The Total Synthesis of Lysergic Acid

Journal of the American Chemical Society
Volume 78, pages 3087-3114 (1956)

By Edmund C. Kornfeld, E.J. Fornefeld, G. Bruce Kline, Marjorie J. Mann, Dwight E. Morrison, Reuben G. Jones and R.B. Woodward1

ASCII by Aurelius, HTML by Rhodium

Abstract

Lysergic acid, the basic fragment derived from the ergot alkaloids, has been synthesized in a fifteen-stage sequence beginning with 3-beta-carboxyethylindole. The starting material was converted to the intermediate 1-benzoyl-5-keto-1,2,2a-3,4,5-hexahydrobenz-[cd]-indole (4), containing three of the four rings present in lysergic acid. This ketone in turn was transformed into the tetracyclic compound, 9-keto-7-methyl-4,5,5a,6,6a,7,8,9- octahydroindolo-[4.3-fe]-quinoline (69), and thence to lysergic acid. The synthetic acid was converted to dl-isolysergic acid hydrazide which had previously been resolved and converted to ergonovine. The present work, therefore, completes also the synthesis of this ergot alkaloid.

The striking physiological properties of ergot early directed attention to this remarkable product of the growth of the fungus Claviceps Purpurea on rye grain. Pre-Christian allusions to its effects have been recorded, and it was identified in 1676 as the causative agent of the dreaded medieval gangrenous scourge, St. Anthony's Fire. The therapeutic powers of ergot were likewise recognized during the middle ages. Its capacity to induce uterine contractions was recorded as early as 1582, and crude preparations were introduced into orthodox medicine early in the nineteenth century.2 However, its present important position in medical practice was made possible only by the extensive researches of the past forty years on the isolation and characterization of the pure active principles. These elegant investigations, in which Arthur Stoll has played a dominant role,3 have led to the isolation of no less than six related bases all of which have been shown to be amides of the same key substance, lysergic acid (1).4 Of the natural bases, ergonovine (2) is a particularly simple representative; the others-ergotamine (3, R=H; R'=- CH2-Ph), ergosine (3, R=H; R'=-CH2CHMe2), ergocristine (3, R=Me; R'=- CH2Ph), ergokryptine (3, R=Me; R'=-CH2CHMe2), ergocornine (3, R=Me; R'=Prbeta)-terminate in common in an unusual polypeptide array.5

Widespread interest in the synthesis of lysergic acid was stimulated even before the structural issues were resolved in 1949. This attention was evoked by the unique polycyclic system present in the acid and by the medical applications of the derived natural bases. The synthesis of dihydrolysergic acid by Uhle and Jacobs6 was the first major accomplishment in the synthetic studies, and this was followed by a number of attempts7 directed at lysergic acid itself. More recently, interest in the synthesis has been heightened by the discovery of the startling capacity of the corresponding diethylamide (LSD) to induce abnormal psychic states.8 In this communication, we describe the synthesis of lysergic acid.9

The presence within the lysergic acid skeleton of a beta-substituted indole system and the ready availability of simple beta-substituted indoles suggested such compounds as starting materials for synthetic work. However, the very high reactivity of the hetero-ring of indole compounds seemed incompatible with further extensive synthetic operations. Moreover, we were aware of the possibility that many tricyclic indole intermediates (viz. various benz-[cd]-indoles) would be susceptible to ready and irreversible isomerization to the more stable naphthalenoid isomers.10

In order to circumvent such problems, we adopted at the very outset the artifice of using dihydroindole derivatives.

Synthesis of Tricyclic Ketone (4)

N-benzoyl-3-(beta-carboxyethyl)-dihydroindole (5)11 was converted by thionyl chloride in ether solution to the corresponding acid chloride, and thence directly, by the action of aluminum chloride in carbon disulfide or ethylene dichloride, to the ketone 4. It is of some interest that when the Friedel-Crafts reaction was carried out in benzene solution, the sole product was the phenyl ketone 6. Attempts to effect direct cyclization of the acid 5 to 4 with sulfuric acid or hydrogen fluoride were unsuccessful; with polyphosphoric acid only a very small conversion was obtained. In another attempt to obtain useful tricyclic intermediates, the oxalyl derivatives (7, R=Me or Et) were prepared by condensation of the appropriate oxalic ester with the corresponding ester of 5 in the presence of alkoxides. When these esters were heated with 80% sulfuric acid, hydrolysis and decarboxylation occurred smoothly, with the formation of the alpha-keto acid 8, but no cyclization to a dihydronaphthalene (cf. 9) was observed.12

The tricyclic ketone 4 was hydrolyzed readily by aqueous hydrochloric acid to the free base (10, R=H), which was dehydrogenated by Pd/C in p-cymene to the known 5-keto-1,3,4,5-tetrahydrobenz-[cd]-indole (11).7a,7f On the other hand, 4 itself, and the corresponding N-acetyl ketone (10, R=Ac), which was prepared by either acetylation of (10, R=H), or by aluminum chloride-catalyzed cyclization of the chloride of N-acetyl-3-(beta-carboxyethyl)-dihydroindole, were dehydrogenated under similar conditions to the naphthalenoid compounds (12, R=Bz or Ac). The stabilization of the naphthalene system in the N-acetylated series as a result of the suppression of interaction between the nitrogen atom and the carbonyl group in the ketonic isomers (cf 11, arrows) has been commented upon by Grob.13 These dehydrogenation studies early in the work tended to confirm the soundness of the synthetic approach based on dihydroindole derivatives. In addition they indicated that the conversion to the indole system would have to be carried out using compounds in which the nitrogen function was in the free base form.

Most of the substances described in the present work fell into one or another of several well-defined structural classes, which could be recognized readily by characteristic ultraviolet and infrared spectra, and liberal use was made of this valuable control. The chromophoric systems in 4 gave rise to absorption at 5.91 (ketone carbonyl) and 6.07mu (amide carbonyl) in the infrared, and at 235 and 326mu in the ultraviolet (cf Fig.1).

All of our subsequent synthetic experiments were based upon the tricyclic ketone 4, of an activated methylene group at C-4, suggested that the construction of ring-D of lysergic acid might be initiated by the attachment of the requisite nitrogen atom at the reactive position.

The 4-bromo derivative 13, an obvious intermediate for such studies, was obtained in excellent yield by bromination of the tricyclic ketone with either bromine or pyridine hydrobromide perbromide. However, early attempts to utilize this compound in the alkylation of amines were unpromising. For instance, the reaction of the bromo ketone 13 with methylamine, even at RT took a complicated course and led in fairly good yield to the naphthalene derivative 14. In addition, initial experiments designed to obtain the potentially useful ketal- ketone 16 by alkylation of methylaminoacetone ethylene ketal 15 were likewise unsuccessful. The side chain amine 15 was obtained by reaction of methylamine with either chloro- or bromoacetone ethylene ketal.14 Further discussion of these alkylations and of the ketal-ketone 16 will be deferred until a later section.

Meanwhile, much of the early effort was directed toward the synthesis of the alpha-amino ketone 17. The preparation of the alpha-oximinoketone 18 by condensation of 4 with butyl nitrite in the presence of potassium ethoxide proceeded smoothly, but the desired reaction of 18 could not be brought about. On the other hand, treatment of the oxime of 4 with potassium ethoxide, followed by acid, gave the desired 17, as hydrochloride, in good yield.15 The free base, however, decomposed immediately on liberation from its salt, and in our hands was of little further synthetic utility. Another sequence of reactions which led into the 4- amino series was initiated by condensation of 4 with ethyl formate, or methyl oxalate, in the presence of sodium methoxide. These reactions afforded 19 and 20, respectively.16 When the hydroxymethylene compound 19 was treated with hydrazoic acid in trifluoroacetic acid17 in the presence of sulfuric acid, the major product was the cyano-ketone 21, though the desired 4-formylaminoketone 22 was formed concomitantly in low yield. Similar treatment of the methoxalyl compound 20 gave the oxazole 23 in poor yield; the Schmidt reaction had evidently proceed normally with subsequent cyclodehydration. Hydrolysis of either the ester 23, or the formylamino compound 22, gave the simple aminoketone 24, which like 17 was deemed too sensitive to be useful.

In the hope that 4-alkylamino ketones might be prepared by the action of amines on oxides of the type 25, the ketone 4 was converted into the enol acetate (26, R=Ac) and the enol ether (26, R=Et) by the action of isopropenyl acetate and ethyl orthoformate, respectively. When these enol derivatives were treated with peracids, however, the product isolated was the hydroxyketone 27.18

The simpler oxide 29 was prepared, however, by peroxidation of the olefin 28. The latter in turn was prepared from 4 by reduction to the alcohol (30, R=OH) with sodium borohydride, conversion to the bromide (30, R=Br) with phosphorus tribromide, and dehydrohalogenation with collidine. The olefin 28 contains the basic chromophoric system present in many subsequent compounds; its maximum absorption in the ultraviolet light is at 264mu (Fig. 2, curve A). The corresponding saturated compound 31, is prepared from 28 by hydrogenation, absorbs at 267 and 292mu (Fig. 3, curve A). We further took advantage of having in hand these simple compounds of known structure by replacing the N-benzoyl groups by N-acetyl functions. The resulting amide 32 possesses bands at 241, 254, 307 and 316mu (Fig. 2, curve B), while the bands of the saturated analog 33 are at 213, 253, 279 and 289mu (Fig. 3, curve B).

The oxide 29 was typical of its class in that is was readily converted to a bromohydrin 34 with hydrogen bromide in benzene-ether, and rearranged to the beta-tetralone (35) by magnesium bromide.19

Of greater synthetic interest was the smooth reaction of 29 with amines. With methylamine at 100°C, for example, the simple alkanolamine 36 was produced, while under similar conditions, methylaminoacetone ethylene ketal 15 gave the ketal alcohol 37.20 Numerous attempts to effect the oxidation of 37 to the corresponding aminoketone 16 were unsuccessful. Reaction of 36 with bromoacetone gave a substance which we formulate as the hemiketal 38 in view of the absence of a carbonyl band below 6mu in its infrared spectrum, and its ready conversion to a methyl ether 29 with methanolic hydrogen chloride. The ether 39 was obtained also from 37 by treatment with hydrogen chloride in methanol. Like 37, 38 could not be oxidized to any well- defined product.

Addition of a carbon chain at C-5

An obvious alternative for building ring-D of lysergic acid involved the elaboration of a carbon chain at C-5. The capacity of the carbonyl group in the tricyclic ketone 4 to undergo addition reactions was therefore utilized.

The initial attempt in that direction involved the Reformatsky reaction. When 4 was treated with methyl or ethyl bromoacetate in the presence of zinc, and the resulting crude hydroxy-ester was warmed with formic acid, the unsaturated ester (40, R=Me or Et) was obtained. The beta-gamma-position of the double bond in these esters easily demonstrated by ultraviolet measurements (vide supra). It is of some interest that only the unconjugated isomers were isolated. The acid (40, R=H), obtained by careful alkaline hydrolysis of either ester (40, R= Me or Et), was then converted to the bromoketone 41 in excellent yield through successive treatments with oxalyl chloride in toluene, diazomethane in DCM and aqueous hydrobromic acid. Reduction of the bromoketone with sodium borohydride furnished directly the oxide 42; clearly the medium was sufficiently basic to effect dehydrobromination of the intermediary bromohydrin. Perbenzoic aid was then used to effect addition of an oxygen atom to the isolated double bond of 42, and the dioxide 43 was obtained.

With methylamine at 100°C, the latter yielded an amorphous, tertiary, tetracyclic base, characterized as the crystalline methiodide. There is little doubt that the substance possesses the structure 44, and in view of its close relation to other tetracyclic substances, described below, it seems likely that intensive further investigation should have enabled us to connect this series with our other synthetic routes. However, the low overall yield in the conversion of 4 to 44, which may be attributed in the main to an insufficient opportunity for stereochemical control, and the availability of superior paths, led us not to make the effort.

Another attempt to utilize the Reformatsky ester (40, R=Me) foundered early on a point of sufficient interest to merit description.

Conversion of the corresponding oxide 45 was easily effected by monoperphthalic acid. We hoped that 45 would yield the diester 46 on treatment with sarcosine methyl ester and were encouraged by the observation that methylamine at 100°C gave the lactam 47. However, the changes depicted in (48, arrows) occurred so readily in the experiment with sarcosine ester that the lactone 49 was the sole product.

The Unsaturated Aldehyde 50

The unsaturated aldehyde 50 containing as it does a very reactive carbonyl group, and a point of entry for the introduction of substituents at C-4, occupied a central position in many schemes for the elaboration of ring- D. Furthermore, it was found to be readily preparable from the tricyclic ketone 4, through the corresponding glycidic ester (51, R=Et) obtained from the ketone by treatment with ethyl chloroacetate in the presence of potassium t-butoxide.21 The ester was smoothly hydrolyzed to the sodium salt (51, R=Na), which was converted to the saturated aldehyde 52 with mineral acids, or directly to derivatives of that compound with appropriate carbonyl reagents. Our major interest in the sodium salt, however, was excited by the observation that it gave the semicarbazone of 50 simply and in high yield when it was treated successively, in acetonitrile solution, with pyridine hydrobromide perbromide and semicarbazide.22 The free aldehyde 50 was readily obtained from the derivative by exchange of the semicarbazide residue to pyruvic acid.23

The synthetic opportunities presented by the unsaturated aldehyde were exploited, inter alia, in at least three directions24;

i. With ethylene glycol and toluenesulfonic acid, the aldehyde was converted to the ethylene acetal 53, which was smoothly oxidized by perbenzoic acid to the oxide 54. The latter reacted with methylamine at 120°C to give the base 55, which in turn combined readily with acrylonitrile to give 56. It will not be difficult to imagine the uses to which we wished to put the nitrile 56; but the recalcitrance of the acetal function stood in the way of all of them. With methanolic hydrogen chloride, for example, the ester 57 was formed, while 6N HCl acid gave 58; hot 90% acetic acid removed the beta-cyanoethyl chain, and gave the familiar 55. Attempts to remove the offending function by more brutal means led only to deep-seated changes of no utility.25

ii. The unsaturated nitrile 59 was easily obtained from the aldehyde 50, through treatment of the corresponding oxime with thionyl chloride. Conversion of the 4,5-double bond to an oxide function, by means of alkaline hydrogen peroxide, was accompanied by hydration of the nitrile group, and the epoxyamide 60 was obtained in substantially quantitative yield. Like other 4,5-oxides, 60 was susceptible to attack by amines; with methylaminoacetone ethylene ketal, the base 61 was readily obtained. The remaining task in this series in the conversion of the alpha-hydroxyamide function to a carbonyl group. It remains undone; the most novel of the various results obtained in this effort was the formation of the pentacyclic lactone 62 when the amide was treated with red lead oxide in acetic acid, in an attempt to bring about oxidative cleavage of the group at C-5.

iii. The aldehyde 50 was converted to the epoxy alcohol 64 by either of two methods. In the first, and preferred method, alkaline hydrogen peroxide was used to convert the aldehyde to the corresponding oxide 63, which was then reduced with sodium borohydride. Alternatively, the latter reagent was used to convert 50 to the unsaturated alcohol 65, which was oxidized to 64 by perbenzoic acid. This method suffered from the concomitant formation of a certain amount of the dihydro alcohol 66 in the reduction step. Although this difficulty was overcome by the use of Ponndorf reaction, the oxidation stage was relatively inefficient.

Reaction of the epoxyalcohol 64 with methylaminoacetone ethylene ketal 15 gave, although poor yield, the desired amino glycol 67, which was very smoothly oxidized to the ketone 16 by slightly more than one mole of periodate in acid solution.

The important intermediate ketal-ketone 16 thus became available in an eleven-stage sequence from the tricyclic ketone 4. However, both the length and the inefficiency of this route led us once again to re- examine the possible direct preparation of 16 from the bromoketone 13. In a new series of experiments it was discovered that reaction of 13 with methylaminoacetone ethylene ketal 15 in a non-polar solvent afforded the ketal-ketone 16 in excellent yield. The cumbersome earlier route, therefore, could be abandoned.

The Tetracyclic Series

With the obtention of the ketone 16, the stage was set for entry into the tetracyclic phase of our work. Thus, it may be noted that the intermediate contains, actually or potentially, all of the functions necessary for closure of ring-D, and for attachment of the lysergic acid carboxyl group as well.

The first step was taken with the hydrolysis of 16 to the diketone 68, best effected by treatment with 6N HCl acid.25 The diketone was then smoothly cyclized, by sodium methoxide in absolute ethanol, to the tetracyclic unsaturated ketone 69, which in turn was converted to the protected unsaturated alcohol 70, by successive treatments with acetic anhydride and sodium borohydride, or vice-versa26.

It was now necessary to replace the hydroxyl function in 70 by a carboxyl group. An initial attempt in that direction was based on the observation of Price and Krishnamurti27 that allyl alcohol is easily and directly converted to allyl cyanide by treatment with cuprous cyanide in concentrated hydrochloric acid. However, when the allylic alcohol 70 was treated under similar conditions with the aim of exchanging -OH for - CN, it was transformed simply into the epimeric alcohol 71.28

Nevertheless, the result was encouraging insofar as it suggested that carbonium ion reactions at C-9 were practicable. We next treated the alcohol 70 with liquid hydrogen cyanide in the presence of boron fluoride etherate. Perhaps not without analogy,29 this reaction led to the very smooth production of the formylamino compound 72.

Our attention was then directed to the preparation of the chloride 73, and after a series of experiments in a number of solvents it was found to be readily prepared as its hydrochloride, by the action of thionyl chloride on 70 in liquid sulfur dioxide. The chloride was extraordinarily susceptible to hydrolysis to the alcohol 71, and initial attempts to replace the chlorine atom by a cyano group in hydroxylic solvents were seriously complicated by the formation of the alcohol 71 and corresponding ethers. The difficultly was surmounted by treating the hydrochloride of 73 with excess sodium cyanide, under which conditions the desired nitrile 74 was formed in good yield. Methanolysis of 74 catalyzed by sulfuric acid gave the ester (75, R=Me), which was hydrolyzed by HCl acid or by sodium hydroxide to the corresponding acid (75, R=H).

Two hydrogen atoms, placed by design at C-5 and C-5a to deprive our synthetic operations of the dangers attendant upon the presence of an indole ring, now alone remained to be eliminated. The position of these atoms, as components of a dihydroaromatic system, provided a basis for supposing that they should be relatively readily removed, and dehydrogenation studies were taken in hand. As was the case in our earlier experiences with related N-acyl derivatives (vide supra, 10 to 12), the N-acetyl derivative of the ester (75, R=Me) was converted by Pd/C in boiling xylene to the known naphthalenoid compound 76.30 On the other hand, when the sodium salt of the acid (75, R=H) was treated in boiling water with Raney nickel, disproportionation occurred, with formation of 77. Under these conditions generation of the desired indole system was accomplished, but reduction of the double bond in ring-D took place as well.

Subsequent studies obviated these difficulties; in similar experiments in which heat-deactivated Raney nickel was used as dehydrogenation catalyst in the presence of sodium arsenate,31 dl-lysergic acid (78) was the sole product isolated.

The synthetic dl-lysergic acid was converted through its methyl ester, into dl-isolysergic acid hydrazide.32 The acid and the hydrazide were shown to be identical with samples prepared32,33 from natural materials by comparison of melting points, mixture melting points, infrared and ultraviolet spectra, X-ray powder diagrams, pKa's, and paper chromatographic behavior.

Since dl-isolysergic acid hydrazide has already been resolved and converted to ergonovine 2,34 the present work completes the synthesis of that alkaloid as well as that of lysergic acid.

Acknowledgements:

We wish to express our warmest appreciation to a number of people whose very cordial assistance during the course of the work was to a large measure responsible for its ultimate success: (1) to Drs. H.L. Breunig, A.W. Hubert and associates who provided more than adequate supplies of several of the early intermediates; (2) to Mr. W.L. Brown, Mr. G. Maciak, Mr. H.L. Hunter, Miss G. Beckmann and Mr. L.S. Hatfield who carried out all of the many analyses; (3) to Dr. H.E. Boaz, Mr. J.M. Forbes, Mr. D.O. Woolf, Miss M. Hofmann, Mrs. H. Arndt and Dr. H.A. Rose who recorded the numerous and very useful physical measurements; (4) to Dr. E.R. Shepard for helpful suggestions; and finally (5) to Dr. T.P. Carney for continued patient encouragement.

Experimental

Melting points were determined in soft glass capillary tubes and are uncorrected. Ultraviolet and infrared measurements were used for control purposes throughout the investigation, and spectra of all pure substances prepared were determined. However, in the sequel, spectra are recorded ordinarily only for the substances in the main line of the synthesis and for key model compounds. In each case in the infrared measurements, the abscissa is plotted in wave lengths (2.6-12mu) and the ordinate in percentage transmission (0-100%). The measurements were made on a Beckman IR 2-T automatic recording spetrophotometer. Determinations were run at 0.15 molar concentration in chloroform solution except those designated with an "M" on the curve. In the latter a mineral oil mull was used. Ultraviolet data were obtained in methanol solution using a Cary model 11 automatic recording spectrophotometer.

1-Benzoyl-3-(beta-carboxyethyl)-2,3-dihydroindole (5)

This was prepared by a modification of Blount and Robinson.35 3-Indolepropionic acid, 94.6g (0.5mol), was dissolved in 600ml of water containing 20g of sodium hydroxide. The solution was mixed with about 100g of Raney nickel catalyst and hydrogenated at RT in a steel hydrogenation bomb at 3000-4000 psi. pressure. Reduction was usually complete in 20-30hours, after which the catalyst was filtered and washed with a little water. Concentrated HCl acid, 85ml, was added to the filtrate, and the solution was cooled. If the reduction was incomplete, unreacted indolepropionic acid separated at this point and was removed by filtration. The filtrate was then benzoylated by the usual Schotten-Baumann procedure using 210ml of 12N sodium hydroxide and 180ml of benzoyl chloride. The solution was kept alkaline throughout the benzoylation, and the temperature was kept below 40°C by cooling. When the benzoyl chloride was completely reacted, the mixture was cooled and acidified with 300ml of concentrated HCl acid. The crude product was filtered and washed with water, after which it was extracted with 4x1L portions of hot water to remove the benzoic acid. The hot syrupy product, after decantation of the aqueous extract, was crystallized from a few volumes of methanol; yield 103g (70%), MP: 151-153°C.

1-Benzoyl-5-keto-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole (4)

1-Benzoyl-3-(beta-carboxyethyl)-2,3-dihydroindole, 118g (0.4mol), was mixed with 200ml of pure thionyl chloride. The solution was allowed to stand for ½ hour, after which it was warmed gently for 15-20 minutes on the steam-bath. Excess thionyl chloride was evaporated completely below 30* in vacuo, and the crude acid chloride was dissolved in 200ml of carbon disulfide. The solution of the acid chloride was then added in a thin stream to a vigorously stirred suspension of 240g of aluminum chloride in 1750ml of carbon disulfide contained in a 5L flask (HOOD!!!). A complex separated, and stirring became difficult. The mixture was heated under reflux and stirred for one hour to complete the reaction, after which it was decomposed very carefully by adding 500g of ice, 250ml of conc. HCl acid and 500ml of water. During the decomposition, stirring was maintained, and cooling was effected by periodic distillation of the carbon disulfide in vacuo, and the product was extracted with 2L of benzene. The extract was washed thoroughly with 500ml of 2N sodium hydroxide in three portions and then with water. It was dried over magnesium sulfate and then evaporated to small volume in vacuo. Slow addition of several volumes of ether caused the yellow ketone to crystallize. It was filtered and washed with ether; yield 85.3g (77%), MP: 146-147°C. A sample was recrystallized for analysis from benzene-ether.

1-Benzoyl-5-keto-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole Semicarbazone

A mixture of 1.5g of N-benzoyl-5-keto-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole, 1.5g of semicarbazide hydrochloride and 2.25g of anhydrous sodium acetate in 25ml of ethanol and 10ml of water was warmed under reflux on a steam-bath for 1.25hours. The solution was cooled, and the product was filtered and washed with water, ethanol and ether; yield 1.7g (94%), MP: 260-262°C dec. It was recrystallized from dilute acetic acid.

1-Benzoyl-3-beta-benzoylethyl-2,3-dihydroindole (6)

In the above Friedel-Crafts cyclization procedure, if benzene was used as solvent in place of carbon disulfide, acylation of the benzene took place rather than cyclization. The product was crystallized from ethanol; MP: 101-102°C, yield 52%.

1-Benzoyl-alpha-oxo-3-Indolinebutyric acid (8)

Diazomethane was prepared from 55g of nitrosomethylurea and 150ml of 40% potassium hydroxide in 500ml of ether in the usual way. The ether was dried over solid potassium hydroxide. Methanol (10ml) was added, and 60g of 1-benzoyl-3-[beta-carboxyethyl]-2,3-dihydroindole was then added in portions with shaking and cooling in an ice-bath. When reaction was complete, the solution was concentrated in vacuo below 25°C, and the residue was dissolved in 400ml of ether. The solution was washed well with dilute HCl acid and with 5% aqueous sodium bicarbonate, dried over magnesium sulfate, and concentrated in vacuo. The 1-benzoyl- 3-[beta-carbomethoxyethyl]-2,3-dihydroindole so obtained did not crystallize but was adequately pure for use in the oxidation reaction below; yield 62.5g (99%)

The methyl ester was mixed with 300ml of dry ether, 36g of methyl oxalate and 12.5g of sodium methoxide, and the mixture was heated under reflux and stirred for 22 hours. Ice-water was added with stirring, and the aqueous layer was separated from the neutral ether layer and then acidified with 7ml of conc. sulfuric acid. The oil which separated was extracted with 300ml of ether in two portions, and the extracts were dried over magnesium sulfate. The ether was distilled, and the 1-benzoyl-3-[beta-methoxalyl-beta-carbomethoxyethyl]-2,3- dihydroindole was obtained as a syrup which did not crystallize; yield 44.2g (55%). Unchanged starting material (32g) was recovered from the neutral ether layer above.

The methoxalyl ester was dissolved in 265ml of 77% sulfuric acid and heated on a steam-bath for 20 minutes during which time the temperature was in the range of 68 to 92°C. The solution was poured onto an excess of ice, and the product was filtered, washed with water, and dried in vacuo. It was crystallized from a mixture of benzene and dioxane; yield 18.4g (51%), MP: 160-161°C dec.

5-Keto-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole (10)

1-Benzoyl-5-keto-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole 30g, was mixed with 300ml of conc. HCl acid and 225ml of glacial acetic acid, and the solution was heated under reflux for 16 hours. It was evaporated to dryness in vacuo, and the residue was dissolved in water. The solution was treated with carbon, filtered and excess ammonium hydroxide was added to the filtrate. The crude yellow ketone was filtered and recrystallized from 60 ml of methanol; yield 13.6g (73%), MP: 124-126°C. An analytical sample recrystallized from methanol separated as golden yellow plates, MP: 126-127°C; ultraviolet , max. wavelength= 244mu (17000), 355mu (1900).

1-Acetyl-5-keto-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole

This was prepared like the acetyl derivatives using one gram of the ketone, and 5ml of each of pyridine and butyric anhydride; yield 90%, MP: 137.5-138.5°C after recrystallization from ethanol.

1-Acetyl-3-[beta-carboxyethyl]-2,3-dihydroindole

In a 1L bomb was placed 71.0g (0.375mol) of 3-indolepropionic acid, 15.0g (0.375mol) of sodium hydroxide, seven teaspoonsful (about 71g) of Raney Nickel catalyst, and distilled water to make the volume 450ml. This mixture was hydrogenated for about 20 hours at a pressure 4300psi at RT. The catalyst was filtered, and the filtrate was acidified with 65ml of conc. HCl acid. Unreduced 3-indolepropionic acid, 12.9g, was recovered by filtration. The aqueous acid solution containing 58.6g (0.307mol) of 3-[beta-carboxyethyl]-2,3-dihydroindole was placed in a 3L, 3-necked flask equipped with a mechanical stirrer, two dropping funnels, a thermometer and a cooling bath. To the stirred mixture was added, through one dropping funnel, a solution of 92g (2.3mol) of sodium hydroxide in 250ml of distilled water. When the reaction mixture became basic, simultaneous addition of 94g (0.921mol) of acetic anhydride was begun, with continued stirring and cooling. At the completion of the additions, the reaction mixture was acidified with 100ml of conc. HCl acid, and the mix was refrigerated overnight. The product which was collected on a funnel weighed 77.7g. Recrystallization of the acid from 700ml of methanol gave 45.3g (63%), MP: 157-158°C.

1-Acetyl-5-keto-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole by Cyclization

The acid chloride was prepared in the customary manner from 23.3g of (0.1mol) of 1-acetyl-3-[beta- carboxyethyl]-2,3-dihydroindole and 50ml (0.698mol) of thionyl chloride. After the excess thionyl chloride had been removed in vacuo, the residual acid chloride was dissolved in 75ml of dry nitrobenzene. The solution was then added to 60g (0.452mol) of aluminum chloride suspended in 300ml of dry nitrobenzene while cooling the reaction mix to keep the temperature below 20°C. The mix was stirred about 4 hours at 25°C, after which it was cooled in a chloroform-dry ice bath and decomposed by adding crushed ice, 200ml of distilled water and 100ml of conc. HCl acid. A heavy precipitate was removed by filtration. The nitrobenzene layer was separated and washed 3x50ml portions of 2N sodium hydroxide solution, and then with distilled water until the washings were neutral. The nitrobenzene was steam distilled, and the residual crystalline product was filtered; yield 3.3g (15%), MP: 175-177°C. It was purified by recrystallization from ethanol.

5-Keto-1,3,4,5-tetrahydrobenz-[cd]-indole (11) from 5-Keto-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole (10)

The indoline derivative, 2.0g, was mixed with 3g of 5% Pd/C and 30ml of p-cymene. The mixture was refluxed for one hour, after which, the solvent was removed in vacuo, and the residue was taken up in 200ml of benzene. The solution was filtered and washed with dilute hydrochloric acid and water and dried over magnesium sulfate. On concentration in vacuo to about 10ml, the indole derivative crystallized; yield 0.3g (15%), MP: 157-158°C. Recrystallization from a few ml of benzene gave the pure ketone, MP: 159.5-160.5°C.

The ultraviolet absorption spectra was identical with that reported by Uhle.36.

1-Benzoyl-5-hydroxy-1,2-dihydrobenz-[cd]-indole (12)

A mix of 20g of 1-benzoyl-5-keto-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole and 20g of 5% Pd/C in 350ml of xylene was heated under reflux for 16 hours. The catalyst was filtered and washed with ethyl Cellosolve. The filtrate was concentrated, and the product was filtered; yield 8.45g (42%), MP: 229-231°C dec. Recrystallization from a mix of chloroform and methanol raised the MP to 231-235°C.

1-Benzoyl-5-acetoxy-1,2-dihydrobenz-[cd]-indole

1-Benzoyl-5-hydroxy-1,2-dihydrobenz-[cd]-indole, 6.2g , was dissolved in a mix of 60ml of pyridine and 60ml of acetic anhydride. The solution was warmed on a steam bath for 1 hour, and the solvents were distilled under reduced pressure. The ester was crystallized from ethyl acetate; yield 4.25g (60%), MP: 168-169°C; ultraviolet max wavelength = 233mu (43900), 321mu (16200), 329mu (14900).

1-Acetyl-5-hydroxy-1,2-dihydrobenz-[cd]-indole (12)

A mix containing 2.0g of 1-acetyl-5-keto-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole and 3.0g of 5% Pd/C in 30ml of p-cymene was heated under reflux for one hour. The solvent was distilled in vacuo, and the residue was extracted with hot methanol and chloroform. The extracts were evaporated, and the residue was taken up in chloroform. The product was filtered (0.25g) and recrystallized from a mixture of acetic acid and methanol, MP: 239-247°C dec.

1-Acetyl-5-acetoxy-1,2-dihydrobenz-[cd]-indole

1.0g of the 5-hydroxy compound was dissolved in a mix of 10ml of acetic anhydride and 10ml of pyridine, and the solution was warmed on a steam bath for 0.5hour. The mix was evaporated in vacuo, and the residual product was taken up in water, filtered and recrystallized from acetone; MP: 156-161°C

1-Benzoyl-4-bromo-5-keto-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole (13)

A solution of 304.7g (1.1mol) of 1-benzoyl-5-keto-1,2,2a,3,4,5-hexhydrobenz-[cd]-indole in 2200ml of glacial acetic acid was warmed to 40°C. While the reaction mix was illuminated with a 250 watt bulb, 352g (1.1mol) of pyridine hydrobromide perbromide was added in portions during minutes with shaking. The solution was warmed to 60°C and was kept at 55-60°C for 0.5 hours. The mix was treated with carbon and concentrated in vacuo to small volume. The residue was taken up in 2200ml of chloroform, and the solution was washed several times with water, dried over magnesium sulfate and concentrated in vacuo. The residue was crystallized from 2200ml of 1:1 acetic acid:ether; MP: 180.5-181.5°C, yield 270g (69%). A second crop (31g) of less pure material was obtained by concentrating the filtrates.

1-Benzoyl-4,5-di-[methyamino]-1,2-dihydrobenz-[cd]-indole (14)

1-Benzoyl-4-bromo-5-keto-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole (19.0g) was dissolved in 300ml of liquid methylamine, and the solution was sealed in an autoclave and kept at 25°C for two days. Methylamine was allowed to evaporate, and the residue was crystallized from about 10 ml of ethanol; yield 9.5g (56%), MP: 197-202°C dec. A sample for analysis was recrystallized twice from toluene, MP: 205-209°C dec; ultraviolet max. wavelength 248mu (21000), 278mu (31300), 370mu (4400).

Methylaminoacetone Ethylene Ketal (15)

A mixture of 3718g of chloroacetone ethylene ketal37 and 7150g of liquid methylamine was heated at 155-170°C for 48 hours (pressure 750-950psi). Methylamine was vented, and the residue was mixed with several volumes of ether. A warm solution of 1400g of potassium hydroxide in 650ml of water was then added slowly with thorough agitation, after which excess solid potassium hydroxide was added to form a thick sludge as a bottom layer. The ether solution was decanted and dried over potassium hydroxide to remove all water, and the ether was distilled. The crude product was distilled to yield 3210g, BP: 50-80°C/12mmHg, of a mixture still containing 25-30% unchanged chloroacetone ethylene ketal, which was removed as follows. The crude product was dissolved in 20L of dry ether, and the hydrochloride of the product was precipitated with HCl gas. The salt was filtered and washed thoroughly with ether to remove all chloroketal; MP: 165-167°C. The HCl salt was then suspended in ether and converted to the free base with potassium hydroxide exactly as above. The pure amino ketal had a BP: 160-162°C.

A similar preparation using 1450g of bromoacetone ethylene ketal37 and 2750g of liquid methylamine heated at 162-173°C for 24 hours gave 843g (80%) of the methylaminoacetone ethylene ketal. [green] In this case, the product was halogen-free, and purification through the HCl salt was unnecessary.[/green]

1-Benzoyl-4-isonitroso-5-keto-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole Potassium salt

A mix of 100ml of anhydrous toluene and 6.5ml of absolute ethanol was placed in a flask protected from atmospheric moisture. To this was added 0.9g (0.023atom) of potassium with stirring and warming to speed solution.

At this point, 5.55g (0.02mol) of 1-benzoyl-5-keto-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole, suspended in 75ml of anhydrous toluene was added, and the mix was warmed until the ketone had dissolved. Immediately, the solution was cooled; 5ml (0.0436mol) of butyl nitrite was added, and the reaction mix was stirred for 4 hours at RT, and allowed to stand for 3 days. The solid product obtained by filtration weighed 6.8g (100%) after washing with anhydrous ether, MP: 145-150°C dec. A sample was recrystallized from absolute ethanol, MP: 167-169°C dec.

1-Benzoyl-5-keto-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole oxime

A mix of 41.7g (0.15mol) of 1-benzoyl-5-keto-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole 17.4g (0.25mol) of hydroxylamine HCl, 12.6g (0.0915mol) of anhydrous potassium carbonate, 750ml of methanol and 20 ml of distilled water was stirred and heated for one hour, cooled and placed in the refrigerator for a few days. The product was filtered and washed with ice-cold distilled water. Dilution of the filtrate with water gave additional product. After drying in vacuo at 50°C, the combined product had a MP: 210-211°C, yield 41.6g (95%).

1-Benzoyl-5-isonitroso-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole Tosylate

Dry pyridine (1000ml) and 87g (0.30mol) of the oxime above were cooled in an ice-bath. To the reaction mix was added with stirring a cold solution prepared by dissolving 228g (1.2mol) of p-toluenesulfonyl chloride in 500ml of dry pyridine while cooling in an ice-bath. Addition was at such a rate that the temperature remained at 1-2°C. When the addition was complete, stirring was continued at 0°C for 2 hours, and the reaction mix was poured onto ice, and the product was filtered, washed with water, and recrystallized from ethanol; yield 100.7g (76%), MP: 152-155°C.

4-Amino-1-benzoyl-5-keto-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole HCl salt (17)

Potassium metal, 0.86g. (0.022gram atom), was added to 50ml of absolute ethanol protected from atmospheric moisture, and the mixture was stirred until solution was complete. After the reaction mix was cooled in ice to 10°C, a suspension of 8.8g (0.02mol) of 1-benzoyl-5-isonitroso-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole tosylate in 200ml of absolute ethanol was added, and stirring and cooling were continued for 8 hours, after which the reaction mix was placed in the refrigerator for three days. Unchanged starting material, 3.69g, was recovered by filtration, MP: 156-157°C. The filtrate was extracted with a total of 225ml of 6N HCl acid solution in several portions. The acid extract was concentrated to dryness in vacuo, and the residue was taken up in 250 ml of hot absolute alcohol, from which the amino ketone hydrochloride, MP: 240°C dec., crystallized. Additional product from the mother liquors brought the yield to 2.30g (60%, based on starting material consumed). Material for analysis had a MP: 248-250°C. dec.

The infrared spectrum (mull) had bands at 5.83, 6.08, 6.27, 6.34, 6.60, 6.78 and 7.19mu.

1-Benzoyl-4-formyl-5-keto-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole (19)

Sodium methoxide, 6.0g and 15ml of ethyl formate were added to 150ml cold, dry benzene. The mix was stirred in an ice-bath, and 27.7g of 1-benzoyl-5-keto-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole was added. The reaction mix was stirred in the cold for 15 minutes and then at 25°C for 2 hours. The sodium enolate which separated was filtered and washed with benzene and ether; yield 28.0g (85%). A sample of the salt was dissolved in water, and the solution was acidified with HCl acid. The aqueous solution was decanted, and the gummy product was crystallized from a mixture of DMF/methanol; MP: 142-145°C.

1-Benzoyl-4-formamido-5-keto-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole (22) and 1-Benzoyl-4-cyano-5-keto-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole (21) from 1-Benzoyl-4-formyl-5-keto-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole

The crude sodium enolate of the 4-formyl derivative, 20g was dissolved in 270ml of trifluoroacetic acid, and 4.6g of sodium azide was added with stirring. The flask was fitted with a reflux condensor and thermometer, and 18ml of conc. sulfuric acid was added dropwise with stirring during 12 minutes. The temperature was kept at 45°C during the addition and for an additional 15 minutes. The reaction mixture was concentrated below 30°C in vacuo to a volume of about 75-100mls. Chloroform, 200ml, and 250ml of ice-water were added, and the organic layer was separated and washed twice with water. Acidic material was then extracted from the mix using an excess of cold dilute sodium hydroxide solution. The neutral chloroform fraction was washed with water, dried over magnesium sulfate and the solvent was distilled under reduced pressure. The 4- formamido derivative was crystallized from a little methanol; yield 2.23g (11.4%), MP: 178-180°C. The ultraviolet spectrum was like that in Fig.1

The aqueous sodium hydroxide extract above, containing the acidic fraction, was acidified was HCl acid and extracted with 150ml of chloroform. The chloroform solution was dried; the solvent was distilled, and the 4-cyano compound was crystallized from methanol; yield 5.83g (32%), MP: 190-191°C. A sample was recrystallized from a mix of DMF/methanol, MP: 190-191°C.

1-Benzoyl-5-keto-4-methoxalyl-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole (20)

To a solution of 30g of methyl oxalate in 800ml of cold benzene was added 15g of sodium methoxide powder with stirring and cooling in ice. A warm solution of 55.5g of 1-benzoyl-5-keto-1,2,2a,3,4,5-hexahydrobenz-[cd]- indole in 350ml of toluene was then added dropwise with continued stirring and cooling during 10 minutes. Stirring was maintained with the cooling bath removed for 2 hours, after which 400 ml of ice-water was added. The aqueous solution containing the sodium enolate was separated and acidified with 30ml of conc. HCl acid. The product was extracted with 500ml of chloroform, and the solvent was distilled in vacuo. The residue was crystallized from a few volumes of ethanol; yield 52.3g (72%), MP: 202-204°C dec. A sample was recrystallized from a mix of acetic acid and methanol.

Methyl-4-benzoyl-4,5,5a,6-tetrahydroindole-[4,3fg]-benzoxazole-8-carboxylate (23)

1-Benzoyl-5-keto-4-methoxalyl-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole, 3.63g, was dissolved in 15ml of conc. sulfuric acid, and 1.0g of sodium azide was added with stirring. The mix was warmed to 40°C, at which point the reaction became exothermic, and the temperature rose spontaneously to 50°C. The solution was cooled to 40°C and kept at that temperature for 15 minutes, after which it was poured onto ice. The amorphous product was filtered, washed with water, and crystallized from ethanol; yield 0.69g (19%), MP: 230°C dec. Purification was effected by recrystallization from a mixture of DMF/methanol; MP: 237-238°C dec.

The infrared spectrum had carbonyl bands at 5.75 (ester) and 6.07mu (amide).

Methyl 4,5,5a,6-tetrahydroindolo-[4,3-fg]-benzoxazole-8-carboxylate HCl

A mix of 1.0g of 4-benzoyl derivatives above and 50ml of methanol was saturated with 12g of dry HCl while cooling. The reaction mixture was kept for four days at 25°C, after which time 0.34g (34%) of unchanged starting material was filtered. The filtrate was decolorized with carbon, and evaporated to dryness in vacuo. The residue was crystallized from methanol; yield 0.26g (49%) based on ester not recovered), MP: 226-227°C dec.

4-Benzoyl-4,5,5a,6-tetrahydroindolo-[4,3-fg]-benzoxazole-8-carboxylic Acid Hydrazide

0.5g of the corresponding methyl ester was dissolved in 5.5ml of hot DMF. Pure hydrazine (7ml) was added, and in a few seconds the hydrazide crystallized. The mixture was diluted with an equal volume of methanol, and the product was filtered, 0.43g (86%), MP: 270°C dec. It was insoluble in all the usual solvents.

4-Amino-5-keto-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole Di-HCl (24) A. by Hydrolysis of Methyl-4-benzoyl-4,5,5a,6-tetrahydroindolo-[4,3-fg]-benzoxazole-8-carboxylate

A mix of the oxazole ester, 7.6g, in 400ml of methanol was saturated with dry HCl, and the solution was heated at reflux while a continuous stream od dry HCl was bubbled in during 4.5 hours. The solution was decolorized with carbon, and the solvent and hydrogen chloride were evaporated under reduced pressure. The product was taken up in methanol, filtered and washed with methanol and ether; yield 3.4g (62%), MP: above 300°C.

Hydrolysis of the ester in aqueous hydrochloric acid gave about the same result.

B. by Hydrolysis of 1-Benzoyl-4-amino-5-keto-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole HCl

A solution of 1.11g of the amino ketone HCl in 50ml of conc. HCl acid was heated under reflux overnight. The mix was treated with carbon, and the solvent was distilled in vacuo. The residual product was digested with a little ethanol, filtered and washed with ethanol and ether; yield 0.6g (68%), MP: above 300°C.

X-ray diffraction patterns of the samples prepared by methods A and B were identical.

C. from 1-Benzoyl-4-formamido-5-keto-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole

The hydrolysis in this case was run in HCl acid exactly as above. Once again, the MP: was indistinct in the region of 280-300°C dec., and identity was proved by comparison of X-ray diffraction patterns.

1-Benzoyl-4-acetyl-5-keto-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole

A mixture of 13.8g of 1-Benzoyl-5-keto-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole in 100ml of acetic anhydride was cooled in an ice-bath. Boron trifluoride gas was bubbled into the solution for 20 minutes with stirring and continued cooling, during which time the temperature of the reaction rose to 40°C. The solution was kept at 25°C for 1.5 hours, after which it was conc. to small volume in vacuo. The residue was dissolved in chloroform, and the solution was washed successively with water, 6N sodium hydroxide, concentrated HCl acid and water. It was dried over magnesium sulfate and concentrated under reduced pressure. The residual ketone was crystallized from benzene; yield 6.0g (38%), MP: 172-175°C.

1-Benzoyl-5-acetoxy-1,2,2a,3-tetrahydrobenz-[cd]-indole (26, R=Ac)

1-Benzoyl-5-keto-1,2,2a,3-tetrahydrobenz-[cd]-indole , 100g was added to 500ml of isopropenyl acetate and 2g of p-tosic acid. The flask was fitted with a glass-helix packed column about 12-15 inches long. The reaction mix was heated, and the rate of heating was regulated so that about 250ml distilled during about 26 hours. Isopropenyl acetate was added periodically to replace that which had distilled. The solution was then treated with decolorizing carbon, filtered, and concentrated under reduced pressure until a thick slurry of crystals had deposited. The mixture was cooled and the product was filtered and washed with cold ethyl acetate and ether; yield 74g (65%). A sample was recrystallized from ethyl acetate for analysis, MP: 145- 146°C.

The infrared spectrum had an ester carbonyl band at 5.70mu; ultraviolet max. wavelength= 263mu (21500), 309mu (4200), 319mu (3660).

1-Benzoyl-4-bromo-5-keto-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole (13) from 1-Benzoyl-5-acetoxy- 1,2,2a,3-tetrahydrobenz-[cd]-indole

The enol acetate, 63.9g , was dissolved in 250ml of chloroform, and the solution was cooled in ice. A mixture of 10.5ml of bromine and 50ml of chloroform was then added dropwise during 10-15minutes with stirring and continued cooling. The solution was warmed to about 55°C, after which 50 ml of methanol was added in portions during five minutes to decompose the acetyl bromide. The mix was cooled, washed with water, dried over magnesium sulfate, and the solvent was distilled in vacuo. The bromo ketone was recrystallized from 1:1 acetic acid-ether, yield 64g (90%), MP: 178-180°C.

1-Benzoyl-5-ethoxy-1,2,2a,3-tetrahydrobenz-[cd]-indole (26, R=Et)

A mix of 5.0g of 1-benzoyl-5-keto-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole, 50ml of ethyl orthoformate, 45ml of absolute alcohol and 2 drops of conc. sulfuric acid was heated at reflux for 4 hours. The volatile materials were removed in vacuo, leaving an oil which crystallized on standing. The oil was triturated with ethyl acetate-petroleum ether and allowed to stand overnight in a refrigerator. The crystalline product, 3.0g (55%) was collected on a funnel and washed with petroleum ether. Several crystallizations from ethyl acetate-petroleum ether gave the enol ether, MP: 97.5-102°C.

1-Benzoyl-4-hydroxy-5-keto-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole (27). A. from 1-Benzoyl-5-acetoxy- 1,2,2a,3-tetrahydrobenz-[cd]-indole.

The enol acetate (10g, 0.031mol) was dissolved in 250ml of dry benzene, and to it was added a solution of 7.3g (0.04mol) of monoperphthalic acid in 135 ml of dry ether. The mixture was allowed to stand at RT for three days. The supernatant solution was decanted from the crystalline phthalic acid which had collected on the sides of the flask, and it was washed with 3x50ml portions of saturated sodium bicarbonate solution. The addition of anhydrous magnesium sulfate to the benzene-ether solution caused the product to separate. Several volumes of chloroform were added, and the magnesium sulfate was separated by filtration. The solvents were removed in vacuo leaving 8g (85%) of the crude hydroxy ketone. Crystallization from ethyl acetate gave analytically pure material, MP: 205-206°C.

The infrared spectrum had carbonyl bands at 5.91 (ketone) and 6.07mu (amide); ultraviolet max. wavelength 244mu (23300), 330mu (4400).

B. from 1-Benzoyl-5-ethoxy-1,2,2a,3-tetrahydrobenz-[cd]-indole

A solution of 1.5g (0.0052mol) of the enol ether in 15ml of chloroform containing 0.86g (0.0062mol) of perbenzoic acid. The solution was allowed to stand in the refrigerator overnight, after which it was washed with two 15ml portions of saturated sodium bicarbonate solution, and then with 15ml of water. It was then dried over anhydrous magnesium sulfate, and the chloroform was removed in vacuo leaving an oil, which crystallized on trituration with 1:1 benzene-petroleum ether. Recrystallization from ethyl acetate gave the hydroxy ketone, MP: 205-207°C. A mix MP with a sample of the hydroxy ketone prepared from the enol acetate showed no depression.

1-Benzoyl-5-hydroxy-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole

25g of 1-benzoyl-5-keto-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole was dissolved in 200ml of hot absolute ethanol. The solution was stirred and heated at reflux while a solution of 2.5g of sodium borohydride in 120ml of absolute alcohol was added dropwise during about 0.5 hours. Refluxing was continued for one hour, after which 50ml of 10% aqueous sodium hydroxide was added, and heating was continued for 0.5 hours. The solution was cooled and then poured into 250ml of 6N HCl acid. Most of the ethanol was distilled in vacuo, and the product was extracted from the residue with 3x200ml portions of 1:1 ether-benzene. The extracts were washed with water, treated with carbon and the solvents were removed. The crude alcohol, 20g (80%) , was sufficiently pure to be used in the subsequent reaction. A sample was crystallized from ethyl-acetate- petroleum ether, MP: 182-183°C. In another experiment, the extraction was omitted, and the crude material was simply filtered and the washed with water, cold methanol and ether. The yield of crystalline material was 72%, MP: 182-183°C.

Oxidation of 1-Benzoyl-5-hydroxy-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole to 1-Benzoyl-5-keto- 1,2,2a,3,4,5-hexahydrobenz-[cd]-indole

To a mix of 2.79g of the tricyclic alcohol in 100ml of acetic acid was added with shaking a solution of 0.73gg of chromic acid in 7ml of water and 9ml of acetic acid. The reaction was kept at 25°C for two days, after which it was warmed on a steam-bath for four hours. The solvents were evaporated under reduced pressure, and the residue was dissolved in chloroform and washed with water. The solution was dried over magnesium sulfate; the solvent was distilled, and the ketone was crystallized from ethanol; yield 1.3g (47%), MP: 138-141°C. Recrystallization from a mix of benzene and ether gave ketone with MP: 145-146°C. A mix MP: with authentic ketone was not depressed.

1-Benzoyl-1,2,2a,3-tetrahydrobenz-[cd]-indole (28)

39.5g of crude 1-benzoyl-5-hydroxy-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole was dissolved in 400ml of benzene, and the mix was cooled in ice while 25ml of phosphorus tribromide was added slowly with swirling. The solution was kept overnight at RT and was then boiled gently under reflux for four hours. It was cooled and poured onto ice. The organic layer was separated, and the aqueous layer was washed with a mix of ether and benzene. The combined extracts were washed well with water and 5% sodium carbonate solution, and the solvents were evaporated in vacuo. The residue of 1-benzoyl-5-bromo-1,2,2a,3,4,5-hexahydrobenz-[cd]- indole weighed 36g (74%) and was pure enough for use in the next step.

The bromide was mixed with 150ml of 2,6-lutidine, and the solution was heated at reflux for 4 hours. The mixture was cooled and poured into 400ml of ice-cold 6N HCl acid. The product was extracted with 1:1 ether- benzene, and the extract was washed with aqueous sodium carbonate, dilute HCl and finally with water. The solution was treated with decolorizing carbon, and the solvents were distilled in vacuo. The residual 1-benzoyl-1,2,2a,3-tetrahydrobenz-[cd]-indole was crystallized from benzene-petroleum ether; yield 15.2g (32% based on the tricyclic ketone), MP: 91-95°C. An analytical sample had MP: 95.5-96.5°C.

1-Acetyl-5-hydroxy-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole

1-Acetyl-5-keto-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole, 3.0g was dissolved in 50ml of hot absolute alcohol, and to the solution was added 4ml of 4%sodium borohydride solution in absolute ethanol. The mix was heated under reflux for ten minutes, after which another 4ml of the sodium borohydride solution was added, and heating was continued for 0.5hours. Water, 25ml was added, and the solution was heated for 10 minutes, cooled, acidified with 2ml of conc. HCl acid and then diluted with several volumes of water. The product was extracted with 75ml of chloroform, and the extract was washed with dilute HCl acid and with 5% sodium bicarbonate solution. The solution was dried over magnesium sulfate, decolorized with carbon, and concentrated to small volume. Petroleum ether was added, and the product was filtered; yield 2.7g (89%), MP: 147-149°C. Recrystallization from benzene raised the MP: 150-151°C.

The ultraviolet type was like that in Fig. 3, curve B.

A solution of perbenzoic acid in chloroform was prepared in the usual fashion and standardized against sodium thiosulfate. 24g of 1-benzoyl-1,2,2a,3-tetrahydrobenz-[cd]-indole was added in portions with swirling to a cold solution of perbenzoic acid containing a 20% excess of the oxidizing agent. The solution was allowed to stand for 44 hours at 0°C, after which it was washed several times with 5% sodium hydroxide solution and then with water. It was dried over sodium sulfate, and the solvent was removed in vacuo. The epoxy compound was crystallized from a mixture of ethyl acetate and petroleum ether; yield 20.6g (81%), MP: 95-100°C. Repeated recrystallization from benzene-petroleum ether raised the MP: 104-105°C; ultraviolet max. wavelength 217mu (34500), 265mu (14600), 295mu (9350).

1-Benzoyl-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole (31)

1-Benzoyl-1,2,2a,3-tetrahydrobenz-[cd]-indole, 8.2g, was hydrogenated at 3 atm pressure in 150ml of ethanol using 4.1g of 5% Pd/C catalyst. The catalyst was filtered, the ethanol distilled, and the residue was crystallized from a mixture of benzene and petroleum ether; yield 6.1g (74%), MP: 107-108°C. A sample for analysis was recrystallized from methanol.

The infrared spectrum had bands at 6.11, 6.20, 6.31, 6.85, 7.15, 7.47 and 7.72mu.

1-Acetyl-1,2,2a,3-tetrahydrobenz-[cd]-indole (32)

A solution containing 2.61g of 1-benzoyl-1,2,2a,3-tetrahydrobenz-[cd]-indole and 4.0ml of 50% aqueous sodium hydroxide in 25ml of ethanol was heated under reflux for 2 hours. About half of the alcohol was distilled, and several volumes of water were added. The mixture was extracted with 100 ml of ether, and the extract was washed with water. The ether solution was then extracted with dilute HCl acid, and the acid extract was neutralized with sodium bicarbonate. The crude 1,2,2a,3-tetrahydrobenz-[cd]-indole was extracted with 100ml of ether and the extract was dried over magnesium sulfate and concentrated to a volume of about 25ml. Methanol, 25ml, and 4.0ml of acetic anhydride were added, and the solution was kept at 25°C for 16 hours. The solvents were distilled, and the acetyl derivative was crystallized from a benzene-petroleum ether mix; yield 1.55g (78%), MP: 120.5-121.5°C.

The infrared spectrum had bands at 6.05, 6.13, 6.37, 6.85 and 7.12.

1-Acetyl-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole (33)

A solution of 2.63g of 1-benzoyl-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole in 25ml of ethanol and 5ml of 50% aqueous sodium hydroxide solution was heated under reflux for 2.25 hours. The solution was concentrated to about 10ml, and the residue was diluted with 100ml of ether and 25ml of water. The ether layer was separated and washed with water. It was then extracted with dilute HCl acid to remove basic material, and the extract was separated and neutralized with excess sodium bicarbonate. The crude 1,2,2a,3,4,5-hexahydrobenz-[cd]- indole was extracted with 100ml of ether, and the solution was dried and concentrated to 25ml. Methanol, 25ml, and 5ml of acetic anhydride were added, and the mix was kept at 25°C for 18 hours. Solvents were distilled, and the acetyl derivative was crystallized from an ether-petroleum ether mixture; yield 1.54g (77%), MP: 104-105°C

The infrared spectrum had bands at 6.04, 6.21, 6.28, 6.86 and 7.14mu.

1-Benzoyl-4-bromo-5-hydroxy-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole

A mix of 3.6g of the bromoketone in 25ml of warm dioxane was treated with a solution of 0.28g of sodium borohydride in 6.7ml of absolute alcohol, and the solution was stirred for 1.5 hours. Water, 20ml, was added, the mix was cooled, and the bromohydrin was filtered; yield 2.12g (54%). It was purified by recrystallization from methanol, MP: 87°C dec. When mixed with a sample prepared by method B, below, the MP: was unchanged.

B. from 1-Benzoyl-4,5-epoxy-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole

The epoxide, 5.0g, was dissolved in a mix of 100ml each of benzene and dry ether. Dry HBr was passed into the solution with cooling until 23g had been absorbed. After standing for 5.5hour at 25°C, the mix was concentrated to small volume in vacuo, and the residue was taken up in a mix of chloroform and ether. The bromohydrin was filtered, washed with ether, and recrystallized from a few volumes of methanol. It deposited as crystals containing one molecule of methanol of crystallization; MP: 80°C dec. yield 4.0g (57%).

1-Benzoyl-4-keto-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole(35)

A suspension of 40g of magnesium turnings in 4L of absolute ether was prepared in a 12L, 3-neck flask fitted with a stirrer and reflux condensor. Bromine, 40ml, was added dropwise with stirring while cooling in ice to moderate the reaction. Two layers formed, and the reaction was completed by warming until the mixture was colorless. The solution was decanted from unreacted magnesium into a 22L, 3-neck flask containing 5L of dry ether. To the resulting solution was added with stirring a solution of 20g of 1-benzoyl-4,5-epoxy- 1,2,2a,3,4,5-hexahydrobenz-[cd]-indole in 1.1L of dry benzene. Stirring was continued for 1.5 hours, after which the mixture was allowed to stand for 18 hours at 25°C. The solvent was evaporated to dryness in vacuo. Dry toluene, 3.1, was added to the residue, and the resulting mix was stirred and heated under reflux for 4 hours. It was then cooled and washed with ice-water. The organic layer was dried over magnesium sulfate, and the toluene was distilled under reduced pressure. The ketone was crystallized from a mixture of benzene and ether; yield 11.4g, MP: 147-149°C; second crop, 3.7g, MP: 143-146°C; total, 15.1g (75%). A sample for analysis had a MP: 149.5-151.5°C; a mixture MP: with the isomeric 5-keto compound was 120-141°C. The semicarbazone of the ketone was prepared in the usual way and was obtained as colorless prisms by recrystallization from aqueous acetic acid; MP: 225-226°C.

20g of 1-benzoyl-4,5-epoxy-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole was placed in an autoclave with 200ml of liquid methylamine, and the reaction mixture was heated at 100°C for 16hours. Excess methylamine was evaporated, and the residue was crystallized with benzene and petroleum ether; yield 25g (90%), MP: 93- 95°C. The amino alcohol crystallized from benzene with 1mol of solvent of crystallization.

The ultraviolet type was like Fig. 3, curve A.

1-Benzoyl-4-acetylmethylamino-5-hydroxy-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole

0.02mole of the methylamino alcohol was dissolved in a mix of 12 ml of methanol and 60ml of ether, and 3ml of acetic anhydride was added. The mix was kept at RT for 1 hour, cooled, and the product was filtered and washed with ether; yield 6.0g (86%), MP: 158-160°C.

0.1mol run gave an 89% yield. The ultraviolet type was like Fig. 3, curve A.

4-Methylamino-5-hydroxy-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole

A mix of 1.54g (0.005mol) of 1-benzoyl-5-hydroxy-4-methylamino-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole, 0.5g (0.013mol) of sodium hydroxide, 10ml of ethanol and 5ml of water was heated under reflux for eight hours. The mix was cooled and extracted with benzene. The extract upon concentration gave material of MP: 184-185°C dec. Further purification using the same solvent raised MP: 185.5-186.5°C dec., yield 0.5g (49%).

1-Benzoyl-4-[N-methyl-N-acetonylamino]-5-hydroxy-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole Ethylene Ketal Hydrochloride (37)

0.01mol of 1-benzoyl-4,5-epoxy-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole was mixed with 5.0g of methylaminoacetone ethylene ketal, and the mixture was warmed on the steam-bath for 19 hours. Excess amine was distilled in vacuo, and the product was dissolved in a few volumes of benzene. It was precipitated as a gum by addition of a few volumes of petroleum ether, and the solvents were removed by decantation. The gum was dissolved in 10ml of acetone, and dry hydrogen chloride was passed into the solution. The hydrochloride crystallized and was filtered and washed with cold acetone and ether; yield 81%, MP: 156-158°C dec. It contained one mole of acetone of crystallization. Larger runs gave yields up to 88%. For isolation of the compound as the free base the following procedure was used.

1-Benzoyl-5-hydroxy-4-(N-methyl-N-acetonylamino)-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole Ethylene Ketal (37)

A mixture of 155.5g of 1-benzoyl-4,5-epoxy-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole of 1-benzoyl-4,5-epoxy- 1,2,2a,3,4,5-hexahydrobenz-[cd]-indole and 310ml of methylaminoacetone ethylene ketal was heated on the steam- bath for 17 hours. Excess amine was recovered by distillation under reduced pressure, and the residue was dissolved in a little benzene. Several volumes of petroleum ether were added to precipitate the product, and the supernatant solution was decanted. The residue was crystallized from 200ml of acetone; yield 98.5g (43%), MP: 126-129°C.

The filtrates were treated with dry hydrogen chloride, and the hydrochloride of the product was filtered and recrystallized from a mixture of ethanol and acetone; yield 88g. (31%), MP: 159-160°C dec. The total yield of product as free base and salt was thus 74%.

The sulfuric acid addition salt was prepared from the above free base in ethanol solution and was recrystallized from dilute ethanol, from which it deposited as crystals containing one mole of ethanol crystallization, MP: 184-185°C.

5-Hydroxy-4-(N-methyl-N-acetonylamino)-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole Ethylene Ketal Dihydrochloride

1-Benzoyl-5-hydroxy-4-(N-methyl-N-acetonylamino)-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole ethylene ketal hydrochloride, 15.0g, was dissolved in 250ml of ethanol, and 75ml of an aqueous solution containing 30g of sodium hydroxide was added. The mix was kept at 25°C for 16 hours, after which it was concentrated in vacuo to small volume. The residue was mixed with water and extracted three times with benzene. The extracts were dried over magnesium sulfate, and the hydrochloride was precipitated with dry hydrogen chloride. The solvent was decanted, and the gummy product was crystallized from a mixture of ethanol and acetone; yield 5.2g (40%). A sample was recrystallized from the same solvents, MP: 166-167°C dec. It crystallized with one molecule of acetone of crystallization.

The same debenzoylated ketal-alcohol dihydrochloride was isolated in about 35% yield in an attempt to oxidize the benzoylketal-alcohol by a modified Oppenauer procedure using potassium t-butoxide and benzophenone.38

1-Acetyl-5-hydroxy-4-(N-methyl-N-acetonylamino)-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole Ethylene Ketal

1.0g of the dihydrochloride salt of 5-hydroxy-4-[N-methyl-N-acetonylamino]-1,2,2a,3,4,5-hexahydrobenz-[cd]- indole ethylene ketal was dissolved in 10ml of methanol, and 0.5g of anhydrous sodium acetate, 5ml of ether and 1ml of acetic anhydride were added. The reaction mix was kept at 25°C for three days, after which it was evaporated to dryness in vacuo, and the residue was mixed with excess aqueous sodium bicarbonate. The acetyl derivative was extracted with chloroform, and the extract was dried over magnesium sulfate and concentrated in vacuo. The product was taken up in ether, and filtered; yield 0.705g (89%), MP: 150-151°C. A sample was recrystallized from acetone, MP: 152-153°C.

The ultraviolet type was like Fig. 3, curve B.

4-Benzoyl-4,5,5a,6,6a,8,9,10a-octahydro-7,9-dimethyl-7H-indolo-[3,4-gh]-[1,4]-benzoxazin-9-ol (38)

A. from 1-Benzoyl-5-hydroxy-4-methylamino-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole

A solution of 14.8g of the amino-alcohol, 2.64g of bromoacetone and 125ml of benzene was heated at reflux for one hour. The supernatant liquid was decanted from the insoluble material and washed well with water and aqueous sodium bicarbonate. The benzene solution was dried, concentrated, and cooled and the crystalline hemiketal was filtered; yield 2.8g (40% based on bromoacetone), MP: 153-155°C. A sample was recrystallized from aqueous acetone, MP: 158-160°C.

The ultraviolet type was like Fig. 3, curve A.

B. from 1-Benzoyl-5-hydroxy-4-(N-methyl-N-acetonylamino)-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole Ethylene Ketal

A solution of 0.5g of the ketal in 20ml of 90% acetic acid was warmed on a steam-bath for 4.6 hours. The acetic acid was evaporated under reduced pressure, and the residue was crystallized from a mixture of acetone and ether; yield 40mg (9%), MP: 150-151°C. A mixture MP: determination with a sample prepared by method A showed no depression.

4,5,5a,6,6a,8,9,10a-octahydro-7,9-dimethyl-7H-indolo-[3,4-gh]-[1,4]-benzoxazin-9-ol

A solution of 1.0g of the 1-benzoyl derivative above in 40ml of 6N HCl acid was kept at RT for five days. The mixture was cooled, and 0.19g of benzoic acid was removed by filtration. The filtrate was concentrated in vacuo to small volume, and the residue was dissolved in water and neutralized with sodium bicarbonate. The product was extracted with chloroform, dried with magnesium sulfate and the solvent was distilled. The hemiketal was crystallized from ethanol; yield 0.13g, MP: 120-122°C.

4,5,5a,6,6a,,8,9,10a-Octahydro-7,9-dimethyl-9-methoxy-7H-indolo-[3,4-gh]-[1,4]-benzoxazine (39). A. from 4-Benzoyl-4,5,5a,6,6a,8,9,10a-octahydro-7,9-dimethyl-7H-indolo-[3,4-gh]-[1,4]- benzoxazin-9-ol (38)

1.0g of the hemiketal in 40ml of methanol was saturated with dry hydrogen chloride and kept at 25°C for 16 hours. The mix was evaporated in vacuo, and the residue was washed with ether to remove methyl benzoate. The crude dihydrochloride was then crystallized from a mixture of absolute ethanol and ether; yield 0.3g (31%), MP: 193-195°C dec. An analytical sample recrystallized from the same mixture had a MP: 199-201°C dec.

B. from 1-Benzoyl-5-hydroxy-4-(N-methyl-N-acetonylamino)-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole Ethylene Ketal

1.0g of the ketal alcohol was treated exactly as above with methanol-hydrogen chloride. 150mg (18%) of the dihydrochloride was obtained, MP: 193°C dec. A mix MP: with the sample prepared in method A was not depressed; ultraviolet max. wavelength 245mu (6800), 292mu (2200).

1-Benzoyl-5-carbethoxymethyl-5-hydroxy-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole

In a flask protected from atmospheric moisture were placed, 16.6g (0.06mol) of 1-benzoyl-5-keto-1,2,2a,3,4,5- hexahydrobenz-[cd]-indole, 30g of activated zinc, 300ml of anhydrous benzene and 15ml of ethyl bromoacetate, and the mix was stirred and heated to reflux. A small amount of ethylmagnesium iodide along with a crystal of iodine were added, and this was repeated twice more at 0.5hr intervals. One hour after the last addition, 15g of zinc and 7.5ml of ethyl bromoacetate were added. At the end of another hour, 15g of zinc and a crystal of iodine were added, and the reaction mixture was stirred and heated under reflux for an additional hour, after which it stood at RT.

After 200ml of benzene had been added, the whole was poured into 300g of ice and water, and the mixture was acidified with a solution of 25g of conc. sulfuric acid in 50ml of water. This dissolved all of the precipitate except in unreacted zinc. The benzene layer was separated, and the aqueous layer was extracted twice with benzene. The extracts were combined and washed with water, dilute acetic acid solution, dilute ammonium hydroxide and finally with water, and dried over anhydrous magnesium sulfate. Concentration at reduced pressure gave an oily residue which deposited crystals on standing overnight at RT; yield 7.7g (35%), MP: 136-138°C. Recrystallization from ethyl acetate gave an analytical sample, MP: 142-143°C.

1-Benzoyl-5-carbomethoxymethyl-1,2,2a,3-tetrahydrobenz-[cd]-indole (40, R=Me)

A mix of 300g of 1-benzoyl-5-keto-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole, 300g of activated zinc, 1.0g of mercuric chloride, 6L of dry benzene and 90g of methyl bromoacetate was prepared in a 12L, 3-neck flask fitted with a stirrer, reflux condensor and heating mantle. The mixture was heated under reflux and stirred, and after an induction period of 10-30minutes the reaction started, and the solution became cloudy. After 3.5 hours 90g of methyl bromoacetate and 20g of zinc were added, and after five hours, 75g of the bromoester and 60g of zinc were added. Refluxing and stirring were maintained for a total of six hours, after which the reaction mix was allowed to cool and stand overnight. 2N HCl acid (1500ml) was then added with stirring. The organic layer was separated and washed with 1500ml of 2N HCl acid, once with water, 3x2N ammonium hydroxide, and again with water. The solution was dried over magnesium sulfate and conc. to about 1200ml in vacuo. Petroleum ether was added slowly until the solution became cloudy; the mix was cooled, and the crude ester was filtered. A solution of the crude hydroxy ester in 2250ml of 98% formic acid and 500 ml of acetic anhydride was warmed for one hour on a steam-bath. The dehydration mix was poured into 6L of cold water, and the product was extracted with about 5L of benzene in two portions. The extract was washed several times with water and 5% aqueous sodiumbicarbonate, and the solvent was distilled under reduced pressure. The product was recrystallized twice from methanol using decolorizing carbon. The yield (two crops) was 300.4g (83%), MP: 114-126°C.

1-Benzoyl-5-carboxymethyl-1,2,2a,3-tetrahydrobenz-[cd]-indole (40, R=H)

A mix of 190g of 1-benzoyl-5-carbomethoxymethyl-1,2,2a,3-tetrahydrobenz-[cd]-indole and 1100ml of absolute alcohol was brought to reflux, and with efficient stirring 250ml of 3N sodium hydroxide was added over a period of 3 hours. After heating for an additional 1.5 hours, the mixture was concentrated to about 600ml in vacuo and diluted with 2L of water. After extraction of unreacted starting material with two 250ml portions of chloroform, the aqueous solution was acidified with 75ml of conc. HCl acid, and the product was extracted with 4x250ml portions of chloroform. The combined chloroform extracts were washed with 250ml of water and dried over anhydrous magnesium sulfate.

After removal of the chloroform in vacuo, 550ml of hot ethyl acetate and several grams of decolorizing carbon were added to the residual oil, and the mix was allowed to boil for 20 minutes. The carbon was filtered, and the filtrate was allowed to stand in a refrigerator overnight. The crystalline product was collected and washed with a small quantity of cold ethyl acetate. The yield of acid, MP: 166-170°C dec., was 157g (86%). A second crop of less pure acid raised the yield to 96%. Several recrystallizations from ethyl acetate gave an analytical sample, MP: 168-.5170°C dec.

1-Benzoyl-5-[gamma-bromoacetonyl]-1,2,2a,3-tetrahydrobenz-[cd]-indole (41)

1-Benzoyl-5-carboxymethyl-1,2,2a,3-tetrahydrobenz-[cd]-indole (50g) was suspended in 750ml of dry toluene in a flask protected from moisture. Oxalyl chloride (40ml) and pyridine (1ml) were added, and the mix was allowed to stir at RT until it was homogeneous (2hours). A small amount of insoluble oil was separated from the reaction mix by filtration, and the filtrate was concentrated to dryness in vacuo. The residual oil was dissolved in dry benzene and again concentrated to dryness in vacuo. The crude acid chloride was dissolved in 1L of dry benzene, and the solution was added dropwise over a period of 2 hours to an ice-cold, well-stirred solution of diazomethane (from 75g of nitrosomethylurea) in DCM. The reaction mix was allowed to stir at RT overnight. After adding 500ml of chloroform (to diminish emulsion formation), 300ml of 48% HBr was added over a period of one hour with ice-bath cooing and vigorous stirring. The mix was allowed to stir an additional 3 hours at RT, then the layers were separated. The organic layer was washed with two 500ml portions of water and 500ml of saturated sodium bicarbonate solution. It was dried over anhydrous magnesium sulfate and concentrated in vacuo to yield a tan solid. The yield of crude bromoketone, MP: 127-129°C, was 48g (77%). The best analysis was obtained from material, MP: 128-132°C, recrystallized from benzene-petroleum ether.

1-Benzoyl-4,5-epoxy-5-[beta, gamma-epoxypropyl]-1,2,2a,3,-tetrahydrobenz-[cd]-indole (42)

To a solution of 20g of sodium borohydride in 960ml of methanol and 40 ml of water was added 20g of 1-benzoyl- 5-(gamma-bromoacetonyl)-1,2,2a,3-tetrahydrobenz-[cd]-indole in several portions. When the spontaneous reaction had subsided, the mix was heated under reflux for one hour. It was then poured into several volumes of water, and the white precipitate which separated was extracted with 2x300ml portions of chloroform. The combined chloroform extracts were washed successively with 150ml of 1N HCl acid, 150 ml of 1N sodium hydroxide and water. The solution was dried over magnesium sulfate and concentrated in vacuo. The residual oil was dissolved in a minimum amount of hot ethanol an allowed to stand at 0°C. The yield of epoxide, MP: 108- 111°C, was 9.5g (59%). Several crystallizations from aqueous ethanol gave an analytical sample, MP: 115- 117°C.

1-Benzoyl-4,5-epoxy-5-[beta, gamma-epoxypropyl]-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole (43)

A mix of 9.5g (0.029mol) of product just above and 6.1g (0.0435mol) of perbenzoic acid in 90ml of chloroform was allowed to stand overnight at 0°C. The chloroform solution was washed with dilute sodium hydroxide and water and then dried over anhydrous magnesium sulfate. The chloroform was evaporated in vacuo, leaving an oil which crystallized upon trituration with aqueous ethanol. The yield of crude di-epoxide, MP: 150-170°C, was 4.7g (42%). Crystallization from ethanol gave an analytical sample, MP: 186-187°C, which contained one mole of ethanol of crystallization.

The ultraviolet type was like Fig. 3, curve A.

4-Benzoyl-9,10a-dihydoxy-7-methyl-4,5,5a,6,6a,7,8,9,10,10a-decahydroindolo-[4,3-fg]-quinoline (44)

A mix of 3.8g of the product just above and about 100ml of liquid methylamine was heated at 100°C in an autoclave for 24 hours. The methylamine was allowed to evaporate and the residual oil was dissolved in chloroform and extracted with 2x50ml portions of dilute HCl acid. The combined acid extracts were washed with 2x25ml portions of chloroform. The acid solution was made alkaline with sodium hydroxide, and the basic solution was extracted with 3x40ml portions of chloroform. The chloroform solution was dried over anhydrous magnesium sulfate and concentrated in vacuo leaving 3.3g of crude amorphous tetracyclic di-alcohol, MP: 95- 105°C. Secondary amine impurities were removed by dissolving the product in about 100ml of ice-cold 6N HCl acid and adding slowly and with stirring, 70ml of 20% sodium nitrite solution. The insoluble neutral material which formed was made basic with cold sodium hydroxide solution. The product (free of secondary amine) was extracted with 3x50ml portions of chloroform, and the solution was washed with water and dried over potassium carbonate. The solvent was removed, and the product was triturated with petroleum ether and filtered, MP: 102-110°C. It was not obtained in crystalline form.

The ultraviolet type was like Fig. 3, curve A. Although the product was not analytically pure, it was suitable for conversion to the crystalline methiodide below.

4-Benzoyl-9-,10a-dihydroxy-7,7-dimethyl-4,5,5a,6,6a,7,8,9,10,10a-decahydroindolo-[4,3-fg]- quinolinium Iodide

When a solution of 0.205g (0.00054mol) of the product just above in 10ml of ethyl acetate and two drops of benzyl alcohol was treated with 4.56g (0.032mol) of methyl iodide with warming overnight, the crystalline methiodide was obtained. It was filtered and purified by recrystallization from ethanol-ethyl acetate solution; MP: 200-202°C.

1-Benzoyl-5-carbomethoxymethyl-4,5-epoxy-1,2,2a3,4,5-hexahydrobenz-[cd]-indole (45)

A mix of 3.5g (0.0105mol) of 1-benzoyl-5-carbomethoxymethyl-1,2,2a,3-tetrahydrobenz-[cd]-indole, 30 ml of chloroform, and a solution of 2.2g (0.012mol) of monoperphthalic acid in 37ml of ether was allowed to stand at RT for one week. The supernatant solution was decanted and washed successively with saturated sodium bicarbonate and water. The organic solution was dried over anhydrous magnesium sulfate and concentrated in vacuo to a viscous oil, which solidified upon trituration with benzene. The yield of the epoxide was 2.0g (55%). Crystallization from benzene gave an analytical sample, MP: 181-182°C.

The ultraviolet type was like Fig. 3, curve A.

4-Benzoyl-5,5a,6,6a-tetrahydro-7-methyl-4H-indolo-[6,5,4-cd]-indol-8-(7H)-one (47)

A mix of 1.0g of 1-benzoyl-4,5-epoxy-5-carbomethoxymethyl-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole and a large excess of liquid methylamine (about 100ml) was heated at 100°C for 6 hours in an autoclave. The methylamine was allowed to evaporate, and the residual oil was dissolved in chloroform. The chloroform solution was washed with dilute HCl acid and water and then dried over anhydrous magnesium sulfate. The chloroform was removed in vacuo leaving a viscous oil which, on standing, deposited a crystalline solid. The mix was triturated with ethanol, and the solid was collected on a filter. Several crystallizations from ethanol gave a pure sample of the lactam, MP: 201-202°C.

Infrared carbonyl bands were at 6.00 (lactam) and 6.08mu (amide).

1-Benzoyl-2,2a,3,4-tetrahydro-4-hydroxybenz-[cd]-indole-delta5-(1H), alpha-Acetic Acid Lactone (49)

A mix of 1.0g of 1-benzoyl-4,5-epoxy-5-carbomethoxymethyl-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole and 4ml of sarcosine methyl ester was heated at 100°C for 2.5 hours. Sarcosine ester was distilled in vacuo, and the residual oil was crystallized from aqueous ethanol. The unsaturated lactone crystallized with one molecule of water of crystallization, MP: 160-161°C.

Infrared carbonyl bands were at 5.84 (lactone) and 6.09mu (amide).

Sodium Salt of 1-Benzoyl-5-carboxymethyl-alpha-5-epoxy-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole (51, R=Na)

Potassium metal, 179g, was added to a mix of 3700ml of dry t-butanol and 4900 of dry benzene contained under nitrogen in a 22L rb-flask. After the potassium had dissolved the solution was evaporated to dryness in vacuo, and the cake of potassium t-butoxide was mixed with 3700ml of dry benzene and 4900ml of toluene. The cake was broken up by mechanical stirring, and while keeping the reaction mixture protected continuously under an atmosphere of nitrogen it was cooled to -5°C. Precooled 1-benzoyl-5-keto-1,2,2a,3,4,5-hexahydrobenz-[cd]- indole, 750g, and 593g of ethyl chloroacetate were then added as rapidly as possible. Stirring in the cooling bath was removed, and the mixture was heated to 75°C with stirring during 40-45minutes. At this point, it was cooled rapidly down RT or below. The mixture was then washed twice with ice-water, twice with cold, dilute sulfuric acid, once with water, and twice with aqueous sodium bicarbonate solution. It was dried over magnesium sulfate, and the solvents were distilled under reduced pressure. The syrupy glycidic ester was dissolved in 5900ml of commercial absolute ethanol, and 350ml of 50% aqueous sodium hydroxide was added. The mixture was warmed rapidly to 70-75°C was kept at that temperature for a few minutes. It was then cooled rapidly below RT by evaporation in vacuo without external heating. The sodium salt crystallized, and the mixture was kept at 0°C for 16 hours. The product was filtered and washed well with absolute ethanol and ether; yield 774g (80%). A further crop of less pure material, 109g (11%), was obtained by dilution of the filtrates with a few volumes of dry ether. A sample of the first crop was recrystallized from a mixture of methanol, ethanol and ether; MP: 230-232°C dec.

1-Benzoyl-5-formyl-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole (52)

0.1mol (35.7g) of the sodium salt above was dissolved in 150ml water and mixed with a solution of containing 100g of sodium bisulfite in 200ml of water. The resulting mix was filtered, kept at 25°C for two hours and then cooled. The bisulfite addition product of the aldehyde was filtered and washed with a little ice- water, alcohol and ether; yield 35g (88%), MP: 128-130°C.

The bisulfite product was dissolved in 350ml of water and decomposed by addition of 25ml of acetic acid and 35ml of conc HCl acid. After long cooling 17.0g of the amorphous aldehyde was deposited, MP: 90-100°C dec. It was never obtained in crystalline form.

Similar results were obtained simply by acidifying aqueous solutions of the Darzen's sodium salt.

1-Benzoyl-5-formyl-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole Semicarbazone

A solution containing 5g of the sodium salt of 1-benzoyl-5-carboxymethyl-alpha-5-epoxy-1,2,2a,3,4,5- hexahydrobenz-[cd]-indole, 5g of semicarbazide hydrochloride and 5g of anhydrous sodium acetate in 250ml of 50% aqueous ethanol was heated at reflux for 1.25 hours, and then concentrated under reduced pressure to a volume of 75ml. 5% sodium carbonate solution (100ml) was added, and the semicarbazone was filtered and recrystallized from 200ml of methanol; yield 2.2g (45%), MP: 200-202 dec.; ultraviolet max. wavelength 235mu (18800), 267mu (12100), 293mu (8050).

1-Benzoyl-5-formyl-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole Oxime

A solution containing 17.9g of the sodium salt produced earlier, 21.9g of anhydrous sodium acetate in 270ml of water and 180ml of ethanol was heated at reflux for 2 hours. It was then concentrated to about 250ml and 500ml of cold water was added. The product was filtered, washed with water and crystallized from methanol; yield 8.5g (55%), MP: 168-170°C.

1-Benzoyl-5-acetoxymethylene-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole

A mix of 71.5g of the sodium salt produced earlier and 25g of anhydrous sodium acetate in 800ml of acetic anhydride and 100ml of acetic acid was refluxed for 1 hour and then concentrated to small volume in vacuo. An excess of methanol was added slowly to decompose any remaining acetic anhydride, and the solution was decolorized and again evaporated. Water (2L) was added slowly with shaking, and the product was filtered and then recrystallized from acetic acid; yield 49.6g (74.5%). The compound showed a double MP: 125-130°C and 166-168°C. The higher MP: form could be obtained by recrystallization from ethanol.

The infrared carbonyl bands were at 5.70 (ester) and 6.10mu (amide).

The enol acetate could be prepared also, but less conveniently, from either 1-benzoyl-5-formyl-1,2,2a,3,4,5- tetrahydrobenz-[cd]-indole or its sodium bisulfite addition product.

1-Benzoyl-5-formyl-1,2,2a,3-tetrahydrobenz-[cd]-indole (50)

335g of 1-benzoyl-5-formyl-1,2,2a,3-tetrahydrobenz-[cd]-indole semicarbazone was mixed with 4400ml of chloroform, 1050ml of pure, freshly distilled pyruvic acid and 15ml of water. The solution was kept at 25°C for 18 hours, after which it was diluted with 1L of chloroform and washed 3x1500ml portions of water and once with 1000 5% sodium bicarbonate. The chloroform solution was dried over magnesium sulfate, and the solvent was distilled in vacuo. The crystalline aldehyde was digested with a little hot methanol, and the mixture was cooled, and the product was filtered and washed with methanol and ether; yield 234.3g (83.5%), MP: 179.5- 180.5°C. Similar runs gave yields in the range of 80-89%. A sample for analysis was recrystallized from ethanol.

1-Benzoyl-5-formyl-1,2,2a,3-tetrahydrobenz-[cd]-indole Semicarbazone

10L of acetonitrile in a 22L rb-flask was warmed to 58°C. The solvent was stirred while 500g of the sodium salt of 1-benzoyl-5-carboxymethyl-alpha-5-epoxy-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole and 448g of pyridine perbromide hydrobromide was added. The reaction mix was illuminated with two 250 watt heat lamps and stirred for 12-15minutes, during which time the temperature rose to about 63°C. The illumination was removed, and 468g of semicarbazide hydrochloride and 460g of anhydrous sodium acetate was added, and the mix was digested on a steam-bath for 3 hours with frequent agitation. The solvent was then removed by distillation under reduced pressure, and the residue was mixed with about 12L of water. The crude semicarbazone was filtered was washed with water. It was purified by recrystallization from a mix of acetic acid and methanol; yield 426g (87%), MP: 231-232°C dec, ultraviolet max. wavelength 257mu (24200).

The unsaturated semicarbazone was also prepared by an exactly analogous procedure starting with 1-benzoyl-5- acetoxymethylene-1,2,2a,3,4,5-tetrahydrobenz-[cd]-indole, but the yield was inferior.

1-Benzoyl-5-formyl-1,2,2a,3-tetrahydrobenz-[cd]-indole Phenylhydrazone

This was prepared from the aldehyde and phenylhydrazine in ethanol solution using a little acetic acid as catalyst. It was recrystallized from a benzene-ethanol mixture; MP: 210-212°C.

1-Benzoyl-5-[2'-dioxolanyl]-1,2,2a,3-tetrahydrobenz-[cd]-indole (53)

A mix of 140g of 1-benzoyl-5-formyl-1,2,2a,3-tetrahydrobenz-[cd]-indole, 250ml of ethylene glycol, 480ml of toluene and 0.4g of p-tosic acid, was heated under reflux for 7.5 hours using a water separator (Dean-Stark app.) to collect water formed in the reaction. The reaction mix was washed thoroughly with aqueous sodium bicarbonate, and the aqueous wash was extracted once with chloroform. The chloroform extract was combined with the original toluene solution, and the mix was dried over magnesium sulfate. The solvents were distilled under reduced pressure, and the residual acetal was taken up in methanol, chilled, and filtered; yield 130g (80%), MP: 150-153°C. An analytical sample recrystallized from a mixture of ethyl acetate and petroleum ether melted at 153-155°C.

1-Benzoyl-5-formyl-1,2,2a,3-tetrahydrobenz-[cd]-indole from the product above.

The acetal, 0.5g in 20ml of 90% acetic acid was warmed for 0.5 hour on a steam-bath. Water was added until the mix became cloudy; the solution was cooled, and 0.35g (85%) of the unsaturated aldehyde was filtered, MP: 176-179°C. A mix MP: with authentic aldehyde was unchanged.

1-Benzoyl-5-[2'-dioxolanyl]-4,5-epoxy-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole (54)

1-Benzoyl-5-[2'-dioxolanyl]-1,2,2a,3-tetrahydrobenz-[cd]-indole, 83.5g (0.25mol) was dissolved in a cold solution containing 0.3 mol of perbenzoic acid in 600ml of chloroform. The reaction mix was kept at 0-5°C for 23 hours, after which it was washed twice with 5% aqueous sodium bicarbonate solution. It was then dried with magnesium sulfate, and the solvent was distilled under reduced pressure. The epoxy-acetal was crystallized from methanol; yield 75.5g (87%), MP: 174-176°C. Recrystallization from methanol raised the MP: 178- 180°C.

The ultraviolet type was like Fig. 3, curve A.

1-Benzoyl-5-[2'-dioxolanyl]-5-hydroxy-4-methylamino-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole (55) A. from 1-Benzoyl-5-[2'-dioxolanyl]-4,5-epoxy-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole

A mix of 15g of the epoxy acetal and 500ml of liquid methylamine was heated in an autoclave at 120°C for 14 hours. The methylamine was evaporated completely, and the dark amorphous product was dissolved in 250ml of methanol, and the solution was decolorized with carbon. The filtrate after removal of carbon was then mixed with a hot solution of 9.3g of picric acid in 60ml of ethanol. The picrate salt which crystallized was filtered and washed with acetone; yield 13.2g (51%), MP: 230°C dec. A sample was recrystallized for analysis from a mixture of DMF and methanol; MP: 240°C dec.

The picrate from several runs, 52.0g , was dissolved by shaking in a mix of 500ml of 40% aqueous ethanolamine and 300ml of chloroform. The chloroform was separated, and the aqueous amine layer was washed with 2x200ml portions of chloroform. The combined chloroform extracts were washed three times with aqueous ethanolamine solution, after which they were dried over magnesium sulfate, and the chloroform was distilled. The residual amino alcohol was crystallized from ethyl acetate; yield 23.7g (71%), MP: 150-152°C. A smaller run gave an 80% conversion. A sample after recrystallization from a mix of ethyl acetate and petroleum ether had a MP: 151-153°C. The pKa in 66% DMF was 8.6.

The ultraviolet type was like Fig. 3, curve A.

The rather hygroscopic hydrochloride salt was prepared and crystallized from a mix of ethanol and ethyl acetate; MP: 221°C dec. It gave analysis indicating a monohydrate.

B. from 1-Benzoyl-5-[2'-dioxolanyl]-5-hydroxy-4-[methyl-beta-cyanoethylamino]-1,2,2a,3,4,5- hexahydrobenz-[cd]-indole

The cyanoethyl derivative below, 3.0g in 100ml of 90% acetic acid was heated on a steam-bath for 17 hours. The solution was decolorized with carbon and evaporated in vacuo. The residue was dissolved in chloroform, and the chloroform solution was extracted with dilute HCl acid. The acid extract was neutralized with excess sodium bicarbonate and extracted with chloroform. The solution was dried; the solvent was distilled, and the amino alcohol was crystallized from ethyl acetate; yield 0.4g (15%), MP: 150-152°C. A mix MP: with a sample of the starting material was depressed, but a mix MP: with an authentic sample prepared as above showed no depression.

1-Benzoyl-5-[2'-dioxolanyl]-5-hydroxy-4-[N-methyl-p-toluenesulfonamide]-1,2,2a,3,4,5-hexahydrobenz- [cd]-indole

A solution of 2.0g of 1-benzoyl-5-[2'-dioxolanyl]-5-hydroxy-4-methylamino-1,2,2a,3,4,5-hexahydrobenz-[cd]- indole and 1.2g of p-tosyl chloride in 8ml of dry pyridine was kept at 25°C for 16 hours. It was then poured into water, and the product was extracted with chloroform. The chloroform solution was dried, and the solvent was distilled. The derivative was crystallized from acetone; yield 1.43g (51%), MP: 200-205°C. A sample was recrystallized from a DMF/methanol mix, MP: 204-206°C.

5-[2'-Dioxolanyl]-5-hydroxy-4-methylamino-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole A. by Alkaline Hydrolysis

1-Benzoyl-5-[2'-dioxolanyl]-5-hydroxy-4-methylamino-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole, 1.0g in 20ml of ethanol was treated with 1ml of 50% aqueous sodium hydroxide, and the solution was refluxed from 0.5 hours and then kept at 25°C for 16 hours. The ethanol was distilled under reduced pressure, and the residue was taken up in water and extracted with chloroform. The extracts were dried over magnesium sulfate and evaporated to dryness. The residual amino alcohol was taken up in alcohol and filtered; yield 0.35g (48%), MP: 207- 210°C. It was recrystallized from a mix of DMF and methanol.

The dihydrochloride salt was prepared and crystallized from a methanol-acetone mix, MP: 250°C dec.

B. by Acid Hydrolysis

A solution of 1.0g of the 1-benzoyl-5-hydroxy-4-methylamino-5-acetal in 50ml of methanol containing 1 ml of conc. sulfuric acid was refluxed for 16 hours. Most of the methanol was distilled and the residue was mixed with 20ml of 6N sodium hydroxide. The product was extracted with chloroform in three portions and isolated as above; yield 0.3g (41%), MP: 208-211°C. A mix MP: with a sample obtained by alkaline hydrolysis showed no depression. Hydrolysis of the benzoyl compound with aqueous acid gave similar results.

1-Benzoyl-5-(2'-dioxolanyl)-5-hydroxy-4-(methyl-beta-cyanoethylamino)-1,2,2a,3,4,5-hexahydrobenz-[cd]- indole (56)

A mix of 6.5g of 1-benzoyl-5-(2'-dioxolanyl)-5-hydroxy-4-methylamino-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole and 50ml of acrylonitrile was warmed briefly to 50°C until homogeneous and then kept at 25°C for 16 hours. Excess acrylonitrile was distilled in vacuo, and the residue was crystallized from ethyl acetate; yield 5.6g (88%), MP: 130-132°C. A sample recrystallized from a mix of ethyl acetate and petroleum ether had MP: 135-138°C. The hydrochloride salt recrystallized from an alcohol ether mixture had MP: 184-186°C. dec.

5-(2'-Dioxolanyl)-5-hydroxy-4-(methyl-beta-carbomethoxyethylamino)-1,2,2a,3,4,5-hexahydrobenz-[cd]- indole (57)

20ml of methanol was saturated with dry hydrogen chloride, and 0.5g of 1-benzoyl-5-(2'-dioxolanyl)-5-hydroxy- 4-(methyl-beta-cyanoethylamino)-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole was added. The solution was kept at 24°C for four days, after which it was concentrated in vacuo to dryness. The residue was taken up in chloroform and water, and the excess sodium bicarbonate. The product was extracted with chloroform; the extract was dried, and the solution was distilled. The residual ester was crystallized from ethyl acetate; yield 0.1g (24%), mp: 138-140°C.

The infrared carbonyl band was at 5.76mu (ester).

5-(2'-Dioxolanyl)-5-hydroxy-4-(methyl-beta-cyanoethylamino)-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole (58)

A solution of 2.0g of the 1-benzoyl derivative above in 40 ml of 6N HCl acid was kept at 25°C for four days. Benzoic acid that precipitated from the mix was filtered, and the filtrate was evaporated under reduced pressure. The residue was crystallized from absolute ethanol; yield 1.3g (63%), MP: 173-175°C dec. The dihydrochloride recrystallized from dilute alcohol contained one mole of ethanol of crystallization.

The free base was obtained by dissolving the salt in water and adding excess sodium bicarbonate. It was extracted with chloroform and the solution was dried over magnesium sulfate and concentrated in vacuo. The product was crystallized from benzene; MP: 130-132°C.

The compound had an infrared band at 4.38mu (nitrile).

1-Benzoyl-5-formyl-1,2,2a,3-tetrahydrobenz-[cd]-indole Oxime

A solution containing 20g of 1-benzoyl-5-formyl-1,2,2a,3-tetrahydrobenz-[cd]-indole, 24g of hydroxylamine hydrochloride and 24g of anhydrous sodium acetate in 400ml of alcohol and 300ml of water was heated under reflux with stirring for 5 hours. The mix was cooled overnight, and the product was filtered, and recrystallized from benzene-methanol, MP: 195-197°C dec. yield 12.5g (62%).

1-Benzoyl-5-cyano-1,2,2a,3-tetrahydrobenz-[cd]-indole (59)

To a suspension of 12.5g of 1-benzoyl-5-formyl-1,2,2a,3-tetrahydrobenz-[cd]-indole oxime in 200ml of dry benzene was added 20ml of thionyl chloride dropwise during 15 minutes while stirring and cooling in an ice- bath. Stirring at 0-5°C was continued for 0.5 hours, after which the solvent was distilled in vacuo at RT. The residue was taken up in benzene, and the solution was again concentrated in vacuo. The product crystallized from a benzene-petroleum ether mix; yield 10.7g (91%), MP: 142-144°C.

The infrared spectrum had bands at 4.48, 6.05, 6.15, 6.82, 7.16 and 7.34mu.

1-Benzoyl-5-carbamyl-5-hydroxy-4-methylamino-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole

A mix of 24.5g of 1-benzoyl-5-carbamyl-4,5-epoxy-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole and 650ml of liquid methylamine was sealed in a steel autoclave and heated in a steam-bath for 18 hours. The methylamine was vented, and the residue was digested with a few volumes of hot methanol. The mix was cooled, and the product was filtered and washed with methanol and ether; yield 25.4g (89%), MP: 140-142°C dec. A sample was recrystallized from a mix of DMF and methanol, MP 141-143°C dec. The amide contained one mole of methanol of crystallization.

The solvent-free form was obtained by drying in vacuo and recrystallizing from benzene., MP: 191-193°C.

Carbonyl bands in the infrared were at 5.93 (unsubstituted amide) and 6.08mu (substituted amide).

5-Carbamyl-5-hydroxy-4-methylamino-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole Dihydrochloride

To 100ml of methanol saturated with dry hydrogen chloride was added 1.0g (0.0028mol) of 1-benzoyl-5-carbamyl- 5-hydroxy-4-methylamino-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole, and the mix was allowed to stand for 3 days. Crystals had formed, MP: 223-226°C dec. yield 0.46g (51%). Recrystallization from a methanol-ether mixture gave an analytical sample, MP: 223-226°C dec.

1-Benzoyl-5-carbamyl-4-dimethylamino-5-hydroxy-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole

In a glass autoclave liner was placed 1.0g (0.003mol) of 1-benzoyl-5-carbamyl-4,5-epoxy-1,2,2a,3,4,5- hexahydrobenz-[cd]-indole and 150ml of anhydrous dimethylamine. The liner and contents were placed in the bomb, which was sealed and heated on the steam-bath at 100°C for 16 hours. After the bomb was opened and the excess dimethylamine was evaporated, the residue was recrystallized from benzene to give the product, MP: 204-205.5°C, yield 0.7g (61%).

1-Benzoyl-5-carbamyl-5-hydroxy-4-(N-methyl-N-acetonylamino)-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole Ethylene Ketal (61)

A mix of 0.450g (0.001mol) of 1-benzoyl-5-carbamyl-5-hydroxy-4-(N-methyl-N-acetonylamino)-1,2,2a,3,4,5- hexahydrobenz-[cd]-indole ethylene ketal and 10ml of 100% hydrazine hydrate was heated under reflux for two hours, then kept at RT for one week. The crude crystalline product MP: 253°C. This was raised to 255°C dec. after recrystallization from a mix of DMF and ethanol.

4-Benzoyl-4,5,5a,6,6a,8,9-hexahydro-9-hydroxy-7,9-dimethyl-7H-indolo-[3,4-gh]-[1,4]- benzoxazine-10a-(6aH)-carboxylic Acid Lactone (62)

Treatment of 1.8g (0.004mol) of 1-benzoyl-5-carbamyl-5-hydroxy-4-[N-methyl-N-acetonylamino]-1,2,2a,3,4,5- hexahydrobenz-[cd]-indole ethylene ketal dissolved in 55ml of acetic acid with 2.74g (0.004mol) of red lead oxide at RT for 18 hours gave a homogeneous solution. Glycerol (1ml) and 3.5g of concentrated sulfuric acid were then added. The mix was filtered, and the filtrate was neutralized with sodium bicarbonate. The precipitate which formed was collected and recrystallized from methanol; MP: 249-250°C dec. yield 0.94g (58%).

The ultraviolet spectrum was like that in Fig. 3, curve A. The infrared spectrum had carbonyl bands at 5.57 (lactone) and 6.08mu (amide) and no bands in the OH or NH regions.

1-Benzoyl-5-formyl-4,5-epoxy-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole, 100g, was dissolved in 5L of acetone. The solution was cooled to 5°C while stirring in ice, and 400ml of cold 30% aqueous hydrogen peroxide and 175ml of cold 5% sodium carbonate solution were added. The mix was stirred in ice for 5 hours, after which 2L of ice-water and 10ml of acetic acid were added. The solution was concentrated in vacuo at 20°C to a volume of about 5L and ice-water was added until the solution became cloudy. The mix was cooled for a few days, and the product was filtered, washed well with water, triturated with ether and refiltered; yield 98.6g , MP: 150-153°C dec., second crop, 4.6g; total yield 92%, the compound was a monohydrate. The sample for analysis was recrystallized from methanol.

The anhydrous epoxyaldehyde could be obtained by vacuum drying of the hydrate at 140°C or better as follows: A mix of the hydrate, 0.98g, in 2.5g of ethyl orthoformate and 0.45ml of absolute ethanol containing a trace of sulfuric acid was refluxed for 2.5hours. The solution was cooled, and the aldehyde was filtered and washed with ether; yield 0.5g (54%), MP: 168-171°C. A sample was recrystallized from ethyl acetate, MP: 173- 174°C.

The ultraviolet spectrum was very similar to that in Fig. 3, curve A.

A larger run (100g) gave a 60% yield of the anhydrous form from the hydrate, and the filtrates on evaporation left a syrup which gave a 2% yield of 1-benzoyl-4-hydroxy-5-keto-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole on crystallization from ethanol; MP: and mixed MP: 203-205°C.

1-Benzoyl-5-hydroxymethyl-1,2,2a,3-tetrahydrobenz-[cd]-indole (65)

1-Benzoyl-5-formyl-1,2,2a,3-tetrahydrobenz-[cd]-indole, 76.5g, was dissolved in 350ml of hot dioxane. The solution was then added during 20 minutes to a stirred solution of 7.32g of sodium borohydride in 330ml of absolute ethanol. The warm mixture was then stirred at RT for two hours, after which 800ml of water was added. The product which separated upon cooling was filtered and washed with acetone; yield 64.5g (74%), MP: 108-110°C dec. The compound was a monohydrate.

The ultraviolet type was like that in Fig 2, curve A.

1-Benzoyl-5-hydroxymethyl-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole by Ponndorf Reduction

A solution of 2.89g of 1-benzoyl-5-formyl-1,2,2a,3-tetrahydrobenz-[cd]-indole in 100ml of dry isopropyl alcohol and 50ml of 1.0M aluminum isopropoxide in isopropyl alcohol was refluxed for 3.5 hours with very slow distillation of about 25 ml of the solvent through a 10in. fractionating column. The reaction mix was then concentrated to small volume, and 200ml of chloroform, 100ml of cold water, 15ml of concentrated HCl acid and 20ml of acetic acid were added. The chloroform layer was separated and washed with water and dilute sodium bicarbonate solution, after which it was dried over magnesium sulfate and concentrated under reduced pressure. The residue was taken up in wet ether, and the crystalline alcohol-hydrate was filtered and washed with ether; yield 2.66g (89%). Recrystallization from dilute acetic acid gave pure alcohol with MP: 108-111°C dec. A mix MP: with a sample prepared by sodium borohydride reduction showed no depression.

1-Benzoyl-5-acetoxymethyl-1,2,2a,3-tetrahydrobenz-[cd]-indole

1-Benzoyl-5-hydroxymethyl-1,2,2a,3-tetrahydrobenz-[cd]-indole hydrate (2.0g) was dried at 120°C in vacuo. The resulting amorphous product was dissolved in 10ml of acetic anhydride and treated with six drops of boron trifluoride etherate. The mix was kept at RT for 3 days and was then concentrated in vacuo below 40°C. The product was taken up in chloroform, and the solution was washed with aqueous sodium bicarbonate. The chloroform was removed, and the acetyl derivative was crystallized from ethyl acetate-petroleum ether; MP: 95-98°C, yield 0.8g (37%).

Carbonyl bands in the infrared were 5.74 (ester) and 6.09mu (amide).

1-Benzoyl-5-trifluoroacetoxymethyl-1,2,2a,3-tetrahydrobenz-[cd]-indole

The alcohol-hydrate, 2.0g, was mixed with 1.0g of sodium trifluoroacetate and 40ml of trifluoroacetic anhydride. The mix was refluxed for 2.66 hours, concentrated to dryness in vacuo, and the residue was taken up in a mix of carbon tetrachloride and ether. The insoluble material was filtered, and the solvent was evaporated. The residual product was crystallized from ether, yield 1.05g (42%), MP: 157-158°C dec. and recrystallized from a mix of chloroform and ether, MP: 161-163°C dec.

The ester had carbonyl band in the infrared was at 5.62mu.

1-Benzoyl-5-benzoxymethyl-1,2,2a,3-tetrahydrobenz-[cd]-indole

1.0g of 1-benzoyl-5-hydroxymethyl-1,2,2a,3-tetrahydrobenz-[cd]-indole hydrate was heated in vacuo at 120°C to remove the water of hydration. The dry alcohol was dissolved in 3ml of pyridine and 0.8ml of benzoyl chloride was added. After a few minutes the mix was warmed to 60°C and kept at that temperature for 5 minutes. It was then poured into cold water, and the gummy product was dissolved in chloroform, and the chloroform was dried over magnesium sulfate, and the solvent was distilled. The benzoate ester was crystallized from an ethyl acetate-pet. ether mix; yield 0.8g (63%), MP: 133-135°C.

1-Benzoyl-4,5-epoxy-5-hydroxymethyl-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole (64)

1-Benzoyl-4,5-epoxy-5-formyl-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole hydrate 48.4g (0.15mol), was dissolved in 180ml of hot dioxane. A solution of 4.13g (0.11mol) of sodium borohydride in 180ml of ethanol was then added slowly with stirring. The reaction mix was stirred for 0.5hours, after which time 500ml of water was added. The crude product which separated on cooling had a MP: 167-170°C, yield 34.6g (75%). It contained inorganic impurities which were removed after repeated recrystallization from methanol; MP: 174-177°C.

The ultraviolet spectrum closely resembled that of Fig. 3, curve A.

The epoxy alcohol was also prepared by treatment of 1-benzoyl-5-hydroxymethyl-1,2,2,a,3-tetrahydrobenz-[cd]- indole with perbenzoic acid in chloroform solution; however, the yield was unsatisfactory.

1-Benzoyl-5-acetoxymethyl-4,5-epoxy-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole

1-Benzoyl-5-acetoxymethyl-1,2,2a,3-tetrahydrobenz-[cd]-indole, 8.0g was dissolved in 60ml of chloroform containing 4.14g of perbenzoic acid. The solution was kept at 0-5°C for 16 hours, after which it was washed with aqueous sodium bicarbonate solution and dried over magnesium sulfate. The chloroform was distilled, and the epoxy ester was crystallized from an ethyl acetate-pet. ether mixture; yield 2.03g (24%). A sample for analysis was recrystallized from a mix of ethyl acetate and methanol; MP: 177-179°C.

1-Benzoyl-5-hydroxy-5-hydroxymethyl-4-[N-methyl-N-acetonylamino]-1,2,2a,3,4,5-hexahydrobenz-[cd]- indole Ethylene Ketal (67)

A mix of 12.0g of 1-benzoyl-4,5-epoxy-5-hydroxymethyl-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole and 50ml of methylaminoacetone ethylene ketal was heated under nitrogen in an oil-bath at 125°C for 16 hours. Excess amine was distilled in vacuo, and the residue was taken up in a little benzene. The crude product was precipitated as a gum by addition of pet. ether. The supernatant liquid was decanted, and the gum was taken up in chloroform. The resulting solution was extracted with cold, dilute HCl acid to remove all the basic material, and the acid extracts were neutralized with sodium bicarbonate. The product was extracted with chloroform; the solution was dried over magnesium sulfate, and the solvent was distilled. The residual amino-glycol was crystallized from ethyl acetate; MP: 148-150°C, yield 1.2g (7.0%)

The picrate prepared in methanol had MP: 185-186°C dec.

1-Benzoyl-2,2a,3,4-tetrahydro-4-[methyl-(2-methyl-1,2-dioxolan-2-yl-methyl)-amino]-benz-[cd]-indol-5- (1H)-one (16) by Oxidation of the Glycol (67)

A mix of 0.88g, (0.002mol) of 1-benzoyl-5-hydroxy-5-hydroxymethyl-4-[N-methyl-N-acetonylamino]-1,2,2a,3,4,5- hexahydrobenz-[cd]-indole ethylene ketal and 0.44g (0.0022mol) of sodium periodate in 10ml of water was treated with 0.2ml of concentrated sulfuric acid. The mixture was shaken occasionally during 0.5 hours at 25- 30°C. The solution was neutralized with sodium bicarbonate, and the amorphous product was filtered and washed with water; yield almost theoretical. The crude product was crystallized from acetone, MP: 135- 136°C.

B. by Alkylation with the Bromo Ketone 12

A solution of 270g (0.76mol) of 1-benzoyl-4-bromo-5-keto-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole and 307g (2.35mol) of methylaminoacetone ethylene ketal in 4500ml of dry benzene was refluxed under nitrogen for 21 hours. The mix was cooled, and 151g (93.5%) of methylaminoacetone ethylene ketal hydrobromide was filtered, MP: 158-159°C.

The filtrate was washed several times with ice-water, after which it was extracted with 2.5L of cold, dilute HCl acid containing 150ml of the concentrated acid. The acid extracts were immediately added to an excess of ice-cold dilute sodium hydroxide. The product was extracted with one liter of chloroform, and the chloroform solution was dried over magnesium sulfate, treated with carbon and concentrated in vacuo. The residual ketal-ketone was crystallized from acetone; MP: and mixture MP: 135-136°C, yield 220g (71%).

The ultraviolet curve was like that in Fig. 1.

5-Keto-4-[N-methyl-N-acetonylamino]-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole (68)

20g of 1-benzoyl-2,2a,3,4-tetrahydro-4-[methyl-(2-methyl-1,3-dioxolan-2-yl-methyl)-amino]-benz-[cd]-indol-5- (1H)-one was dissolved in a mixture of 250ml of concentrated HCl acid and 250ml of water, and the solution was kept under nitrogen at 37°C for five days. The mixture was cooled, treated with carbon, filtered and the filtrate was concentrated in vacuo to small volume. The residue was treated with excess sodium bicarbonate; the product was extracted with cold chloroform, and the solvent was removed in vacuo at RT. The crude diketone was powdered, slurried with about 75ml of 1:1 benzene-ether, and filtered; yield 9.8g (77%), MP: 105-107°C. A sample for analysis was recrystallized from benzene-ether or ethanol; MP: 109-110°C; ultraviolet max. wavelength 241mu (17700), 346mu (2200). A monohydrochloride was obtained from dilute ethanol; MP: 200°C dec.

4-Acetyl-4,5,8,9-tetrahydro-7,9-dimethyl-7H-indolo-[3,4-gh]-[1,4]-benzoxazin-9-ol (viii)

5-Keto-4-[N-methyl-N-acetonylamino]-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole, 0.15g was dissolved in 5ml of methanol, and 0.5 ml of acetic anhydride was added. The solution was kept at RT for 2.5 hours and was then diluted with an equal volume of ether. The acetyl derivative which separated 0.075g (43%), had a MP: 160-162°C.

The compound shows an OH band and a single amide carbonyl band in the infrared and must, therefore, be formulated as the cyclic hemiketal. Ultraviolet absorption indicated the naphthalene system: max. wavelength 236mu (20000), 273mu (36500), 279mu (35000), 314mu (7060), 326mu (7500), 357mu (3900), and 363mu (4000).

9-Keto-7-methyl-4,5,5a,6,6a,7,8,9-octahydroindolo-[4,3-fg]-quinoline (69)

25g of 5-Keto-4-[N-methyl-N-acetonyl]-amino-1,2,2a,3,4,5-hexahydrobenz-[cd]-indole was mixed with 550ml of absolute ethanol. The mix was stirred under nitrogen and cooled to -15°C. Sodium methoxide, 16.9g, was then added, and the mixture was stirred at -10°C to -12°C for ten minutes. The reaction mixture was cooled to - 25°C, and the product was filtered on a 6.5-inch buchner funnel and washed with a little cold ethanol and ether. With the very minimum exposure to air (contains sodium methoxide!) the crude ketone was immediately slurried with a little ice-water and refiltered. It was washed with ice-water, ethanol and ether; yield 16.2g (69%), MP: 145-147°C. An analytical sample was recrystallized from dilute ethanol; MP: 155-157°C; ultraviolet max. wavelength 210mu (10000), 266mu (13000), 306mu (20000).

The dihydrochloride was prepared and recrystallized from aqueous acetone; MP: 270°C dec.

The oxime was prepared in dilute ethanol and was recrystallized from a mixture of DMF and ether, MP: 235-236°C dec.

The semicarbazone recrystallized likewise from DMF-ether had MP: 225°C dec.

4-Acetyl-9-keto-7-methyl-4,5,5a,6,6a,7,8,9-octahydroindolo-[4,3-fg]-quinoline

9-Keto-7-methyl-4,5,5a,6,6a,7,8,9-octahydroindolo-[4,3-fg]-quinoline, 24g, was added to 80ml of cold acetic anhydride. The mixture was kept at 25°C for about 5 minutes, after which it was thoroughly cooled, and the product was filtered and washed with ether; yield 20.5g (76%), mp: 167-170°C. A second crop was obtained by evaporation of the filtrate; this raised the total yield to 82%. A sample was recrystallized from acetone-ethanol; MP: 169-170°C; ultraviolet max. wavelength 216mu (6400), 259mu (21000), 301mu (17600); pKa's in 66% DMF = 4.30

The hydrochloride was prepared in ethanol and was recrystallized from aqueous ethanol; MP: 250°C dec.

The oxime was prepared in the usual manner, MP: 250°C dec. after recrystallization from DMF-ether.

The semicarbazone MP: 245-246°C dec, after recrystallization from aqueous ethanol.

9-Hydroxy-7-methyl-4,5,5a,6,6a,7,8,9-octahydroindolo[4,3-fg]-quinoline

10g of 9-Keto-7-methyl-4,5,5a,6,6a,7,8,9-octahydroindolo-[4,3-fg]-quinoline in a mixture of 200ml of methanol and 10ml of water was treated with 1.5g sodium borohydride. The mixture was stirred for 2 hours, after which it was diluted with 150ml of methanol and 25ml of water, heated to boiling, treated with carbon, and concentrated in vacuo. The product was filtered and washed with water and methanol; 8.6g (85%), MP: 210-220°C dec. A sample was recrystallized from alcohol; ultraviolet max. wavelength 242mu (22000), 318mu (1900).

The infrared spectrum (mull) had bands at 3.15, 6.27, 6.86, and 7.03mu.

The dihydrochloride was prepared in ethanol and was recrystallized from aqueous ethanol; MP: 242-243°C dec.

Acetylation of the free base in alcohol solution gave the 4-acetyl derivative described as below.

4-Acetyl-9-hydroxy-7-methyl-4,5,5a,6,6a,7,8,9-octahydroindolo-[4,3-fg]-quinoline (70)

10g of 4-acetyl-9-keto-7-methyl-4,5,5a,6,6a,7,8,9-octahydroxindolo-[4,3-fg]-quinoline was added to a mixture of 150ml of methanol and 10ml of water. Sodium borohydride, 1.5g was added, and the reaction was allowed to proceed at RT to small volume, and a mix of 15ml of conc. HCl acid and 60ml of water was added. The hydrochloride which separated on cooling was filtered and washed with methanol, 9.0g (79%). A sample was recrystallized from dilute ethanol; MP: 245-246°C dec.

The free base was obtained by neutralization of the hydrochloride with aqueous sodium bicarbonate. It was extracted with chloroform; the solution was dried over magnesium sulfate, and the chloroform was distilled. The crude product was crystallized from ethyl acetate; MP: 182-184°C; ultraviolet max. wavelength 243mu (33400), 251mu (38700), 306mu (3500), 316mu (3000); pKa's in 66% DMF = 6.02.

The hydrobromide was prepared and recrystallized from dilute ethanol; MP: 243-244°C.

The acetate ester of the alcohol was prepared using excess acetic anhydride (15ml) with 0.5g of the alcohol at 25°C for 12 hours. The acetic anhydride was distilled in vacuo, and the hydrochloride of the acetate was prepared in methanol and recrystallized from dilute ethanol; MP: 186-187°C dec.

The methiodide of the unsaturated alcohol was obtained using 1.5 parts of methyl iodide in 1:1 nitromethane: methanol as solvent. It was recrystallized from water, MP: 257-258°C dec.

The methochloride was obtained by treating the methiodide with silver acetate, filtering the silver iodide, and the adding hydrochloric acid to the filtrate. It was crystallized from a mix of methanol and ethyl acetate; MP: 240-241°C dec.

4-Acetyl-4,5,5a,6-tetrahydro-9-hydroxy-7-methylindolo-[4,3-fg]-quinolinium Hydroxide Betaine (ix)

A mix of 1.0g of 4-acetyl-9-keto-7-methyl-4,5,5a,6,6a,7,8,9-octahydroindolo[4,3-fg]-quinolinium, 10g of 5% Pd/C, and 35ml of xylene was heated under reflux for four hours. The catalyst was filtered and extracted with hot methanol and chloroform. The combined filtrates were evaporated under reduced pressure, and the residue was recrystallized from water; yield 0.61g (57%), MP: 255-256°C dec. The compound was a monohydrate; ultraviolet max. wavelength 246mu (29000), 351mu (6900); pKa's in 66% = 6.06; pKa's in water = 4.82.

The acetic acid salt was prepared by crystallization from acetic acid; MP: 175°C.

The hydrochloride crystallized from water as a hemihydrate, MP: 267-268°C.

9-Hydroxy-7-methyl-4,5,5a,6-tetrahydroindolo-[4,3-fg]-quinolinium Chloride Hydrochloride

A mix of 1.0g of the betaine above in 10ml of conc. HCl acid and 10ml acetic acid was heated under reflux for 2.5 hours. The mix was cooled, and the product was filtered and washed with ethanol and ether; yield 0.9g (77%), It was recrystallized from ethanol containing a little aqueous HCl acid, MP: 284-285°C dec.; ultraviolet max. wavelength 235mu (20000), 284mu (5600), 324mu (9900).

4-Acetyl-9-hydroxy-7-methyl-4,5,5a,6,7,8,9,10-octahydroindolo-[4,3-fg]-quinoline (x)

1.0g of the betaine above in a mix of 20ml of ethanol and 5ml of water was treated with 0.081g of sodium borohydride, and the solution was refluxed for 10 minutes and kept at 25°C for one hour. The solvent was distilled, and the residue was taken up in a mix of chloroform and water. The chloroform the solvent was distilled. The residue was recrystallized twice from a nitromethane-ethyl acetate mix; yield 0.2g (21%), MP: 193-196°C dec.; ultraviolet max. wavelength 252mu (2060), 328mu (1500). In acid solution the 328mu band was absent.

0.4g of starting betaine was recovered from the aqueous layer.

4-Acetyl-9-beta-hydroxy-7-methyl-4,5,5a,6,7,8,9-octahydroindolo-[4,3-fg]-quinoline (71) A. from the 9-Chloro compound

1.0g of 4-acetyl-9-chloro-7-methyl-4,5,5a,6,6a,7,8,9-octahydroindolo-[4,3-fg]-quinoline hydrochloride below was dissolved in 2ml of water and the solution was refrigerated overnight. The crystalline product which separated was filtered and washed with cold water and ethanol; yield 0.4g (42%). It was recrystallized from aqueous ethanol; MP: 195°C dec.

The free base was obtained by neutralization of the hydrochloride with aqueous sodium bicarbonate. It was crystallized from ethanol or ethyl acetate; MP: 195-197°C dec. The ultraviolet spectrum was like the epimeric alcohol above; pKa's in 66% DMF = 6.68

B. from 4-Acetyl-9-hydroxy-7-methyl-4,5,5a,6,6a,7,8,9-octahydroindolo-[4,3-fg]-quinoline

0.5g of the alpha-alcohol hydrochloride was dissolved in 15ml of concentrated HCl acid, and the solution was kept at 25°C for 2 hours. Excess acid was removed in vacuo, and the residue was neutralized with aqueous sodium bicarbonate. The product was extracted with chloroform, and the solution was dried over magnesium sulfate. The solvent was distilled, and the residue was crystallized from ethyl acetate; yield 0.33g (75%), MP: 194-196°C dec. When mixed with the epimeric alcohol, the MP: 163-165°C dec. Conc. sulfuric acid likewise caused epimerization of the alpha-alcohol, although in slightly lower yield.

4-Acetyl-9-beta-acetoxy-7-methyl-4,5,5a,6,6a,7,8,9-octahydroindolo-[4,3-fg]-quinoline Hydrochloride

4-Acetyl-9-hydroxy-7-methyl-4,5,5a,6,6a,6,7,8,9-octahydroindolo-[4,3-fg]-quinoline, 1.0g, was dissolved in 25ml of acetic acid, and the mix was added 10ml of a saturated solution of boron fluoride in acetic acid. The mixture was warmed briefly until it was homogeneous, and it was then kept at 25°C for 2 hours. The solvent was distilled in vacuo, and the residue was taken up in water and chloroform. Excess sodium bicarbonate was added; the chloroform extract was separated and dried, and the solvent was distilled. The crude acetate ester was dissolved in methanol, and the hydrochloride was precipitated with dry hydrogen chloride. It was recrystallized from methanol-ethyl acetate, MP: 176-177°C dec. and the MP: was depressed when mixed with the alpha-epimer described above.

9-Formamido-7-methyl-4,5,5a,6,6a,7,8,9-octahydroindolo-[4,3-fg]-quinoline

5g of 9-hydroxy-7-methyl-4,5,5a,6,6a,7,8,9-octahydroindolo-[4,3-fg]-quinoline was suspended in 500ml of liquid hydrogen cyanide, and 200ml of boron fluoride etherate was added slowly while cooling in ice. The solution was kept at 25°C for 16 hours, after which it was concentrated in vacuo to thick syrup. Water and chloroform were added, and the mixture was neutralized with excess solid sodium bicarbonate. The product crystallized and was filtered and then digested with hot water and refiltered and washed with water; yield 4.67g (84%), MP: 242-244°C dec. A sample was recrystallized from DMF-methanol; ultraviolet max. wavelength 243mu (29000), 320mu (2000).

The infrared spectrum (mull) had bands at 3.09, 6.00, 6.24, 6.57, 6.83, 6.88 and 7.13.

4-Acetyl-9-formamido-7-methyl-4,5,5a,6,6a,7,8,9-octahydroindolo-[4,3-fg]-quinoline (72)

5.0g of 4-acetyl-9-hydroxy-7-methyl-4,5,5a,6,6a,7,8,9-octahydroindolo-[4,3-fg]-quinoline was suspended in 500ml of liquid hydrogen cyanide, and to the mix was added slowly and while cooling in an ice- bath, 200ml of boron trifluoride etherate. The solution was kept at 25°C for about 20 hours, after which it was concentrated in vacuo to a thick syrup. The residue was then mixed with chloroform and cold water, and excess solid sodium bicarbonate was added to neutralize the acid. The chloroform layer was separated, and the aqueous layer was extracted twice with chloroform. The combined extracts (300ml) were dried over magnesium sulfate, and the chloroform was distilled. The residue was crystallized from 50ml of methanol; yield 4.5g (85%), It was recrystallized from a mix of DMF-methanol; MP: 225-226°C dec.

When the epimeric beta-alcohol was used as starting material in place of the normal alcohol above, the same amide was obtained in the same yield, MP: 227-228°C dec. A mix MP: showed no depression.

The infrared spectrum (mull) had bands at 3.15, 6.03, 6.32, 6.57, 6.83, and 7.13mu.

9-Amino-7-methyl-4,5,5a,6,6a,7,8,9-octahydroindolo-[4,3-fg]-quinoline A. by Acid Hydrolysis of the 9-Formamido Compound

A solution of 5.0g of the 9-formamido compound in 125ml of conc. HCl acid was heated at reflux under nitrogen for 4.5hours. The solution was decolorized with carbon and concentrated in vacuo until a thick slurry of crystals deposited. The product was filtered and washed with alcohol and ether; yield 5.2g, (81%), MP: 303-305°C dec. A sample was recrystallized from aqueous ethanol; ultraviolet max. wavelength 245mu (27600), 296mu (1500), 330mu (1240).

B. by Acid Hydrolysis of the 4-Acetyl-9-formamido Compound

A solution of 2.0g of the formamido compound in 50ml of conc. HCl acid was heated at reflux under nitrogen atmosphere for 18 hours. The solution was evaporated to dryness in vacuo and the residual trihydrochloride salt was recrystallized from aqueous methanol; MP: 291-292°C dec. yield 2.1g (93%).

The salt from either A or B (2.3g) was dissolved in water, and the solution was treated with excess sodium bicarbonate. The free base which separated was filtered and washed with water, methanol and ether; yield 1.6g (90%), MP: 165-166°C after recryst. from ethyl acetate.

C. by Basic Hydroysis of the 9-Formamido Compound

A mix of 2.5g of the formamido compound, 5.6g of potassium hydroxide and 100ml of water was heated at reflux under nitrogen for 17 hours. The solution was cooled, and the product was filtered and washed with water; yield 1.99g (89%), MP: 165-166°C. A mix MP: with the sample obtained by acid hydrolysis showed no depression.

9-Acetamido-4-acetyl-7-methyl-4,5,5a,6,6a,7,8,9-octahydroindolo-[4,3-fg]-quinoline

The 9-amino-4-desacetyl compound, 0.4g, was dissolved in 10ml of acetic anhydride, and the solution was kept at 25°C for 0.5 hours. Excess acetic anhydride was removed in vacuo, and the residue was crystallized from methanol-ether; yield 0.44g (82%). The diacetyl derivative was recrystallized from aqueous methanol, MP: 215-217°C dec. ultraviolet max. wavelength 243mu (42000), 250mu (45000), 305mu (3200), 315mu (2800).

7-Methyl-4-[3',4',5'-trimethoxybenzoyl]-9-[3',4',5'-trimethoxybenzamido]-4,5,5a,6,6a,7,8,9- octahydroindolo-[4,3-fg]-quinoline in 11ml of pyridine was added 0.88g of 3,4,5-trimethoxybenzoyl chloride. The solution was kept at 0°C for 17 hours, after which it was poured into an excess of aqueous sodium bicarbonate solution. The mixture was extracted 3 times with chloroform, and the extracts were washed with water and dried over magnesium sulfate. The solvent was distilled and the product was crystallized from methanol; yield 0.9g (57%). A sample for analysis was recrystallized from a mix of DMF and methanol; MP: 275-280°C dec.

4-Acetyl-9-chloro-7-methyl-4,5,5a,6,6a,7,8,9-octahydroindolo-[4,3-fg]-quinoline hydrochloride (73)

4-Acetyl-9-hydroxy-7-methyl-4,5,5a,6,6a,7,8,9-octahydroindolo-[4,3-fg]-quinoline hydrochloride, 3.1g, was dissolved in 75ml of liquid sulfur dioxide contained in a glass liner in a steel autoclave. Thionyl chloride, 1.2 ml was added, and the vessel was sealed and kept at 25°C for 6 hours. The autochlave was vented, and the reaction mix was removed. Sulfur dioxide was allowed to evaporate while the volume of the solution was kept constant by the slow addition of dry ether. The amorphous chloro hydrochloride was filtered, washed with ether, and dried in vacuo, MP: 130-135°C dec. yield 3.5g.

Use of the 9-beta-epimeric alcohol in this reaction gave the same chloride in comparable yield.

4-Acetyl-9-cyano-7-methyl-4,5,5a,6,6a,7,8,9-octahydroindolo-[4,3-fg]-quinoline (74)

Dry, powdered sodium cyanide, 40g, was added to 300ml of ice-cold liquid hydrogen cyanide. The mix was stirred and cooled in ice, and 7.5g of the crude amorphous hydrochloride described above was added. Stirring was continued for 30 minutes after which the hydrogen cyanide was quickly distilled under reduced pressure below about 10°C. The residue was mixed with chloroform and ice-water, and the resulting mixture was filtered. The organic layer was separated, and the aqueous phase was extracted twice with chloroform. The combined extracts were dried over magnesium sulfate, decolorized and the solvent was distilled in vacuo. The product was crystallized from ethyl acetate, yield 3.3g (54%), MP: 181-182°C; ultraviolet max. wavelength 243mu (37000), 249mu (40000), 306mu (3300), 316mu (2900).

9-Carbomethoxy-7-methyl-4,5,5a,6,6a,7,8,9-octahydroindolo-[4,3-fg]-quinoline (75, R=Me)

The product just above, 1.0g, was mixed with 15ml of methanol and 0.25ml of water. The mix was cooled and 2ml of concentrated sulfuric acid was added slowly. The solution was sealed in a glass tube under nitrogen and heated at 100°C for 23 to 24 hours. The mixture was treated with decolorized carbon and then concentrated in vacuo to about 10ml. It was poured onto a mix of chloroform (30ml), ice and 10g of sodium bicarbonate. The chloroform layer was separated, and the aqueous phase was extracted with 3x10ml portions of chloroform. The combined extracts were dried over magnesium sulfate, evaporated to dryness, and the product was crystallized from benzene; yield 0.51g (53%), MP: 159-160°C. It was recrystallized from ethyl acetate; MP: 160-161°C; ultraviolet max. wavelength 242mu (10600), 318mu (1920); pKa's in 66% DMF = 6.20.

Methanolysis under the same conditions using hydrogen chloride as catalyst gave the ester in lower yield. Methanolysis under milder conditions gave mixtures containing the above ester along with some deacetylated nitrile, MP: 170-171°C.

4-Acetyl-9-carbomethoxy-7-methyl-4,5,5a,6,6a,7,8,9-octahydroindolo-[4,3-fg]-quinoline

Acetylation of one part of the tetracyclic ester using four parts of acetic anhydride in about 25 parts of methanol gave the acetyl derivative, MP: 140-142°C (from benzene-ether).

4-p-Toluenesulfonyl-9-carbomethoxy-7-methyl-4,5,5a,6,6a,7,8,9-octahydroindolo-[4,3-fg]-quinoline (76)

9-Carbomethoxy-7-methyl-4,5,5a,6,6a,7,8,9-octahydroindolo-[4,3-fg]-quinoline, 0.5g, and 0.38g of p- tosyl chloride were mixed in 10ml of pyridine while cooling in an ice-bath. The mix was kept below 25°C for 3 hours, after which most of the excess pyridine was distilled under reduced pressure at RT. Chloroform and excess aqueous sodium bicarbonate solution were added to the residue. The chloroform layer was separated, and the aqueous layer was again extracted with chloroform. The combined extracts were dried over magnesium sulfate, and the solvent was distilled in vacuo. The residue was crystallized from ethanol; MP: 172- 173°C, yield 0.31g (40%).

4-Acetyl-9-carbomethoxy-7-methyl-4,5,5a,7,8,9,10-hexahydroindolo-[4,3-fg]-quinoline (76)

A mix of 0.575g of 4-acetyl-9-carbomethoxy-7-methyl-4,5,5a,6,6a,7,8,9-octahydroindolo-[4,3-fg]- quinoline, 15ml of xylene and 0.5g of 5% Pd/C was heated at reflux under nitrogen for 16 hours. The catalyst was filtered, and the filtrate was cooled. The first crop of yellow crystalline product was collected and washed with benzene; yield 0.20g (35%). Some less pure ester could be obtained by concentrating the filtrates. The melting point after recrystallization from benzene was 177-178°C.

The ultraviolet spectrum was identical to that reported by Stoll7d and by Atherton.7e

9-Carboxy-7-methyl-4,5,5a,6,6a,7,8,9-octahydroindolo-[4,3-fg]-quinoline (75, R=H)

A solution of 1.0g of 9-carbomethoxy-7-methyl-4,5,5a,6,6a,7,8,9-octahydroindolo-[4,3-fg]-quinoline in 30ml of conc. HCl acid and 5ml of water was heated under reflux for 3 hours. The light yellow solution was evaporated completely to dryness in vacuo. A sample of the dihydrochloride salt, obtained thus in quantitative yield, was dissolved in a little water, and the solution was passed through a column of ion exchange resin IR-45 to remove HCl acid. The eluate was evaporated to give the amino acid, MP: above 300°C. A sample was recrystallized from water for analysis.

9-Carboxy-7-methyl-4,6,6a,7,8,9,10,10a-octahydroindolo-[4,3-fg]-quinoline (77)

A mix of 1.0g of the product above and 40ml of @ N sodium hydroxide solution was heated under reflux for 19 hours. The solution was treated with decolorizing carbon, filtered and 10g of wet Raney nickel was added. Refluxing was continued for 3 hours under nitrogen. The catalyst was filtered, and the pH was adjusted to 5.8 by addition of dilute HCl acid. The crude product which separated, 0.5g, containing impurities was purified by reprecipitation from dilute ammonium hydroxide solution with carbon dioxide, MP: 315-316°C dec. The compound retained water of crystallization when dried at 120°C, and was not completely anhydrous after drying at 180°C.

The same dihydrolysergic acid was formed when 4-acetyl-9-cyano-7-methyl-4,5,5a,6,6a,7,8,9-octahydroindolo- [4,3-fg]-quinoline was hydrolyzed with alkali and the hydrolysate was treated with Raney nickel.

The infrared spectrum (mull) had bands at 2.9, 3.1, 6.20, 6.38, 6.89, and 7.30mu. The ultraviolet spectrum was that of an unconjugated indole system; max. wavelength 222mu (30000), 281mu (6200), 291mu (5100).

Synthetic dl-Lysergic Acid (78)

A mix of 9-carbomethoxy-7-methyl-4,5,5a,6,6a,7,8,9-octahydroindolo-[4,3-fg]-quinoline, 3.9g, and 78ml of 1.5% potassium hydroxide solution was refluxed for 30 minutes under nitrogen. Hydrated sodium arsenate, 8.5g, and Raney nickel (16g, wet), previously deactivated by boiling in xylene suspension,31 was added, and the mix was heated under reflux and stirred in a nitrogen atmosphere for 20 hours. The solution was treated with carbon, and the crude lysergic acid was precipitated by neutralization to pH 5.6. It was filtered and washed with water; yield 1.04g, MP: 240-242°C dec. A second crop, 0.16g, MP: 233-235°C dec. was also obtained; total yield 30%. The acid could be purified by dissolving it in dilute ammonium hydroxide, treating with decolorizing carbon, and reprecipitating with carbon dioxide., MP: 242-243°C dec; a mix with dl- lysergic acid made from natural d-lysergic acid39 was likewise 242-243°C dec.

The anhydrous acid was obtained by drying in vacuo for several hours at 150°C.

The ultraviolet spectrum in dilute aqueous alkaline solution was identical with that of the sample derived from natural sources. Max. wavelength 223mu (20400), 238mu (20500), 310mu (9100). The pKa's in 66% DMF (4.92 and 8.04) was the same for both samples, and the X-ray diffraction patterns and paper chromatographic behavior were identical.

dl-Isolysergic Acid Hydrazide from Ergocristine

A sample was obtained by reaction of anhydrous hydrazine with ergocristine in the usual manner.32 It was recrystallized from a mix of DMF and methanol; MP: 225-228°C dec; max. wavelength 229mu (18000), 240mu (18300), 310mu (7630).

Synthetic dl-Isolysergic Acid Hydrazide

Crude synthetic dl-lysergic acid, 0.4g, was powdered and mixed with 23ml of benzene, 2ml of methanol, and 25ml of approximately 2.5% diazomethane in cold ether. The mixture was shaken periodically during 45 minutes. Solvents were evaporated under reduced pressure, after which the residue was taken up in about 20ml of 1:1 benzene:methanol and decolorized with carbon. Solvents were again evaporated, and the crude dl-methyl lysergate was dissolved in 10ml of methanol an 2ml of anhydrous hydrazine. The solution was heated at reflux under nitrogen for 1.5 hours, after which solvents were removed in vacuo, and the dl-lysergic acid hydrazide was crystallized from methanol; yield 0.050g, MP: 224-227°C dec. A mix MP: with natural dl-lysergic acid hydrazide showed no depression. Ultraviolet and infrared spectra and X-ray diffraction patterns for natural and synthetic specimens were identical in every respect.

References

  1. Harvard University; other authors, The Lilly Research Laboratories.
  2. G. Barger, "Ergot and Ergotism", Gurney and Jackson, London, (1931)
  3. A.Stoll, Chem. Revs. 47, 197 (1950)
  4. W. Jacobs and L. Craig isolated lysergic acid [J. Biol. Chem. 104, 547, (1934) and 106, 393, (1934)] and deserve the major credit for the determination of its structure. Their deductions were incomplete only in respect to the placing of one double bond, and stereochemical points. These final details were established by Stoll [A. Stoll, A. Hofmann and F. Troxler, Helv. Chim. Acta. 32, 506, (1949).: A. Stoll, Th. Petrzilka, J. Rutschman, A. Hofmann and Hs. Gunthard, ibid., 37, 2039, (1954)] A comprehensive account of the structural work is given in a review by A. Glenn, Quart. Revs. 3, 192, (1954).
  5. This structure, which was documented by Stoll on the basis of extensive and thorough reductive studies [A. Stoll, Th. Petrzilka and B. Becker, Helv. Chim. Acta, 33, 57, (1950) : A. Stoll and A. Hofmann and Th. Petrzilka, ibid, 34, 1544, (1951)], is unusual in that it represents the alkaloids as ortho amide derivatives containing a free hydroxyl group. No other authentic members of this class are known, and it may be questioned whether factors are present in these molecules which would confer on the ortho amide structures stability vis-a-vis the ring chain tautomeric isomers (i). It is of interest that the large-ring lactone structure (i), earlier considered favorably by Stoll was first suggested by Barger as long ago as 1938 (Barger, "Handbuch der experimentellen Pharmakologie", suppl. Vol. VI, (1938), pp. 84, 221). Attention may be directed here to the presence of large lactone rings in a number of other metabolic products of microorganisms [enneatin A and B: Pl. A. Plattner and V. Nager, Helv. Chim. Acta, 31, 2192, (1948); picromycin: H. Brockmann and R. Oster, Naturwissenshaften, 42, 155, (1955); erythromycin: P. Wiley, K. Gerzon, E. Flynn, M. Sigal and U. Quarck, J. Am. Chem. Soc., 77, 3677, (1955); Magnamycin: unpublished observations by R.B. W.]
  6. F. Uhle and W. Jacobs, JOC, 10, 76, (1945)
  7.  
    1. F. Uhle, J. Am. Chem. Soc., 71, 761, (1949); 73, 2402, (1951)
    2. A. Stoll and J. Rutschmann, Helv. Chim. Acta., 33, 67, (1950); 34, 382, (1951);
    3. A. Stoll , Th. Petrzilka and J. Rutschmann, ibid, 33, 2254, (1950);
    4. A. Stoll, and Th. Petrzilka, ibid, 36, 1125, (1953);
    5. F. Atherton, F. Bergel, A. Cohen, B. Heath-Brown and A. Rees, Chemistry & Industry, 1151, (1953);
    6. C. Grob and co-workers, Helv. Chim. Acta, 33, 1796, (1950); 33, 1955, (1950); 35, 2095, (1952); 36, 839, (1953);
    7. H. Plieninger and co-workers, Ber. 86, 25, (1953); 66, 404, (1953); 87, 228, (1954); 87, 882, (1954); 88, 370, (1955); 88, 550, (1955);
    8. E. Hardegger and co-workers, Helv. Chim. Acta, 38, 463, (1955); 38, 468, (1955);
    9. J. Barltrop and D. Taylor, J. Chem. Soc., 3399, (1954); 3403, (1954);
    10. P. Julian and H. Printy, J. Am. Chem. Soc., 75, 5301, (1953);
    11. A. Berrie, G. Newbold and F. Spring, J.Chem. Soc., 2042, (1952);
    12. A. Glenn "Synthetic Compounds Structurally Related to the Ergot Alkaloids", Ph.D. Thesis, London, (1951);
  8. W.A. Stoll, Schweis. Arch. Neurol. Psychiat., 60, 279, (1947).
  9. Preliminary communication: E. Kornfeld, E. Fornefeld, G.B. Kline, M. Mann, R.G. Jones and R. Woodward, J. Am. Chem. Soc., 76, 5256, (1954).
  10. We considered that this possibility would apply generally to any intermediate containing the tricyclic system (ii), and one additional double bond. Our presumption of the greater stability of the naphthalenoid isomers was based on the fact the resonance energy of naphthalene in much greater than that of indole [L. Pauling, "The Nature of the Chemical Bond", Cornell Univ. Press, Ithaca, N.Y., (1942)]. It has been amply justified by subsequent events. Thus, lysergic acid itself falls into the above category, and has recently been found to suffer ready irreversible isomerization to (iii) in the presence of acids (ref. 7d). Furthermore, at least two attempts by other groups (ref. 7d and 7e) to synthesize lysergic acid were based apparently on the opposite presumption, i.e., isomerization of (iii) to 1, but such an isomerization could not be effected after (iii) had been obtained by synthesis.
  11. B. Blount and R. Robinson, J. Chem. Soc., 3158, (1931). In our hands, the hydrogenation of 3-(beta-carboxyethyl)-indole proceeded most smoothly using Raney nickel in aqueous solution.
  12. Contrast the behavior of simpler compounds of the type 7: L. Fieser and E. Hershberg, J. Am. Chem. Soc., 58, 2314, (1936); L. Fieser and H. Holmes, ibid, 58, 2319, (1936).
  13. C. Grob and P. Payot, Helv. Chim. Acta., 36, 839, (1953).
  14. M. Kuhn, J. Prakt. Chem., 156, 103, (1940).
  15. The Neber Reaction, of which this is an example, has been studied recently by D. Cram and M. Hatch, J. Am. Chem. Soc., 75, 33, and 38, (1953).
  16. Another acyl compound, the 4-acetyl derivative of 4, was prepared by condensation of the ketone with acetic anhydride in the presence of boron fluroride etherate, but was not further investigated.
  17. It was hoped that trifluoroacetic acid would bring about the Schmidt reaction without the strong mineral acid, but no reaction set in until the latter was added.
  18. Some oxides of the type 25, have been reported, [N. Leeds, D. Fukushima and T. Gallagher, J. Am. Chem. Soc., 76, 2943, (1954).] to be formed by the action of peracids on the enol acetates of the 17-keto-steroids.
  19. F. Straus and A. Rohrbacher, Ber. 54, 40, (1921); M. Stoll and A. Commarmont, Helv. Chim. Acta., 31, 1077, (1948.
  20. No rigorous proof of the direction of opening of this oxide with amines was obtained. However, in the 5-substituted-4,5-epoxy series described below reaction with amines takes place at the 4-position.
  21. W.S. Johnson, J.S. Belew, L.J. Chinu and R.H. Hunt have independently discovered the superiority of this catalyst in the Darzens reaction [J. Am. Chem. Soc., 75, 4995, (1953).].
  22. So far as we are aware, there are no previous instances of this reaction sequence, in which advantage is taken of the intermediacy of an enolate in the decarboxylation of glycidates: The dehydrobromination, induced by semicarbazide, follows well-explored paths [W. McGuckin and E. Kendall, J. Am. Chem. Soc., 74, 5311, (1952)].
  23. E. Hershberg, JOC, 13, 542, (1948)
  24. We contemplated the possibility of constructing the desired system in a single step, using the aldehyde 50 in a "quasi-physiological" reaction with methylaminomethylmalonic acid (cf. (iv) to (v)) or suitable components for, or equivalents of that substance. However, a number of attempts to reduce this conception to practice were uneventful. Indeed, the aldehyde was uniformly disappointing in all reactions with simple amines or with active methylene compounds; in the former case, prior to the reaction at the aldehyde group, rather than at C-4, seemed to be the rule and, in all cases, dehydrogenation (or disproportionation) with formation of naphthalene derivatives was a complicating factor.
  25. Other instances are available of the difficulty of hydrolyzing acetals and ketals neighbored by a basic nitrogen atom [R. Moggridge and A. Neuberger, J. Chem. Soc. 745, (1938); C. Grob and H. Utzinger, Helv. Chim. Acta., 37, 1256, (1954)]. The basis for such behavior is reasonable: the usual easy heterolytic cleavage of the C-O bond in such substances involved the generation of a stablilized cation (cf (vi) to (vii)) the formation of which is strongly suppressed by a proximate N+. In the case at hand, the positive pole engendered by the presence of a hydroxyl group, and perhaps also steric and hindrance effects, must enhance the difficulty. It will be noted, nevertheless, that an important later stage in our synthesis involves the hydrolysis of a basic ketal [vide infra. (16-68)].
  26. It is worthy to note that the ketones of this series are susceptible to very ready aerial oxidation. Thus, in an attempt to effect the acetylation of 68 with acetic anhydride in methanol, the sole product isolated was (viii). Special attention may be directed to the facile dehydrogenation of the N-acetyl derivative of the tetracyclic ketone 69 to the interesting betaine (ix), and the reduction of the latter by sodium borohydride to an unsaturated alcohol (x) isomeric with 70.
  27. C. Price and I Krishnamurti, J. Am. Chem. Soc., 72, 5334, (1950).
  28. The stereochemistry implied in the formula 70 and 71, and the complete solution, symbolized for 71 in (xi) may be developed as follows:
    (a) the tightly fused A/B/C ring system can be constructed only with a quasi-axial hydrogen atom at C-5a.
    (b) C-6a will enjoy the more stable of the two possible orientations, i.e., that containing a quasi-equatorial C-N bond; since opportunities for equilibration have been provided in both the tricyclic and tetracyclic classes (cf. 68 and 69), by the proximity of a carbonyl or a vinylogous carbonyl function.
    (c) sodium borohydride reduction of the carbonyl groups ordinarily leads to equatorial alcohols (cf 70);
    furthermore, the relative basicity of 70, (pKa, 6.02) and 71 (pKa, 6.68) confirms our assignment, in that stablilization of the conjugate acid by hydrogen bonding is possible in the stronger base 71, but not in 70. A related case, that of quinine and epiquinine, is discussed by R. Turner and R.Woodward (R. Manske and H. Holmes, "The Alkaloids", Academic Press, Inc, N.Y., N.Y., (1953), p. 32.). Assignment of the configuration of the epimeric 10-hydroxydihydrodesoxycodienes was recently made in a similar fashion [H. Rapoport and S. Masamune, J. Am. Chem. Soc., 77, 4330, (1955).]
  29. J. Ritter and J. Kalish, ibid, 70, 4048, (1948).
  30. References 7d and 7e.
  31. E. Kleiderer and E. Kornfeld, JOC, 13, 455, (1948).
  32. A. Stoll and A. Hofmann, Z. Physiol. Chem., 250, 7, (1937).
  33. S. Smith and G. Timmis, J. Chem. Soc., 1440, (1936).
  34. A. Stoll and A. Hofmann, Helv. Chim. Acta., 26, 922, 944, (1943).
  35. B. K. Blount and R. Robinson, J. Chem. Soc., 3158, (1931).
  36. F. C. Uhle, J. Am. Chem. Soc., 71, 761, (1949).
  37. M. Kuhn, J. Prakt. Chem., 156, 103, (1940).
  38. R.B. Woodward, N.L. Wendler, and F.J. Brutschy., J. Am. Chem. Soc., 67, 1425, (1945).
  39. S. Smith and G.M. Timmis, J. Chem. Soc., 1440, (1936).