry Ice chest, was tiest added dropwise with 5tirring over a period of thirty minutes. The mixture was stirred overnight while it was dlowc.ii to warm to room tempelattire. The alcohol was the11 t,iken off under reduccd pres5ure (less than 100 mm.), tgi trrnperature never being permitted t o rise above 50 . To the sirupy residue there tvas added 30 ml. (315 g., 5.2: moles) of glacial acetic acid and 300 g. of ice. The e x e s acetic acid was neutralized with sodium bicarbonate The supernatant oily layer was separated, the residue extracted with ether, and the combined material was dried over anhydrous sodium sulfate. The ether was removed and the residue distilled from a modified Claisen flask (8) The author's data suggest, however, t h a t a small percentage ot prininry alcohol rrsultirir from :ittnek at the secotidarv c:rrhnrl epoxirtr was formed I : ,i :c,iiIi of S?;I reaction. , ' + I b f d t i n n p o i n t s cori-ectrri, lmilinr p o i l i t 5 *iiicorrecti:ri , !(It hlanske, i b i d . , 63, 1 LOlj t l < 4 3 1 j
tlir
i ~ ti l i r
Oct.. 1950
ACETOACETIC ESTERWITH SOMEUNSYMMETRICAL EPOXIDES
equipped with a 30-cm. Vigreux side arm. After a small forerun of unreacted acetoacetic ester, there was obtained 436 g. (61.4%) of the colorless lactone of 2-acetyl-4hydroxypentanoic acid (a-aceto-y-valerolactone,I ) b. p. 88-9O'at 2 mm. (118-119' a t 8 mm.); nZ6D1.4489; da64 1.1013. Anal.11 Calcd. for C,HloOs: C, 59.1; H, 7.0. Found: C, 58.9; H , 7.0. S-Hydroxy-2-hexanone, 11.-In a 500-ml. flask equipped with a reflux condenser, a mixture of 100 g. (0.70 mole) of I , 50 ml. of 12 N hydrochloric acid and 250 ml. of distilled water was warmed to 70' on a steam-bath and immediately allowed to cool while carbon dioxide was evolved for the next three hours. The mixture was then neutralized and the solvent saturated with potassium carbonate. The supernatant oil was taken off, the aqueous residue extracted with ether, and the combined extracts dried over potassium carbonate. The solvent was removed, and upon distillaticn of the residue through a modified Claisen flask, there was obtained 56 g. (69.0%) of 5-hydroxy-2hexanone, 11, b. p . 79-82' at 16 mm.I2 A sample for analysis and physical properties was taken at 2 mm.I*; b. p. 61Oat 2 mm.; nZ6D1.4312; dZ6a0.9626. Anal. Calcd. for CoH1202: C, 62.1; H , 10.3. Found: C, 62.1; H , 10.3. The semicarbazone melted a t 15~.0-151.5"in agreement with the reported12 value 149-150 . Anal. Calcd. for C7Hlb02N3: N, 24.3. Found: N, 24.7. Oxidation of 11.-In a 500-ml. 3-necked flask packed in ice and equipped with reflux condenser, dropping funnel and mechanical stirrer, 40 g. (0.34 mole) of 11, previously mixed with 50 g. of ice and 75 ml. of 12 N sulfuric acid, was added dropwise to 34 g. (0.11 mole) of sodium dichromate. After the initial strongly exothermic reaction had subsided, the mixture was warmed on a steam-bath for fifteen minutes, then allowed t o stand overnight. The mixture was then neutralized with sodium carbonate and distilled with steam until no more oil separated from saturated potassium carbonate solution. The entire distillate was saturated with potassium carbonate and extracted with ether. The combined extracts were dried over anhydrous sodium sulfate. The solvent was removed and the residue was distilled to give 24.0 g. (61.0%) of 2,5-hexanedione (acetonylacetone, III), b. p. 186-192 ; n * l ~1.4227.14 Dioxime: m. p. 136.2-136.9'; reported12 137". Bis-2,4-dinitrophenylhydrazone:m. p. 255.8-257.0"; reported16 257 '. The residue from the initial steam distillation was acidified with sulfuric acid and then extracted with ether. The combined extracts were dried over anhydrous sodium sulfate. KO high boiling residue remained after removal of the solvent.I6 Acid Cleavage of 1.-Sodium t-butoxide (9.6 g., 0.10 mole) and 28.4 P. 10.20 mole) of I were dissolved in 150 ml. of t-butyl alcohol freshly distilled from sodium. The mixture, protected from atmospheric moisture, was re(11) Microanalyses by Clark Microanalytical Laboratory, Urbana, Illinois. (12) Lipp and Scbeller, Bcr., 42, 1960 (1909), reported the boiling point at 10 mm. as 80-81'. (13) Samples taken at higher pressures and temperatures were all high in carbon and low in hydrogen. Lipp and Scheller" and Wohlgemuth, Comgl. rend., 119, 80 (1914), both reported similar difficulties with 7-ketoalcohols, blaming loss of water through dihydrofuran formation. I t is possible, however, that the ether formed between the two tautomeric modifications of the ketoalcobol may be involved [see Stevens and Stein, THIS JOURNAL, 69, 1045 (1940)l. (14) Gray, J . Chcm. SOL.. 79, 681 (1901),reported n Z 61.4232. ~ (15) Armstrong and Robinson, ibid., 1650 (1934). (16) If any a-aceto-@-methyl-y-butyrolactonehad been formed in the original condensation of acetoacetic ester with propylene oxide, some a-methyllevulinic acid, b. p. 153-156O, might be expected at this point.
437 1
fluxed for twenty-four hours and was then made neutral to litmus with 12 N hydrochloric acid. The precipitated salt was removed by filtration and the filtrate distilled. After removal of the t-butyl alcohol and water from the filtrate by distillation at atmospheric pressure, the residue was distilled in vacuo to give 3.1 g. (16.0%) of the lactone of 4-hydroxypentanoic acid (7-valerolactone, IV), b. p. 51-52' a t 1 mm.; nS6D1.4312. The physical constants agreed with those of an authentic sample of y-valerolactone and the same derivative, the hydrazide of y-hydroxyn-valeric acid, m. p. 64.5-65.Oo,l1 was obtained from both the product and the comparison sample. A small amount (2.0 9.) of I was also recovered in the distillation and considerable non-distillable gummy residue remained. CY-Aceto-7-phenyl-7-butyrolactone,V.-This preparation was carried out by the same general method as was that of I. Styrene oxide (570 ml., 600 g., 5.00 moles) was used in place of propylene oxide. It was not necessary t o pre-chill the styrene oxide, and the reaction flask was kept in a water-bath at room temperature. Benzene was used for extractions instead of ether. Distillation of the final product gave 572 g. (56%) of the colorless lactone of 2-acetyl -4-hydroxy-4-phenylbutanoic acid (CY-aceto-?phenyl-y-butyrolactone, V), b. p. 153-165' a t 2 t o 3.5 mm.; n 2 6 1.5395; ~ dZ6d 1.1772. In several runs there was always some pressure fluctuation during the distillation, indicating some decomposition. The refractive index of the product remained essentially constant throughout the distillation. Anal. Calcd. for C12H1203: C, 70.6; H , 5.9. Found: C, 70.6; H , 5.8. S-Hydroxy-5-phenyl-2-pentanone,VI.-A mixture of 100 g. (0.49 mole) of V dissolved in 200 ml. of absolute ethanol and 100 ml. of 6 N hydrochloric acid contained in a 500-1111. erlenmeyer flask was maintained a t 40-50" on a steam-bath with stirring for twenty-four hours. At the end of this time evolution of carbon dioxide had subsided. The resulting solution was saturated with potassium carbonate, and the alcoholic layer was separated and dried over anhydrous sodium sulfate. After removal of the alcohol, there was obtained 50.0 g. (57%) of pale yellow 5-hydroxy-5-phenyl-2-pentanone,VI, b. p. 127-129 a t 1 mm. Redistillation of this product for the preparation of an analytical sample left more than 3070 of the material in the distillation flask as a clear yellow gum, perhaps the dimeric ether.'* The analytical sample gave the physical constants nsD 1.5311, dZ54 1.1000. Anal. Ca1cd:for CllH1402: C, 74.1; H, 7.9. Found: C, 74.1; H, 7.5. Phenacylacetone, W.-To an ice-cooled solution of 19.7 g. (0.11 mole) of freshly prepared VI dissolved in 100 ml. of acetone there was added slowly with stirring a solution of 20 g. (0.13 mole) of potassium permanganate and 0.5 g. of sodium hydroxide in 100 ml. of acetone and 200 ml. of water. The solution was allowed t o warm t o room temperature overnight and was then acidified with 12 N sulfuric acid and decolorized by means of sulfur dioxide. The acetone was removed by distillation and the residual mixture extracted several times with benzene. The combined extracts were dried over anhydrous sodium sulfate and distilled to yield 8.1 g. (42%) of pale yellow 1phenyl-l,4-pentanedione (phenacylacetone, VII) , b. p. 109-112' at 0.35 mm.; n 3 0 ~1.5250. o u r product, although phenacylacetone is reportoed by HelbergerlD as a bright yellow solid melting a t 29 , could not be induced t o crystallize, nor could an authentic sample prepared by the Friedel-Crafts reaction of levulinyl chloride with benzene. Two derivatives, 2-methyl-l,5-diphenylpyrrole, m. p. 82.5-83.0", and 2-methyl-5;phenyl-l-(4-aminophenyl)-pyrrole, m. p. 136.8-137.1 , prepared by the method of Helberger,lg proved identical with the corresponding derivatives obtained from the authentic sample (17) Darapsky, cl at., J . prakl. Chcm., [2] 147, 146 (19361, reported the value 8 6 O . (18) See Stevens and Stein, THISJOURNAL, 6% 1045 (1940). (19) Helberger, Ann., 6SS, 269 (1936).
of phendCykdCetOne, and the melting point valiics agree
ith those of Helberger. dfurii Acidification of the basic washings with I2 Acid precipitated 2.0 g. of a colorless solid which melted iharply a t 122 and gave no depression in a mixed melting point determination with benzoic acid. The material gave no indication of the phenylhydrazone formation charactexistic of a-phenyllevulinic acid, which might hi vxpected t o appear at this point if any of the origin 11 lactone had been a-aceto-(3-phcnyl-)-butyrolactoiir 1-Phenyl-1-pentmol, VIII.--A %elution of 46.0 g ( 1 tiiole) of freshly distilled \'I in 400 mi. of triethylene glycol \\
