Model Reactions for the Biosynthesis of Thyroxine ... - ACS Publications

Thyroxine is formed in about 20% yield when. DIHPPA4 is permitted to react with DIT4 at room temperature in a neutral or slightly alkaline solution.5...
3 downloads 18 Views 9MB Size
Notes

-

Model Reactions for the Biosynthesis of Thyroxine. VIII. The Reaction of Various Analogs of 4-Hydroxy-3,5-diiodophenylpyruvic Acid with 3,5-Diiodotyrosine1~* TETSVO s H I B A 3 AND H. J. CAHNMANN

modest yield. In these cases, the analog used had the same aliphatic side chain as DIHPPA, and the phenolic ring still contained one iodine atom or two bromine atoms in ortho position to the phenolic hydroxyl. 4Hydroxy-3,5-diiodophenylacetaldehyde(X) did not react with DIHPPA. Upon alkalinization of the mixture, the aldehyde apparently underwent an aldol condensation leading to the tetraiodo aldehyde XII.

A'ational Institute of Arthritis and Metabolic Diseases, A'ational Institutes of Health, Bethesda, Maryland Experimental'

TERCOMATSUVRA, AKIRANISHINAGA, A N D HARUSHIGE SAKAMOTO Department of Synthetic Chemistry, Faculty of Engineering, Kyoto lhiversity, Kyoto, J a p a n Received M a y 6 , 1964

Thyroxine is formed in about 20% yield when DIHPPA4 is permitted to react with D I T 4 a t room temperature in a neutral or slightly alkaline solution.5 Analogs of thyroxine are formed when in this reaction the amino acid is replaced with certain analogs of DIT. The influence of structural changes in the D I T molecule upon the yield of the corresponding analog of thyroxine has been inve~tigated.68~The present Note deals with the influence of structural changes in the keto acid component of this coupling reaction upon the yield of the coupling product. Various conipounds structurally related to DIHPPA were permitted to react with D I T (and in a few cases with desamino analogs of DIT) following the usual procedure described in the previous papers of this series. The yield of thyroxine (or its analogs) was determined by isolation and weighing or, in the case of very poor yields, by paper chromatography. The syntheses of most of the analogs of DIHPPA used have not been previously reported and are therefore described in the Experimental part. The yields obtained in the various coupling reactions are reported in Table I. It has been shown6-8 that in the reaction between DIHPPA and D I T , the latter can be replaced with various analogs of D I T , without a drastic reduction in the coupling yield. In contrast, the requirement for DTHPPA is much more specific. It can be seen from Table I that no coupling took place in most instances when DIHPPA was replaced with analogs of this keto acid. Only with a few analogs was a coupling product obtained, and then only in poor or (1) Paper VII: T. Matsuura, H. Kon, and H. J. Cahnmann, J . Org. Chem., 49,3058 (1964). (2) This investigation was supported in part by U. 9. Public Health

Service Research Grant AM 07955 from the National Institute of Arthritis and Metabolic Diseases. (3) Visiting Scientist from the Department of Chemistry, Faculty of Science, Osaka University, Osaka, Japan. (4) Abbreviations: DIHPPA. 4-hydroxy-3.5-diiodopbenylpyruvic acid: D I T , 3,b-diiodotyrosine. (5) R. I. Meltzer and R. J. Stanaback, J . Orp. Chem., 16, 1977 (1961): (6) A. Nishinaga and T. Matsuura, ihid., 39, 1812 (1964). (7) T. Shiba and H. J. Cahnmann, ihid., 29, 1652 (1964). ( 8 ) T. Shiba and H. J. Cahnmann, ihid., 29, 3063 (1964).

3061

S-Iodovanillin.-A solution of 16.2 g. (0.1 mole) of iodine monochloride in 20 ml. of 207, HC1 was added a t once to a warm solut,ion (50") of 14.2 g. (0.1 mole) of vanillin in 400 ml. of 20% HC1. After 1 week, the precipitate formed (21.6 g., 81y0),was collected, and recrystallized from aqueous ethanol yielding needles, m.p. 182-183" (lit.lo m.p. 181-182"). 4-Hydroxy-3-iodo-5-methoxyphenylpyruvic Acid (IV).-Iodovanillin was converted to 4- (4-acetoxy-3-iodo-5-methoxybenzal)2-methyl-5-oxazolone following a procedure similar to the one described below for the synthesis of 4-(4-acetoxy-3,5-dibromobenzal)-2-methyl-5-oxazolone; m .p. 192-194" (lit . l l m.p. 196-197'), yield 86% (lit." 27%). This azlactone was converted to IV by acid hydrolysis as described for the synthesis of DIHPPA.6 Recrystallization of the crude keto acid from ethanol-acetic acid gave needles, m.p. 241-243" dec., in 58% yield. A n a l . Calcd. for CloHsIO~:C, 35.75; H, 2.70. Found: C, 36.26; H , 3.00. 4-~4-Acetoxy-3,5-dibromobenzal j -2-meth yl-5-oxazolone .-A mixture of 22.6 g. (0.08 mole) of 3,5-dibromo-4-hydroxybenzaldehyde,l2 10 g. (0.08 mole) of acetylglycine, 7 g. (0.08 mole) of freshly fused sodium acetate, and 75 ml. of acetic anhydride was heated on a boiling water bath for 2 hr. After cooling, 10 ml. of petroleum ether (b.p. 40-60") and 10 ml. of water were added to the reaction mixture which was then broken up by means of a glass rod. The crystalline mass was filtered and washed with water, dried, and recrystallized from benzene to give needles, m.p. 180-185", yield 27 g. (86%). An analytical sample was recrystallized a second time; m.p. 187-189'. A n a l . Calcd. for CI3HgBr2NO4:C, 38.74; H, 2.25; Br, 39.66; N, 3.48. Found: C, 38.50; H , 2.47; Br, 39.51; E, 3.62. 3,5-Dibromo-4-hydroxyphenylpymvic Acid (V).-The oxazolone was converted to V by acid hydrolysis as described for the synthesis of IV. Recrystallization of the crude keto acid from aqueous ethanol gave needles, m.p. 208" dec., in 61% yield. A n d . Calcd. for C9H6Br2O4: C, 31.71; H, 2.23. Found: C, 31.98; H, 1.79. N-(a-Acetamido-4-acetoxy-3,5-diiodocinnamoyl)glycine.-To a suspension of 3.3 g. (6.6 mmoles) of 4-(4-acetoxy-3,5-diiodobenzal)-2-methyl-5-oxazolone5 and 0.5 g. (6.6 mmoles) of glycine in 30 ml. of acetone was added 13.2 ml. of 0.5 'V NaOH. After standing for 1.5 hr., 6.6 ml. of 1 A' HCI was added; the reaction mixture was concentrated to about 10 ml. and then kept overnight a t 4". The precipitate formed was recrystallized from aqueous ethanol yielding 3.0 g. (78%), m.p. 222-223". A n a l . Calcd. for CljH,412N20s: C, 31.49; H , 2.47; I , 44.66; N, 4.90. Found: C, 31.28; H , 2.59; I , 44.59; N, 4.80. N-(a-Acetamido-4-hydroxy-3,5-diiodocinnamoyl)glycine .This product can be obtained by mild alkaline hydrolysis of the corresponding acetoxy compound described in the preceding (9) The elemental analyses were carried out by Schwarzkopf Microanalytical Laboratories, Woodside, ,N. Y., and by Mr. J. Coda and his a8sociates of the Faculty of Science, Osaka City University, Osaka, Japan, Melting points were determined in capillary tubes and are uncorrected. (IO) G . D. Thorn and C. B. Purves, Can. J. Chem., 32,373 (1954). (11) L. C. Raiford and C. H. Buurman. J . O w . Chem., 8 , 466 (1943). (12) C. Paal. Ber., 48, 2407 (1895).

NOTES

3062

VOL. 29

TABLE I YIELDSOBTAINED I N VARIOUS COUPLING REACTIONS DIHPPA and analog6

-DIT

and andogs-

RI

HO

@-

RI

RZ

Compd.

a

I I1 I11

H

IV

V VI VI1 VI11 IX X XI See ref. 5.

Rt

Ri

I I

Yield of thyroxine or its analogs, yo

Ra

I

.I H H CHsO

CHzCOCOOH CHzCOCOOH CHzCOCOOH CHzCOCOOH

Br

Br

CHzCOCOOH

I I I I I I

I I

CHzCOCONHCHzCOOH CHzC( COOH)=NOH CH=C( CO0H)NHCOCHo COCOOHc CHzCHO CHOc

* See ref. 7.

I I I I

I I I Br I I I I

I I

H NHz H H NHz NHz NHz NHz NHz NH*

20a 2b None4 4 1 9 3 1 Trace Trace None None None None

See ref. 18.

preparation. It is, however more conveniently prepared directly from 4-(4-acetoxy-3,5-diiodobenzal)-2-methyl-5-oxazolone. A solution of 1.65 g. (22 mmoles) of glycine in 22 ml. of 1 N NaOH was added to a suspension of 9.94 g. (20 mmoles) of this oxazolone' in 60 ml. of acetone. After stirring for 4.5hr. the oxazolone was completely dissolved. Then 20 ml. of 2 N NaOH was added and stirring was continued for 1 hr. The precipitate, formed after acidification with 2 N HC1 and cooling, was recrystallized from ethanol yielding 6.73 g. (63'%), m.p. 24C-242" dec. Anal. Calcd. for C13H1~12N206: C, 29.45; H , 2.28; I , 47.88; N, 5.29. Found: C, 29.67; H , 2.46; I , 47.71; N, 5.04. N-(4-Hydroxy-3,5-diiodophenylpyruvyl)glycine (VI) .-A mixture of 1.06 g. (2 mmoles) of N-(a-acetamido-l-hydroxy-3,5-diiodocinnamoyl)glycine, 15 ml. of acetic acid, and 6 ml. of 1 N HCl was heated for 2 hr. on a steam bath. The reaction mixture was clarified by filtration, and the filtrate was evaporated. The residue was triturated five times with 20 ml. of ether. The combined ether extracts were mixed with petroleum ether (1: 1). The precipitate obtained (0.12g., 12.3%) melted a t 219-223". Anal. Calcd. for C l 1 H 9 I ~ N 0 ~C, : 27.02; H , 1.86; I , 51.90; N, 2.86. Found: C, 27.26; H , 2.16; I , 52.19; N, 2.63. Z-Oximino-3-(4-hydroxy-3,5-diiodophenyl)propionic Acid (VII). -Hydrogen was bubbled through a solution of 10 g. (0.14mole) of hydroxylamine hydrochloride in 60 ml. of water and 150 ml. of 2 N NaOH, in order to remove oxygen. Then 4.7 g. (11 mmoles) of DIHPPA'3 was added and the mixture was heated in an atmosphere of hydrogen14 on a steam bath for 20 min., then cooled and acidified with 4 N HC1 to about pH 4. On standing overnight a t 4" a crystalline precipitate formed, which, on recrystallization from aqueous ethanol, gave 2.6 g. (52y0)of crystals, m.p. 173-176". Anal. Calcd. for CoH71zN04.1.5Hz0: C, 22.81; H , 2.13; I , 53.55; N, 2.96. Found: C, 22.83; H , 2.59; I , 53.35; N, 3.07. ~-Acetamido-4-hydroxy-3,5-diiodophenylcinnaic Acid (VIII). -A suspension of 2.49 g. (5 mmoles) of 4-(4-acetoxy-3,5-diiodobenzal)-2-methyl-5-oxazolone~ in 15 ml. of 20y0 aqueous KOH was heated for 20 min. on a steam. bath. The reaction mixture was cooled and acidified (congo red) with 1 N HC1. The crystals formed were recrystallized from aqueous ethanol yielding 1.56 g. (66y0),m.p. 245-247' dec. (lit.1sm.p. 234-235"). Anal. Calcd. for CllH9IzNO4: C, 27.93; H, 1.92; I , 53.66; N,2.96. Found: C, 28.15; H,2.06;I , 53.43; N,2.95. Acetyl-Z,6-diiodochavicol (O-Acetyl-4-allyl-Z,6-diiodophenol) . -A solution of 13 g. (123mmoles) of Na2C03 in 100 ml. of water (13) Commercially available from Osaka Laboratory of Synthetic Organic Chemicals, 74 Ueda Higashimachi, Nishinomiya, Japan. (14) In alkaline medium DIHPPA undergoes a rapid breakdown to the aldehyde X I , when oxygen is present. (15) M. Nakano and S. Tsuchiya. Gunma J . Med. Sei., 10, 280 (1961).

was mixed with a solution of 6.7 g. (50 mmoles) of chavicol (4allylphenol)l6 in 50 ml. of 1 Y NaOH. The mixture was cooled in an ice bath and a solution of 25.4 g. (0.1mole) of iodine and 25.5 g. of potassium iodide in 100 ml. of water was added over a period of 1 hr. An aqueous suspension of the precipitate formed was acidified with 4 N HC1. An oil formed which solidified in the cold yielding 17.1 g. (78%). A solution of 16.6 g. of this crude diiodochavicol in 50 ml. of pyridine was cooled in an ice bath and 50 ml. of acetic anhydride was added. After standing overnight a t room temperature, the mixture was poured into ice-water, acidified with 4 N HC1, and extracted with ether. After successive washings with 2 N HCl, water, 1 N NaOH, and again water, the ether extract was dried and evaporated. Distillation of the residue gave 15.0 g. (82%) of a viscous oil, b.p. 146-148" (0.5mm.), which crystallized upon standing a t 4' for 1 month; m.p. 42-45'. Anal. Calcd. for C1lHl&Oz: C, 30.86; H , 2.35. Found: C, 30.91; H , 2.50. Semicarbazone of 4-Acetoxy-3,5-diiodophenylacetaldehyde.To a mixture of 2.2 g. (5 mmoles) of acetyl-2,6-diiodochavicol, 50 ml. of ether, 50 ml. of water, and 50 mg. (0.2mmole) of osmium tetroxide, was added 2.3 g. (12 mmoles) of sodium periodate in small portions.17 After stirring for 4.5 hr., the ether layer wm separated and concentrated to about 10 ml. A solution of 2.2 g. (20mmoles) of semicarbazide hydrochloride in 10 ml. of water was added and the mixture was rendered homogeneous by the addition of about 10 ml. of pyridine. After standing overnight, the mixture was poured into ice-water. The precipitate formed gave, after recrystallization from ethanol, 1.5 g. (62y0)of fine prisms, m.p. 177-179". Anal. Calcd. for CllH1112N303:C, 27.12; H , 2.28; N,8.63. Found: C,27.18;H , 2.47; N,8.65. The 2,4-dinitrophenylhydrazone waa obtained by adding a solution of 2,4-dinitrophenylhydrazinein a mixture of aqueous ethanol and sulfuric acid to the above-mentioned concentrated ether solution. Recrystallization from ethanol gave yellow needles, m.p. 21C-213'. Anal. Calcd. for ClsH1~IzN40e: C, 31.50; H , 1.98; N, 9.18. Found: C, 31.82; H , 2.14; N,9.11. Semicarbazone of 4-Hydroxy-3,5-diiodophenylacetaldehyde .A suspension of 1.2 g. (2.5 mmoles) of the above-mentioned semicarbazone in a mixture of 10 ml. of ethanol and 10 ml. of 2 N NaOH was warmed to 40' in order to obtain a clear solution which was permitted to stand a t room temperature for 1.5 hr. Water was then added and the mixture waa acidified with 4 N HC1. The precipitate formed gave, after recrystallization from ethanol, 0.57 g. (52y0)of needles, m.p. 175-176". Anal. Calcd. for CoH91ZN30z: C, 24.41; H , 2.04; N, 9.44. Found: C,24.61;H , 2.36; N,9.42. (16) V. Grignard and .I. Ritz, Bull. 8oc. chtm. France, [518, 1181 (1936). (17) Cf. R. Pappo, D. S. Allen, Jr.. R. U. Lemieux, and W. 9. Johnson, J . O w . Chem., P i , 478 (1956).

OCTOBER,1964

NOTES

4-Hydroxy-3,5-diiodophenylacetaldehyde(X).-To a solution of 1.18 g. (3.1 mmoles) of the semicarbazone of 4-hydroxy-3,5diiodophenylacetaldehyde in 30 ml. of acetic acid and 15 ml. of water, was added 3 ml. (43 mmoles) of pyruvic acid. The solution was kept for 20 hr. a t 40" in an atmosphere of nitrogen. Water was then added until a slight turbidity persisted. On standing overnight a t 4', 0.81 g. of crystals were obtained. The mother liquors yielded another crop of 0.11 g. The crude product was recrystallized from aqueous acetic acid yielding 0.29 g. (28%) of needles, m.p. 111-113'. Anal. Calcd. for C8H61202:C, 24.75; H , 1.55. Found: C, 24.92; H, 1.76. Coupling Reactions of Analogs of DIHPPA with DIT.-These reactions were carried out essentially as described previously6 with several minor modifications imposed by the nature of the analog used. The amount of analog used in a run ranged from 0.5 to 6 mmoles. The analog VI11 was insoluble in butanol and was therefore dissolved in 2 S NaOH and the pH was kept a t 7.6 by the slow addition of 4 S HC1. In several cases, the addition of t-butylhydroperoxide was omitted since preliminary tests indicated that it had little if any influence on the yield of the analog of thyroxine. The method of working up the reaction mixture also depended on the nature of the starting material and on the yield of coupling product. I n the case of very small yields the thyroxine f&med could only be detected by paper chromatography in the usual solvent systems.18 3,5,3'-Triiodo-J'-methoxythyronine.-A coupling reaction was carried out with 2.02 g. (6 mmoles) of the keto acid IV and 2.01 g. (4.6 mmoles) of D I T . The crude product was extracted with acetone in order to remove side products of the reaction, mainly 5-iodovanillin. The undissolved material, m.p. 218-219' dec., weighed 0.13 g. (4YG). It was chromatographically pure. Crystallization from a 553 sodium carbonate solution gave the sodium salt in the form of fine crystals. 3,5,3'-Triiodo-5'-methoxythyropropionic Acid .-A coupling reaction of 2.02 g. (6 mmoles) of the keto acid IV and 1.94 g. (4.6 mmoles) of 3,5-diiodophloretic acid's gave a crude product which was crystallized from benzene to give 17 mg. (lye) of prisms, m.p. 196-198'. Anal. Calcd. for C16H131306:C, 28.79; H, 1.97. Found: C, 28.26; H , 2.08. 3 ',5'-Dibromo-3,5-diiodothyronine.-A coupling reaction was carried out with 2.02 g. (6 mmoles) of the keto acid V and 2.17 g. (4.6 mmoles) of D I T dihydrate. The crude product was extracted with acetone in order to remove side products of the reaction, mainly 3,5-dibromo-4-hydroxybenzaldehyde. The undissolved material, m.p. 243' dec., weighed 0.28 g. (9%). It was chromatographically pure. Treatment of a solution in methanolammonia (95: 1) with dilute acetic acid gave a colorless precipitate, m.p. 244' dec. (lit. m.p. 245-246' dec.19 and 244.5' dec.20). 3',5'-Dibromo-3,5-diiodothyropropionic Acid.-The crude product obtained in a coupling reaction between 2.02 g. (6mmoles) of keto acid V and 1.94 g. (4.6 mmoles) of 3,5-diiodophloretic acid18 was crystallized from benzene yielding 0.1 g. (3%) of needles, m.p. 203-205". The infrared spectrum was identical with that of a sample prepared by a different synthetic route.2l 3,5,3',5'-Tetrabromothyropropionic Acid .-A coupling reaction between 1.27 g. (3.5 mmoles) of the keto acid V and 0.59 g. (2.9 mmoles) of 3,5-dibromophloretic acid22 gave a crude product which, after crystallization from benzene, yielded 18 mg. (1%) of needles, m.p. 184-186'. The infrared spectrum was identical with that of a sample prepared by a different synthetic route.3' Behavior of 4-Hydroxy-3,s-diiodophenylacetaldehyde(X)in the Coupling Reaction with DIT.-The reaction waa carried out in the usual manner with 0.25 g. (0.64 mmole) of the aldehyde X and 0.25 g. (0.53 mmole) of DIT. When the residue obtained after evaporation of the butanol extract was acidified, 0.1 g. of a precipitate was obtained which, after crystallization from ethyl acetate, gave fine needles, m.p. 248-249' dec. No thyroxine was detected by paper chromatography of the mother liquor. Further investigation showed that the presence of D I T was not required for the formation of this product which waa tentatively identified as the aldehyde X I I , formed by aldol condensation of 2

molecules of X . The aldehyde X was fairly stable when it was treated with oxygen a t pH 7.6 However, when alkali was added the condensation product XI1 formed. Structure XI1 is compati-

(18) T. Matsuura and H. J. Cahnmann. J . A m . Chem. Soc., 81, 871 (1959). (19) K.Schuerraf, H e h . Chim. Acto, 19, 405 (1929). (20) C. R. Harington and W. McCartney, J . Chem. Soc.. 899 (1929). (21) T. Matsuiira. unpublished work. (22) T. Matsuura and H. J. Cahnmann, J . A m . Chem. Soc., 89, 2055 (1962).

3063

I

I.

XI1 ble with the elemental analysis and the infrared spectrum. A carbonyl band a t 1730 cm.? indicated that the carbonyl is not in conjugation with a double bond.Z3 The ultraviolet spectrum showed bands a t 220 mp (e 37,000), 245 (26,000), 260 (24,000), 348 (28,400), and shoulder a t 415 (6300). Anal. Calcd. for ClGHloLOa: C, 25.33; H, 1.32. Found: C, 25.21; H , 1.55. (23) L. J. Bellamy, "The Infrared Spectra of Complex Molecules," 2nd Ed., John Wiley and Sons, Inc., New York, N. Y., 1958, p. 133.

Model Reactions for the Biosynthesis of Thyroxine. IX. Synthesis of Peptides of L-Thyroxine1p2 TETSUO SHIBA AN ~D H. J. CAHNMANN National Institute of Arthritis and Metabolic Diseases, National Institutes of Health, Belhesda, Maryland Received May 6, 1964

Only a few peptides of thyroxine4J have been synthesized.6 The method used, heating of halogenoacylthyroxine with ammonia, permits the preparation of only a limited number of peptides, leads to racemization, and cannot be applied to the synthesis of those peptides in which the carboxyl group of thyroxine is peptide linked. The numerous methods for the synthesis of peptides published during the last few decades are not easily applicable to the synthesis of peptides of thyroxine since most of them involve the removal of a blocking group by either strong acid or hydrogenolysis, which leads to partial or total deiodination. It is also not possible to synthesize peptides of thyroxine by iodinating peptides of thyronine since this would lead to peptides of 3',5'-diiodothyronine. I n the present investigation a new principle, etherification of peptides of diiodotyrosine with 4-hydroxy-3,5diiodophenylpyruvic acid, was applied to the synthesis of peptides of thyroxine. This reaction is analogous to that in which the same keto acid converts diiodotyrosine to thyroxine7 and various analogs of diiodotyrosine to the corresponding analogs of thyroxine.899 The (1) Paper VIII: T . Shiba, H. J. Cahnmann, T. Matsuura, A. Nishinaga. and H. Sakamoto, J . Orq. Chem., 49, 3061 (1964). (2) A preliminary report of this work was presented at the 141st National Meeting of the American Chemical Society, Washington, D.C., March, 1962. (3) Visiting Scientist from the Department of Chemistry, Faculty of Science, Osaka University, Osaka, Japan. (4) J. N. Ashley and C. R. Harington, Biochem. J . , 99, 1436 (1928). (5) E. Abderhalden and E. Schwab. Fermentfomhunq., 11, 164 (1930). (6) The formation of N-acylated peptides of thyroxine in the aerobic incubation of N-acylated peptides of diiodotyrosine was described by R. Pitt-Rivers [Biochem. J . , 4.9, 223 (194811 and by R. Pitt-Rivers and A. T. James [ibid., TO, 173 (1958)l. Attempts to convert these acyl derivatives to the free thyroxine peptides failed, because removal of the acyl group was accompanied by fission of the peptide bond (R. Pitt-Rivers, personal communication). (7) R. I. Meltrer and R. J. Stanaback, J . Org. Chem., 46, 1977 (1961). ( 8 ) A. Nishinaga and T. Matsuura, i b i d . , 99, 1812 (1964). (9) T.Shiba and H. J. Cahnmann. ibid., 99, 1652 (1964).

NOTES

3064 Woodward's reagent's

VOL.29

+ H-gly-OEt, HC1Cbz-L-tyr-gly-OEt Woodward's reagent Cbz-L-tY-OH + H-gly-OBz, tos +Cbz-L-tyr-gly-OBz (1) Woodward's reagent Cbz-gly-OH + H-L-tyr-OMe, HC1 Cbz-L-tyr-OH

(2) N a O H

Cbz-C1

H-gly-L-t yr-OH

\+ /

NaOH __j

Cbz-L-tyr-gly-OH

Hn-PD

H-L-tyr-gly-OH

HrPd

---

+-i

(1)

(1) H-gly-OEt, HC1

+

Cbz(or OzN-Cbz)-gly-L-tyr-OH

( 2 ) Woodward's reagent

or 0zN-Cbz-C1 NaOH

Cbz(or OZN-Cbz)-gly-L-tyr-gly-OEt

-+Cbz(or OzN-Cbz)-gly-L-tyr-gly-OH

reaction takes place under mild conditions (room temperature, neutral pH), does not lead to racemixation,l0 and permits the preparation of peptides in which either the amino group or the carboxyl group of thyroxine, or both, are peptide linked. Glycyl-L-thyroxine, Lthyroxylglycine, and glycyl-L-thyroxylglycine were prepared in this manner in 7-1 1yoyield from the corresponding peptides of diiodotyrosine. The same synthetic principle is presumably applicable to the synthesis of any other peptide of thyroxine. The required peptides of diiodotyrosine were prepared by iodination of the corresponding peptides of tyrosine. Tyrosylglycine was refractory to iodination with iodine monochloride or potassium triiodide. When, however, the free amino group was blocked by adsorption of this peptide on a strong cation exchanger (Dowex 50), iodination proceeded smoothly. Glycyl-L-tyrosine was a commercial product. LTyrosylglycine and glycyl-L-tyrosylglycine were prepared by reactions 1 and 2." Since most of the compounds shown in this reaction scheme had been prepared previously, although by other methods, only new intermediates are described in the Experimental part. The method of Woodward, et a1.,12for the synthesis of the tyrosine peptides always gave solid reaction products, while the mixed anhydride and dicyclohexylcarbodiimid methods often yielded oily products. The only shortcoming of the use of Woodward's reagent is that the reaction product sometimes contains traces of sulfur that poison the catalyst in the subsequent hydrogenolysis. Therefore, when the Cbz-peptide, obtained after treatment of the reaction product with sodium hydroxide, showed a positive sodium nitroprusside reaction, it was dissolved in ethyl acetate and extracted several times with a solution of sodium carbonate. The precipitate obtained after acidification of the combined extracts was then free of sulfur. Paper chromatography was used routinely to follow the purification of synthetic intermediates and end products (see Table I). The fact that peptides of diiodotyrosine can replace free diiodotyrosine in the reaction with 4-hydroxy-3,5diiodophenylpyruvic acid again shows that in this coupling reaction, the structural requirements are less stringent for the amino acid component than for the The conversion under mild keto acid component. conditions of diiodotyrosine to thyroxine within a peptide is of particular interest because it is generally as'

198,9

(10) T. Shiba a n d H. J. C a h n m a n n , J. Ore. Chem. 21, 1773 (1962). (11) Abbreviations: Cbz. carbobenzyloxy; OIN-Cbz. p-nitrocarbobenzyloxy; t y r , tyrosyl; gly, glycyl; M e , methyl; E t , ethyl: Ba, benzyl; tos. p-toluenesulfonate. (12) R. B. Woodward, R. A. Olofson, a n d H. Mayer. J . A n . Chem. Soc., 83. 1010 (1961).

HzPd

d

H-gly-L-tyr-gly-OH

(2)

TABLE I Rf VALUESOF PEPTIDES OF TYROSINE, DIIODOTYROSINE, AND

Peptide

THYROXINE" Solvent 16

Solvent 2c

gly-L-tyr gly-L-DIT g1y -L-T4 L-tyr-gly I,-DIT-gly

0.12 0.11 0.13 0 . 28d 0.27 0,57d 0.14 0.12 0.09 0.32d L-T~-& 0.37 0 . 5sd gly-L-tyr-gly 0.09 0.12 gly-L-DIT-gly 0.05 0.2id gly-L-TA-gly 0.33 0.51d Abbreviations : gly, glycine; tyr, tyrosine; DIT, 3,j-diiodotyrosine; T,, thyroxine. 6 1-Butanol-dioxane-2 N ammonia ( 4 : 1:5). 1-Butanol-acetic acid-water ( 7 8 :10: 12). d Elongated spot.

sumed that the biosynthesis of thyroxine from diiodotyrosine takes place within a protein, thyroglobulin. ExperimentaP

p-Nitrocarbobenzyloxyglycyl-L-tyrosine.-To a stirred and ice-cooled solution of 1.90 g. (8.0 mmoles) of glycyl-L-tyrosine in 8 ml. of 1 N NaOH were added simultaneously over a period of 30 min. 10 ml. of 1 N NaOH and a solution of 2.16 g. (10.0 mmoles) of p-nitrocarbobenzyloxychloride in 10 ml. of dioxane. Stirring was continued for 30 min. with cooling and for another 30 min. a t room temperature. A small amount of insoluble material was removed by filtration and the filtrate was acidified (congo red) and extracted three times with ethyl acetate. The combined extracts were washed once with 1 N HC1, then three times with a 5% solution of sodium carbonate. After acidification of the combined alkaline washings with HCl, the mixture was permitted to stand overnight a t 4'. The crystals formed were washed with water and dried yielding 1.68 g. (51y0),m.p. 168170'. An analytical sample was recrystallized from water; m.p. 172-174". Anal. Calcd. for C19H~N308: C, 54.67; H , 4.59; N , 10.07. Found: C, 54.68; H , 4.76; N , 10.23. p-Nitrocarbobenzyloxyglycyl-L-tyrosylglycine Ethyl ester.A suspension of 1.27 g. (5.0 mmoles) of N-ethyl-5-phenylisoxa~olium-3'-sulfonate~~ in 30 ml. of nitromethane was added to a solution of 2.09 g. (5.0 mmoles) of p-nitrocarbobenzyloxyglycylL-tyrosine in 0.51 g. (5.0 mmoles) of triethylamine and 20 ml.of nitromethane. After stirring for 30 min. the reagent was completely dissolved. A solution of 0.70 g. (5.0 mmoles) of glycine ethyl ester hydrochloride in 0.51 g. (5.0 mmoles) of triethylamine and 50 ml. of nitromethane was then added and stirring was continued overnight. The residue obtained after evaporation of the solvent was dissolved in ethyl acetate. The solution was washed with 1 N Na2C03, 1 N HC1, and water. The solution was evaporated and the residue was crystallized from ethanol-water m.p. 95-97" dec. yielding 1.96 g. (78y0), (13) T h e microanalyses were made b y Mr. McCann and his associates of this I n s t i t u t e a n d b y Schwarzkopf Microanalytical Laboratories, Woodside. N. Y. Melting points were taken in capillary tubes and are uncorrected. (14) Woodward's reagent K . Pilot Chemicals, Inc.. Watertown 72. Mass. (cf. ref. 12).

OCTOBER,1964

NOTES

Anal. Calcd. for C Z ~ H Z ~ TC, ~ O54.97; ~ : H , 5.22; IV, 11.15. Found: C, 54.94; H , 5.23; N , 11.29. p-Nitrocarbobenzyloxyglycyl-L-tyrosylglycine .-To 1236 g (3.7 mmoles) of p-nitrocarbobenzyloxyglycyl-L-tyrosylglycine ethyl ester was added 1 .V NaOH (9.5 ml.) with stirring a t such a rate that a pH of about 12 was maintained. After 1.5 hr. the solution was clarified by filtration, and the filtrate was acidified (congo red) with 4 X HC1, then left a t 4" for a few hours. The crystals formed were washed with water and dried yielding 1.19 g. (6870), m.p. 218-220' dec. Anal. Calcd. for CzlHzzN409: C, 53.16; H , 4.67; S , 11.81. Found: C, 53.07; H , 4.93; N, 11.81. Glycyl-3,5-diiodo-~-tyrosine.-A solution of 5.1 g. (31 mmoles) of iodine monochloride in 11.3 ml. of 20% HCl was added with stirring to a solution of 3.6 g. (15 mmoles) of glycyl-~-tyrosine~~ in 25 ml. of 1 ,V HCl. After 1.5 hr. an aqueous solution of sulfur dioxide was added until the reaction mixture became pale yellow. The precipitate formed on adjusting the pH to 4 with 6 N S H r OH was washed with water, and purified by reprecipitation (pH 7 ) from its solution in 6 N NHdOH yielding 5.1 g. (69%) of needles, m.p. 222-224" dec.; ref. 16 gives a sintering point of 122-128' and a melting point of 290-292" dec. For paper chromatogra hy, see Table I . Anal. Caycd. for CllH12IZX:204: C, 26.96; H , 2.47; I , 51.79; IG,5.72. Found: C,27.08; H,2.69; I,51.89; N , 5.82. 3,5-Diiodo-~-tyrosylglycine.-Dowex 50 X 12, H-form (about 20 ml. of wet resin) was added to a solution of 1.42 g. (6 mmoles) of ~-tyrosylglycine'~ in 60 ml. of water until the pH of the mixture was 3.2. (Addition of more resin did not cause a further change in pH). Then 3 ml. of a 4 M solution of iodine monochloride (12 mmoles) in 1 N HCl was added with stirring over a period of about 5 min. After stirring for another 30 min. the resin was collected by filtration and washed with water, then with 60 ml. of 2 N NHdOH, and again with water. The combined alkaline filtrates were concentrated to about 60 ml. When the pH was adjusted to 4, a precipitate formed which was collected after cooling and washed with ice-cold water yielding 1.66 g. (57%), m.p. 195.5-196.5' dec. Dissolution in 2 N KH4OH and reprecipitation with 2 N HCl gave 1.20 g. of small crystals (m.p. unchanged) which were dried in vacuo a t room temperature. A second crop, m.p. 184-185' dec. (0.50 g.), was obtained from the filtrate. For paper chromatography, see Table I. Anal. Calcd. for C ~ ~ H ~ Z I I N S O ~ * OC, . ~ 26.47; H ~ O : H, 2.63; I, 50.86; E,5.61. Found: C, 26.68; H , 2.64; I , 50.83; N , 5.53. Glycyl-3,5-diiodo-~-tyrosylglycine.-A solution of 1.7 g. (6.7 mmoles) of iodine and 1.6 g. of potassium iodide in 10 ml. of water was added slowly (1 hr.) to a stirred solution of 0.89 g. (3.0 mmoles) of glycyl-L-tyrosylglycine18 in 10 ml. of a 2% aqueous solution of methylamine. A 20% aqueous solution of methylamine was added dropwise as needed in order to maintain the pH of the reaction mixture between 7.5 and 9. After stirring for another hour the excess of iodine was reduced with an aqueous solution of sulfur dioxide and the p H was brought to 6.5 with 4 N HC1. The microcrystalline precipitate was washed with a small amount of ice-cold water and dried in vacuo a t room temperature yielding 1.21 g., m.p. 211-212" dec. A second crop was obtained from the mother liquors; total yield, 1.34 g. (80y0). For paper chromatography, see Table I. Anal. Calcd. for C13H1&N306.0.5Hz0: C, 28.08; H , 2.90; I , 45.64; N , 7.56. Found: C, 28.06; H , 3.13; I , 45.49; N , 7.47. Glycyl-L-thyroxine.-The reaction of glycyl-3,5-diiodo-~tyrosine (2.45 g., 5.0 mmoles) with 4-hydroxy-3,5-diiodophenylpyruvic acid (2.59 g., 6.0 mmoles) was carried out essentially as described previouslylo for the reaction of 3,5-diiodo-~-tyrosinewith the same keto acid. The solution of the keto acidl9 was freshly prepared and the extraction with I-butanol was omitted. The reaction mixture was evaporated and the residue was washed with small amounts of a saturated solution of sodium

sulfate and of ice-cold water. The crude product was purified by dissolution in 1 N NH40H and reprecipitation (pH 5) with 4 N HC1. The substance was dried in vacuo a t room temperature yielding 0.46 g. (11%), m.p. 183-185' dec. For paper chromatography, see Table I. Anal. Calcd. for Cl~H141nN20a.2H~0: C, 23.47; H , 2.09; I , 58.35; ?;, 3.22. Found: C, 23.23; H , 2.14; I , 58.11; N , 3.36. L-Thyroxylglycine .-Solid 4- hydroxy- 3,5-diiodophenylpyruvic acid (1.73 g., 4.0 mmoles) was added in small portions to a buffered solution (pH 6.7) of 3,5-diiodo-~-tyrosylglycine hemihydrate (1.47 g., 2.9 mmoles). Other reaction conditions were similar to those described previously (cf. glycyl-L-thyroxine). After reprecipitation a t pH 7 of the crude product, the peptide was dried in vacuo a t room temperature yielding 0.23 g. (9'%), m.p. 207-210" dec. For paper chromatography, see Table I. Anal. Calcd. for C ~ ~ H ~ ~ I ~ N ~ OC,~ 23.96; . H Z O H: , 1.89; I , 59.58; N, 3.29. Found: C, 23.77; H , 1.87; I , 59.10; N, 3.54. Glycyl-L-thyroxylg1ycine.-Reaction conditions were similar to those described in the preceding paragraph. Starting materials were 1.73 g. (4.0 mmoles) of 4-hydroxy-3,b-diiodophenylpyruvic acid and 1.64 g. (2.9 mmoles) of glycyl-3,5-diiodo-~tyrosylglycine hemihydrate. The crude reaction product was washed by centrifugation, then reprecipitated a t pH 4.5, and dried in vacuo a t room temperature yielding 0.18 g. (7%) of fine crystals, m.p. 198-201' dec. Anal. Calcd. for C19H1714X30~.2H~O: C, 24.62; H , 2.28; I , 54.76; N , 4.53. Found: C, 24.89; H , 2.29; I, 54.76; N , 4.76. Inspection of paper chromatograms of the substance in short wave ultraviolet light revealed the presence of a faintly fluorescent spot. The fluorescent impurity was removed by extracting a solution of the tripeptide in 1 N NaOH several times with 1butanol, washing the combined butanol extracts with water, then removing the butanol by evaporation under reduced pressure. Recrystallization of the residue gave a chromatographically pure product. For Rr values, see Table I.

.

(15) Nutritional Biochemicals Corp.. Cleveland, Ohio. (16) E. Abderhalden and M. Guggenheim, Ber., 41, 1237 (1908). (17) M.P. 259-264' dec.; after recrystallization from acetone-water, 282-283' dec.; cf. E. Abderhalden, R. Abderhalden, H. Weidle, E. Baertich and W. Morneweg [Fermentforachung., 16, 98 (1938)l and H. Zahn and K. Ziegler [ A n n . , 610, 132 (195711. (18) M.P. 239-242' dec.; cf. ref. 17 and T. Yamashita. J . Biochem. (Tokyo). 48, 651 (1960). (19) Commercially available from Osaka Laboratory of Synthetic Organic Chemicals, Nishinomiya, Japan.

3065

Synthesis of 2-(l-Nonenyl)-4-quinolinol (Pyo 111) and Some Related 2,4-Disubstituted Quinolines WILLIAMJ. GOTTSTEIN,HELENROBERTS, IBERT C. WELLS,'AND LEE C. CHENEY Research Division, Bristol Laboratories, Division of Bristol-Myers Company, Syracuse 1, New York Received May 19, 1964

I n 1945 Hays and co-workers2 reported the isolation of five closely related antibiotic metabolites of Pseudomonas aeruginosa. These substances were designated Pyo Ib, Pyo IC,Pyo 11, Pyo 111, and Pyo IV. I n 1952 Wells3 reported the structural elucidation and synthesis of Pyo I b , Pyo IC,and Pyo 111. Pyo I b and Pyo ICwere shown to be the homologous 2-heptyl-4-quinolinol and 2-nonyl-4-quinolino1, respectively. Application of the Conrad-Limpach reaction4 afforded synthetic Pyo I b and Pyo ICin yields above 25%. Structural studies featuring hydrogenation and ozonization6 provided convincing evidence that the structure of Pyo I11 was 2-(l-nonenyl)-4-quinolinol. By use of chromatography an 0.8y0yield of Pyo I11 was obtained by (1) Department of Biological Chemistry, State University of New York Medical Center at Syracuse University, Syracuse. N. Y. (2) E. E. Hays. et al., J . Biol. Chem., 169, 725 (1945). (3) I. C. Wells, ibid., 196, 331 (1952). (4) N. J. Leonard, H. F. Herbrandson, and E. M. Van Heyningen. J . A m . Chem. Soc.. 68,1279 (1946). (5) I. C. Wells. W. H. Elliott, 9. A. Thayer, and E. A. Doisy, J . Biol. Chem., 196, 321 (1952).

NOTES

3066

VOL.29

TABLE I SOME HOMOLOGS O F 2-( l-NONENYL)-4-QUINOLINOL AND

INTERMEDIATES

R

cyl

N ' R'

R

R'

COzH COzH COZH C02H COZH COzH

CHzCHzCHaa C~H19-n~ C11H23-n C(Br)HCH&H3 C(Br)HCloHZ1-n CH=CHCHsC

NO.

Ia Ib

Formula

C13HlSx02 CIQHZSNOZ CziH2&Oz Cl3Hl2BrNOZ CZ1Hz8BrS02 C13HllSOz

---Calcd., C H

70--

%-

--round.

N

C

H

N

M.P.; "C.

Yield,

%

155-158 124-125 4 12 124-127 4 95 164-166 3 05 116-117 206-208

56 90 78 48 61 68

CH=CHCgHipn CziHz7X02 77.50 8 . 3 6 '4.30 77.35 8 . 3 9 4 . 2 8 138-139 CH=CHCHI C12H1Jz .HCI 65.30 5 . 9 4 65.50 6.10 238 dec. IVC CH=CHCgHlg-n CzoHzgX2.HNO3 66.82 8 . 1 3 11.69 66.80 8 . 1 2 11.78 173-175 Va CH=CHCHa ClzHllNO 77.81 5.99 77.39 6 . 1 7 210 dec. VC -OH CH=CHCgHig-n CzoHz,_1'0 80.76 9.15 80.45 8 . 8 1 144-145 Lit.7 m.p. 156' (water). * Lit.7m.p. 145" (ethanol). Lit.g m.p. 153-155' trihydrate (ethanol).

47 27 44 20 40

IC

IIa IIc IIIa

IIIc IS'a

76 77 53 62 73

22 02 08 06 22

8 8 4 6 5

42 76 28 93 4 28 77 10 11 4 76 53 25 95 3 45 61 75 20 73 24

8 44 8 84 4 26 6 78 5 43

COzH -NH2 -NHz -OH

Recrystn. solvent

Methanol Methanol Methanol Methanol Methanol Dimethylformamide water Methanol Acetone Methanol Acetone Acetone

5

cyclization of the ani1 derived from crude methyl 3-oxo4-dodecenoate in boiling diphenyl ether. Although this synthesis served to confirm the assigned structure, attempts to improve the yield were not encouraging. Moreover, the other known routes to 4-quinolinols are likewise rendered inappropriate for a practical synthesis of Pyo I11 by virtue of its conjugated double bond which is the most distinguishing feature of the metabolite. It is noteworthy that no other examples of 2-(1alkenyl)-4-quinolinols have been found in the literature. I n the reaction sequence I b to Vb is portrayed a new general four-step synthesis of Pyo I11 (Vb). The preparation of lower and higher homologs (Ia and c to Va and c) attests to the general nature of the method, which has the added advantages of convenience, inexpensive starting materials, satisfactory over-all yields, and relatively mild operating conditions.

IIIa, R=CH3

b, R=C7H;rn C,

R=CgHig-n

2-Nonylcinchoninic acid (Ib) was readily prepared in yields above 90% from isatin and 2-undecanone by the Pfitzinger reaction.' Ziegler bromination of I b with N-bromosuccinimide produced 2-(l-bromononyl)cinchoninic acid (IIb) which was dehydrohalogenated with potassium hydroxide in t-butyl alcohol to obtain 2-(1nonenyl) cinchoninic acid (IIIb). The Weinstocks modification of the Curtius reaction was used to advantage for the conversion of acid I I I b into the corresponding amine IVb in a unit operation. Treatment of the carboxylic-carbonic mixed anhydride prepared from IIIb, ethyl chloroformate, and triethylamine with aqueous sodium azide followed by the introduction of methanol produced the corresponding methyl carbamate. The latter was hydrolyzed in situ with alkali, and the amine (IVb) was isolated as the nitrate salt. Diazotization of IVb in trifluoroac.etic acid and subsequent hydrolysis of the diazonium salt in situ gave 2(l-nonenyl)-4-quinolinol(Pyo 111, Vb) in a yield of 50-6070. The representative homologs of Pyo I11 and their intermediates selected for Table I were prepared by the same general procedures described in detail for the synthesis of compounds Ib to Vb inclusive. It is noteworthy that 2-propenylcinchoninic acid (IIIa)g was the only example of the 2-(l-alkenyl)cinchoninic acids found in the literature. It was isolated as a trihydrate in 9.8% yield as a product of the interaction of pyruvic acid, crotonaldehyde and aniline under the conditions of the Doebner-Miller reaction. Although many derivatives of 4-aminoquinoline and 4aminoquinaldine are known, none of the 2-alkenyl-4aminoquinolines has been hitherto reported.

OH Experimental'o

2-(l-Bromononyl)cinchoninic Acid (IIb) .-To a mixture of 29.8 g. (0.1 mole) of 2-nonylcinchoninic acid7 and 17.8 g. (0.1

( 6 ) R. C. Elderfield "Heterocyclic Compounds." Vol. 4, John Wiley and Sons. Inc., New York, N. Y.,1952,Chapter 1.

(7) N. P. Buu-Hal and R. Royer, J . Chem. Soc., 106 (1948). (8) J. Weinstock, J . Ore. Chem., 16, 3511 (1961). (9) M. Richter and P. Boyde, J . prakt. Chem., 9, 124 (1959): Chem. Abstr.. 64, 6715 (1960). (10) All melting pointa are uncorrected.

OCTOBER, 1964

NOTES

mole) of N-bromosuccinimide in 500 ml. of carbon tetrachloride there was added 100 mg. of benzoyl peroxide. The mixture was stirred and heated under reflux on the steam bath for 1 hr. in the presence of two No. 2 Photoflood lamps. The succinimide was collected by filtration. Evaporation of the solvent from the filtrate left a tan solid residue which was recrystallized from methanol to yield 15 g. (3Yyc) of pale yellow crystals, m.p. 125126". Anal. Calcd. for C19H24BrN02:C, 60.32; H , 6.40; N, 3.70. Found: C,60.28; H,6.35; Yj, 3.61. 2-(l-Nonenyl)cinchoninicAcid (IIIb) .-To 25 g. (0.066 mole) of 2-(l-bromononyl)cinchoninicacid there were added 50 g. of finely ground potassium hydroxide and 125 ml. of t-butyl alcohol. The mixture was stirred and heated under reflux for 1 hr., during which time a viscous mass of salt had formed on the bottom and sides of the flask. The mixture was diluted with 100 ml. of water and acidified to pH 2 with dilute sulfuric acid. The solid was collected by filtration and washed well with water to remove the residual salts. Recrystallization from methanol gave 16.2 g. (825;) of light yellow crystals, m.p. 137-138'. Anal. Calcd. for C1,H2,N02: C, 76.73; H, 7.80. Found: C,77.00; H , 7.94. 4.-Amino-2-(l-nonenyl)quinoline (IVb) .-To a solution of 2.97 g. (0.01 mole) of 2-(1-noneny1)cinchoninic acid and 1.38 ml. (0.01 mole) of triethylamine in 25 ml. of tetrahydrofuran a t 5" there was added dropwise 1.4 g. (0.013 mole) of ethyl chloroformate. The mixture was stirred for 0.5 hr. while a solution of 0.86 g. (0.013 mole) of sodium azide in 20 ml. of water was added dropwise. After stirring an additional hour a t S o , the solution was poured onto 100 g. of cracked ice and the azide was extracted into ether. .4 few milliliters of saturated sodium chloride solution was added to break up the resulting emulsion. The organic layer was separated and dried over anhydrous magnesium sulfate. Anhydrous methanol (10 ml.) was added to the filtered solution and the ether was removed under reduced pressure. The residue was boiled for 2 hr. with 10 ml. of anhydrous methanol in 35 ml. of benzene. The benzene was removed under reduced pressure and the carbamate was heated under reflux for 5 hr . in aqueous methanolic potassium hydroxide prepared from 2 8 . of potassium hydroxide pellets, 20 ml. of methanol, and 5 ml. of water. The mixture was diluted with .io ml. of water and the amine was extracted into ether. The ether was removed under reduced pressure and the residue was dissolved in 20 ml. of acetone. The pH was adjusted to 2 by the addition of dilute (1 : 1) nitric acid. The amine salt was collected by filtration and washed with acetone to yield 1.07 g. (33y0) of off-white crystalline 4amino-2- ( 1-nonen yl) quinoline nitrate, m .p . 163-1 64 O . Anal. Calcd. for Cl8H25X303: C, 65.23; H , 7.60; N, 12.68. Found: C,65.58; H, 7.70; N,12.63. 2-(l-Nonenyl)-4-quinolinol(Vb, Pyo 111).-To a light yellow solution of 2.64 g. (0.08 mole) of 4-amino-2-(l-nonenyl)quinoline nitrate dissolved in 20 ml. of trifluoroacetic acid a t 5' there was added 0.52 g. (0.08 mole) of sodium nitrite. After the deep red solution was stirred for 0.5 hour a t 5", 7 ml. of water was added all a t once, and the solution was stirred for an additional 0.5 hour. The mixture was poured into 50 ml. of water and extracted with ether. The ether layer was separated, washed with water, and evaporated under reduced pressure to obtain a red oil. The oil was dissolved in 25 ml. of acetone and diluted with a saturated aqueous sodium carbonate solution to pH 9. The tan crystals were collected by filtration, washed with water, and dissolved in 20 ml. of methanol. After a carbon treatment, the filtrate was evaporated to obtain a yellow crystalline solid. Recrystallization from acetone yielded 1.2 g. (56%) of pale yellow needles, m.p. 151-152". Anal. Calcd. for C18Y23NO: C, 80.25; H , 8.61; N, 5.20. Found: C, 80.05;H, 8.60; N,5.08. The infrared spectrum was identical with the spectrum of authentic 2-(l-nonenyl)-4-quinolinol (Pyo 111).2 The trans character of the double bond in the nonenyl side chain was clearly shown by the strong band a t 965 cm. -I which is characteristic of a trans-disubstit I ethylene." Likewise the other compounds bearing alkex. -.oups showed a similar absorption band. Lack of absorption in the region of 690 cm.-' indicated the absence of cis isomer in every case.

Acknowledgment.-The authors wish to thank Richard 1 4 . Downing for the microanalyses and Donald Evans and David Whitehead for the infrared spectra.

(11) L. J. Bellamy, " T h e Infrared Spectra of Complex Molecules," 2nd Ed., John Wiley a n d Sons, Inc., New York, N. Y. 1958, p. 45.

3067

A Convenient Synthesis of Tricyclic 2- Quinolizidones' LEONMANDELL, B. A. HALL,A N D K. P. SINGH Departmat of Chemistry, Emory University, Atlanta, Georgia 30322 Received M a y 12, 1964

We have previously shown the utility of our quinolizidine synthesis in the preparation of matrine,2 niatridine,3 and various tricyclic systems with nitrogen a t a bridgehead.4 This work reports a modification of that scheme which allows its application to the synthesis of 2-quinolizidones typified by 11. This new route is outlined below.

zCH-CN

0

",,A

% 2. CHZ=CH-COpEt OI

I

NC

/R%

IV

I11

::

I1

The steps are all unexceptional and follow closely the basic pathway we investigated earlier. Key in the synthesis is the Stork enamine synthesis5 for the preparation of I and the reductive cyclization of I to provide 11. We have discussed the latter reaction previously.4 The structure and stereochemistry of the quinolizidone (11) were established by lithium aluminum hydride reduction of I1 to a mixture of hexahydrojulolidines, which consisted of 60% I11 and 40% IV. Experimentala Ethyl P-[2-( l-Oxo-6-cyanoethyl)] cyclohexylpropionate (I).The pyrrolidine enamine was prepared by refluxing 35 g. (0.23 mole) of the above ketone with 57 g. (0.81 mole) of pyrrolidine in 200 ml. of anhydrous benzene. After 18 hr. the theoretical amount of water had been collected in an azeotrope separator, The excess pyrrolidine and benzene were removed in vacuo and (1) We wish t o acknowledge t h e support of this research b y t h e National Institutes of Health through Research G r a n t RG-7902. This work waB taken in p a r t from t h e M.S. Dissertation of B. A. Hall a n d t h e P h . D . Dissertation of K. P. Singh. (2) L. Mandell. K. P. Singh, J. T. Gresham, a n d W. Freeman, J. Am. Chem. Soc., 86, 2682 (1963). (3) L. Mandell a n d K. P . Singh, ibid., 88, 1766 (1961). (4) L. Mandell, J. U. Piper, a n d K. P. Singh, J . O r g . Chem., 88, 3440 (1963). ( 5 ) G . Stork, A. Brimolere, H. Landesman, J . Szmuszkovicz, a n d R. Terrell, J. Am. Chem. Soe., 86, 207 (1963). (6) Melting points a n d boiling points a r e uncorrected. Elemental analyses were done b y Drs. G. Weiler a n d F. B. Strauss, Oxford, England.

NOTES

3068

the crude enamine was taken up in 200 ml. of N,N-dimethylformamide. Twenty-three grams (0.23 mole) of ethyl acrylate was added dropwise to the enamine solution a t reflux and the reaction mixture was refluxed for 40 hr. The solvent was removed in oacuo on a steam bath and the enamine was hydrolyzed by refluxing for 2 hr. with 150 ml. of water and 8 ml. of glacial acetic acid. After cooling, the mixture was extracted thoroughly with ether and the ether extract was washed with dilute acid, base, and water, and finally dried over sodium sulfate. The ether was removed and the residue was distilled to afford 18.6 g., b.p. 160-162" a t 1 mm. Allowing for recovered starting material the yield waa 63%. A n a l . Calcd. for ClrHzlNOa: C, 66.9; H , 8.4; N , 5.6. Found C, 66.9; H, 8.5; N, 5.6. 3-Oxohexahydrojulolidine (11).-Twenty grams (0.80 mole) of I was hydrogenated in 100 ml. of absolute ethanol a t 1800 p.s.i. and 125' using 7 g. of W-7 Raneynickel as catalyst. The catalyst was removed by filtration, the solvent was removed in vucuo, and the residue was distilled, giving 8.3 g. (54%) of 11, b.p. 144-150' a t 0.1 mm. A n a l . Calcd. for C12HlQNO: C, 74.5; H, 9.9; N, 7.2. Found: C, 74.2; H, 9.9; N , 7.1. Reduction of I1 to Hexahydrojulo1idine.-To a solution of 1.2 g. of powdered lithium aluminum hydride in 40 ml. of dry tetrahydrofuran was added 6.1 g. of I1 in 20 ml. of tetrahydrofuran. The reaction was refluxed 4 hr. and then treated first with ethyl acetate and finally with water to destroy the unused hydride. The mixture was filtered, the gelatinous precipitate was washed well with tetrahydrofuran, and the filtrate and washings were combined and dried over sodium sulfate. Removal of solvent and distillation gave 3.6 g. (727,) of hexahydrojulolidine, b.p. 110-120" a t 5 mm. Three grams of this product was chromatographed following the procedure of Bohlmann7 to give 1.5 g. of 111, picrate m.p. 225226", and 1.0 g. of IV, picrate m.p. 186-187". (7) F. Bohlmann a n d C. A r n d t , Chem. Ber.. 91,2167 (1958).

VOL. 29

2 ) . For preparative use this reaction has the disadvantage that the quinol ether is often difficult to separate from the simultaneously formed I. Reoxidation of I to I1 with potassium ferricyanide in this reaction has been practiced, but this procedure is very t e d i o ~ s . ~ J Another method of preparing quinol ethers of this type is the reaction of 4-bromo-2,4,6-tri-t-butyl-2,5cyclohexadien-1-one with alkali phenolate^.^ The reaction can be carried out in an organic solvent; however, separate preparation of the starting materials from the phenols is required, Our interest in oxidations with active manganese dioxides has now led to a very convenient method for the preparation of quinol ethers. We observed that 2,4,6-tri-t-butylphenol (I) in benzene solution in contact with active manganese dioxide instantly forms the blue solution of 2,4,6-tri-t-butylphenoxyl(11) in apparently high yield. Successive additions of 1 mole of a phenol (IIIa-h) per mole of 2,4,6-tri-t-butylphenol (I) to the I\InOz-containing radical solution leads to irreversible discharge of the blue color. Since the 2,4,6tri-t-butylphenol formed in the reaction is immediately reoxidized by the manganese dioxide present , the isolation of the quinol ether is no longer difficult. Filtration and evaporation of the solvent give the quinol ethers (IVa-h) in excellent yield and high state of purity. We have used this method to prepare quinol ethers previously not reported (see Table I), including the reaction product of 2,4,6-tri-t-butylphenoxylwith the unsubstituted phenoxy radical. TABLE I

Preparation of Quinol Ethers HANS-DIETERBECKER

General Electric Research Laboratory, Schenectady, New York Received M a y 19, 1964

Chemical reactions of radicals obtained by oxidation of sterically hindered phenols such as 2,4,6-tri-t-butylphenol (I) have been widely studied during recent years.'-3 An interesting result was reported by Muller, Ley and Schlechte, who found that 2 moles of 2,4,6-tri-t-butylphenoxyl(11) react with 1 mole of a phenol (111) to form 1 mole of 2,4,6-tri-t-butylphenol (I) and 1 mole of a 4-quinol ether (IV)4(reactions 1 and

R I

I1

c!

IIIa-h

IVa-h

(1) E. Mtiller, A. Riecker, and K. Schefder, Ann. Chem., 646, 92 (1961), and previous papers. (2) C. D. Cook and N. D. Gilmour, J. 370. Chem., ,S. 1429 (1960),and previous papers. (3) For a recent review, Bee H. Mueso. Anoew. Chem., 76, 965 (1963). (4) E.MUller, K, Lby, a n d G. Schlechte, Chem. Ber., $0, 2660 (1957).

Ar in I11 and I V

Quinolid bandR in I V (om.-])

a, Phenyl b, 2,4,6Trichlorophenyl c, Pentachlorophenyl d, 4-Phenylmercaptophenyl e, 4-Methoxyphenyl f , 4-Benzoyloxyphenyl g, 4-Benzyloxyphenyl h, 4Phenoxyphenyl Liquid film. * I n potassium bromide.

1645, 1665" 1645, 1666* 1642, 1664b 1645, 1665b 1636, 1662b 1640, 1662b 1640, 1662b 1645, 1665"

The infrared spectra (see Table I) of all quinol ethers prepared exhibit double peaks a t about 6 k , typical for compounds with quinolid structure. Compared with known procedures for quinol ether formation, this new method appears to be the most convenient, since the starting materials are easily available, the reaction is carried out in an organic solvent, and the procedure does not involve any separation problems. When we extended the reaction of phenols with active manganese dioxide to the cooxidation of 4bromo-2,6-di-t-butylphenol (V) and pentachlorophenol (reaction 3), we isolated in excellent yield the aromatic quinone ketal VII, 4,4-bis(pentachlorophenoxy)-2,6di-t-butyl-2,5-cyclohexadiene-l-one. Its structure is confirmed by analytical data (see Experimental) , its infrared spectrum (bands at 1655 and 1680 cm.-', In KBr), and its conversion into pentachlorophenol by treatment with and 2,6-di-t-butyl-benzoquinone-l,4 boiling 96% ethanol. (6) T. Matauura and H. J. Cahnmann, J. A m . Chem. Soc., 82,2055 (1960). (6) E.Adler and H.-D. Becker, Acta Chem, Scand., 16, 849 (1961); H.-D. Becker. t b i d . , 16, 683 (1961); 16, 78 (1962).

OCTOBER,1964

NOTES

3069

cases a slightly higher ratio than 1: 1 of the phenol (111) to tri-tbutylphenol is required for complete diiappearance of the blue ?H OH radical. Recrystallization of the quinol ethers did not cause any major losses of the materials. 4-Phenoxy-2,4,6-tri-t-butyl-2,5-cyclohexadiene-l-one (IVa) .2,4,6-Tri-t-butylphenol (2.62 g., 10 mmoles) was added to a suspension of 12 g. of active MnO2 in 100 ml. of benzene and the Br c1 mixture was shaken under N2 for 1 min. One gram of phenol c1 'Cl (10.6 mmoles) was then added in five portions ovir a perioh of 30 V min. Between each addition the mixture was shaken for a VI few minutes. After the last addition of phenol, the blue color of the solution disappeared. The manganese dioxide was removed by filtration using a fine sintered-glass funnel, and the yellow filtrate was evaporated a t room temperature under vacuum, yielding 2.75 g. of IVa as a yellow oil, yield 73%. Anal.13 Calcd. for Cz4H&: C, 82.49; H, 9.05; mol. wt., (3) 378.53. Found: C, 82.41; H, 9.42; mol. wt., 336 (thermoelectric measurement in dioxane). 4-( 2,4,6-Trichlorophenoxy)-2,4,6-tri-t-butyl-2,5-cyclohexadic1 c1 c1 ene-1-one (IVb) .-2,4,6-Tri-t-butylphenol (524 mg., 2 mmoles was added to a suspension of 5.2 g. of active MnOz in 100 ml. of VI1 benzene. After shaking for 1 min., 400 mg. of 2,4,6-trichlorophenol (2.04 mmoles) was added in four portions (shaking for 1 Upon addition of 2,4,ti-tri-t-butylphenol to a freshly min. between each addition). After the last addition, the soluprepared, colorless solution of VI1 in benzene, the blue tion still showed a slightly greenish color. Further addition of 20 color characteristic of 2,4,6-tri-t-butylphenoxyl is mg. of 2,4,6-trichlorophenol gave a yellowish solution. The mixformed after a few seconds. This indicates that the ture was filtered as described under IVa; the very slightly greenish filtrate was evaporated a t room temperature in vacuo, yielding aromatic quinone ketal dissociates into phenoxy 850 mg. of a slightly yellowish green crystalline residue (93?&), radicals (reaction 4) that can oxidize 2,4,6-tri-tm.p. 122-123'. It was recrystallized from hot ethanol; m.p. butylphenol. 127-1 29 O . Anal. Calcd. for C24HalClsO~ (457.86): C, 62.96; H , 6.82; C1, 23.23. Found: C, 62.94; H, 6.95; CI, 23.40. 4-(Pentachlorophenoxy)-2,4,6-tri-t-butyl-2,5-cyclohexadiene-lone (IVc).-This compound was prepared similarly to IVb from pentachlorophenol; yield 90%, m.p. 157-158" (recrystallized from a hot mixture of ether-ethanol, 8:2), greenish-yellow + crystals. VI1 r c1 clQcl c1 (4) Bnal. Calcd. for Cz4H28C1502 (526.74): C, 54.73; H, 5.55; C1, 33.65. Found: C, 54.51; H, 5.57; C1, 33.44. CI 4-(4-Phenylmercaptophenoxy)-2,4,6-tri-t-butyl-2,5-cyclohexc1 CI c1 cl+c' adiene-1-one (IVd) .-This compound was prepared in the same manner as IVb from 2,4,6-tri-t-butylphenol and p-hydroxydiphenyl thioether. The light greenish-colored oil obtained Oxidation of V alone with i\lnOp gives 3,5,3',5'after evaporation of the filtrate crystallized after treatment with a little ethanol; yield 86%, m.p. 112-113", yellowish crystals. tetra-t-butyl-diphenoquinone, while oxidation of pentaAnal. Calcd. for CaoH8,OpS (462.60): C, 77.89; H , 8.28; chlorophenol with b h O 2 affords the previously deS, 6.91. Found: C, 77.85; H, 8.57; S, 7.10. scribed 2,3,4,5,6-pentachloro-4-pentachlorophenoxy-2,5- 4- (4-Methoxy-phenoxy)-2,4,6-tri-t-butyl-2,5-cyclohexadiene1 cyclohexadiene-1-one.? We presunie that the formation one (IVe).-This was prepared in the same manner as IVb from of VI1 involves the quinol ether VI which could undergo 2,4,6-tri-t-butylphenol and 4-methoxyphenol. The green-colored filtrate yielded a green oil (91%) which formed yellow crystals either a nucleophilic* or radical displacement reaction upon treatment with a little ethanol; m.p. 74-75'. It was reto yield VII. crystallized by dissolving in a little ether and addition of a little The formation of a mixed aliphatic-aromatic quiethanol; m.p. 77-78' (slightly yellow crystals). none ketal by oxidation of substituted hydroquinone Anal. Calcd. for CZ5H36O3(384.54): C , 78.08; H , 9.44. monoalkyl ethers has previously been r e p ~ r t e d . ~ Found: C, 78.31; H, 9.67. 44 4-Benzyloxy-phenoxy)-2,4,6-tri-t-butyl-2,5-cyclohexaAside from one example of a "dinier polyhalogen diene-1-one (IVf).-This substance was prepared as described aroxyl''lo obtained by oxidation of 4-pentachloroabove from 2,4,6-tri-t-butylphenol and hydroquinone monobenphenoxy-2,3,5,6-tetrachlorophenol,no aromatic quizoate; yield 87%, yellow needles, m.p. 132-133". I t was renone ketals have been reported previously. crystallized by addition of ethanol to a warm ether solution. Anal. Calcd. for C31H3804 (474.61): C, 78.45; H, 8.07. Found: C, 78.25; H, 8.08. Experimental11 4-(4-Benzyloxy-phenoxy) -2,4,6-tri-t-butyl-2,5-cyclohexadiene1-one (IVg) .-This substance was prepared as described above "Active manganese dioxide" was prepared according to from 2,4,6-tri-t-butylphenol and hydroquinone monobenzyl Attenburrow,12 el al. Benzene was dried over sodium. All exether. The slightly greenish filtrate left an oil after evaporation periments were carried out in screw-cap bottles which were of the solvent a t room temperature under vacuum. Addition of flushed with nitrogen whenever opened. Depending on the a little ethanol caused rapid crystallization; yield 71TG, slightly quality of the MnOs, phenols can be strongly adsorbed, thus preyellowish crystals. It was recrystallized by dissolving in warm venting their reaction with the tri-t-butylphenoxyl. In these ether and adding ethanol; m.p. 104-105". Anal. Calcd. for C31H4003 (460.63): C, 80.83; H, 8.75. (7) R . Reed, J r . , J . Am. Chem. Soc., 80,219 (1958). (8) T h e active manganese dioxide is slightly alkaline. Found: C, 80.89; H, 8.96. (9) C. Martius a n d H. Eilingefeld, Ann. Chem.. 607, 159 (1957). 4-(4-Phenoxy-phenoxy)-2,4,6-tri-1-butyl-2,5-cyclohexadiene-l(IO) E. Muller, A. Rieker. and W. Beckert. Zeif Naturforsch., 17b, 567 one. (IVh).-2,4,6-Tri-t-butylphenol (1.31 g., 5 mmoles) was (1962). added to a suspension of 13 g. of active manganese dioxide in 125 (11) Melting points were taken on a Fisher-Johns a p p a r a t u s and are n o t

1

L

b

-

corrected. (12) J. Attenburrow, A. F. B. Cameron, J. H. C h a p m a n , R . N. E v a n s , R . A . Hems, A . I3. A. Jansen, and T. Walker, J . Chem. Soc., 1094 (1952).

(13) Analyses were carried o u t b y Schwarzkopf Microanalytical Laboratory, Woodside 77. N. Y.

NOTES

3070

ml. of benzene, and 930 mg. of p-phenoxyphenol (5 mmoles) was added in four portions as described above. Filtration and evaporation of the yellow solution gave IVh as a yellow oil, yield 1.85 g. (82%). A n a l . Calcd. for C ~ O H ; {C, ~ O80.68; ~ H, 8.59; mol. wt., 446.60. Found: C , 80.50; H, 8.71; mol. wt., 436 (thermoelectric measurement in dioxane). 4,4-Bis (pentachlorophenoxy)-2,6-di-t-butyl-2 ,5-cyclohexadiene1-one (VII).-Active manganese dioxide (26 9.) was added to a solution of 2.66 g. of pentachlorophenol (10 mmoles) and 1.43 g. of 4-bromo-2,6-di-t-butylphenol in 300 ml. of benzene. The suspension was shaken under NZ for 30 min. The MnOz was removed by filtration through a fine sintered-glass funnel and washed with 200 ml. of benzene. Evaporation of the light yellow filtrate gave a light yellow, crystalline residue which was washed with 150 ml. of absolute ethanol; yield 3.05 g. (83%), m.p. 155". The substance was recrystallized by dissolving in cold chloroform. Addition of absolute ethanol gave fine, colorless needles of VII, m.p. 166-167" dec. A n a l . Calcd. for CnsHzoCl,OOs: C, 42.49; H, 2.72; C1, 48.25; mol. wt., 734.80. Found: C, 42.24; H , 2.82; CI, 48.33; mol. wt., 698 (thermoelectric measurement in chloroform). Hydrolysis of VI1.-A suspension of VI1 (500 mg.) in 25 ml. of 967, ethanol was refluxed for 45 min. The clear, yellow solution formed was evaporated in vacuo, yielding a semisolid yellow residue. Treatment with 5 ml. of cold petroleum ether (b.p. 3&60") and filtration gave 305 mg. (84Y0) of pentachlorophenol, m.p. 189-190'. Mixture melting point with authentic material gave no depression. Evaporation of the filtrate gave a yellow oil that formed yellow crystals of 2,6-di-t-butylbenzoquinone; yield 110 mg. (73W), m.p. 67-68" m.p. 67-68'). The infrared spectrum was identical with that of authentic material. Oxidation of 4-bromo-2,6-di-t-butylphenol.-Asolution of 2.85 g. of 4-bromo-2,6-di-t-butylphenol (10 mmoles) in 175 ml. of benzene was shaken with 20 g. of active manganese dioxide for 35 min. The manganese dioxide was removed by filtration and washed with 50 ml. of benzene. Evaporation of the dark brown filtrate gave a solid, brown residue that was treated with 20 ml. of methanol, filtered, and dried at 105". The yield of 3,3',5,5'tetra-1-butyldiphenoquinone was 1.9 g. (927,). The substance formed red-brown crystals upon heating; m.p. 245-246" (lit.15 m.p. 245-247'). Mixture melting point with authentic material gave no depression. Oxidation of Pentachlorophenol.-A solution of 2.66 g. of pentachlorophenol (10 mmoles) in 175 ml. of benzene was shaken with 20 g. of active manganese dioxide for 35 min. The yellow filtrate obtained after removal of the manganese dioxide was evaporated a t room temperature in vacuo, yielding a solid yellow residue that was washed with 20 ml. of ethanol. Filtration gave 1.5 g. of yellow crystals having m.p. 165-167". The substance was dissolved in a little benzene, cold methanol was added, and the mixture was kept a t -20" a few minutes. Then most of the solvent was rapidly removed in vucuo, yielding 1.2 g. (457,) of 2,3,4,5,6-pentachloro-4-pentachlorophenoxy-2,5-cyclohexadiene-l-one, m.p. 177-178" (lit.' m.p. 177-178"). Its infrared spectrum was completely identical with that of authentic material.

VOL.29

catalyzed self-condensation. The base-catalyzed selfcondensation of the two isoniers of phthalide, coumaranone (11) and isocoumaranone (III),had been reported earlier.435

0 I1

0

I

I11

Treatment of phthalide in dimethyl sulfoxide with potassium t-butoxide a t room temperature leads to a dimer (IV) that can be isolated in 75y0yield. Upon being heating above its melting point, IV readily eliminates water to yield the stilbene V (phthalanylidenephthalide). This product is formed directly if the selfcondensation of phthalide is carried out a t 65 '. On the

C II 0

II 0

IV

V

basis of its ultraviolet spectroscopical comparison with the known biphthalide the trans structure can be assigned to V (Fig. 1). The niechanism of the formation of IV probably involves the phthalidyl carbanion (Iaj that undergoes an ester condensation which is followed by intramolecular acetalization.

0 Ia

(14) E. MPller and K. Ley, Chem. Ber.. 88, 601 (1955). (15) H. H a r t a n d F. 4.Cassia, Jr., J . A m . Chem. SOC.,73, 3179 (1951).

Self-condensation of Ph thalide HANS-DIETER BECKER General Electric Research Laboratory, Schenectady, New York Received M a y 1 , 1964

Phthalide (I) is known to react with aldehydes or esters in the presence of base to yield aldol condensation It has now been products or phthalidyl observed that phthalide also can undergo a base(1) R. L. Horton and K. C. Murdock, J . Org. Chem., 9 5 , 938 (1960). (2) H. Zimmer and R. D. B a r r y , tbrd.. Z7, 1602 (1962). (3) W. Wislicenus, Ber.. 20, 2061 (1887).

When the base-catalyzed reaction of phthalide was extended to 3-alkoxy-substituted phthalides (phthalaldehydic acid pseudo esters) in dimethyl sulfoxide or dimethylformaniide, the only crystalline product was a colorless substance that analyzed for CleH804(formed in 15% yield). The color produced by H~SOK-KNO~, the solubility properties, and the melting point (3350j6 were all in agreement with biphthalide (VI). However, biphthalide has been reported to form yellow crystal^.^ It has been assigned the trans structure when its syn(4) K. Fries and W. Pfaffendorf. %hid., 44, 114 (1911); see also W. Baker and R. Banks, J . Chem. Soc., 279 (1939). ( 5 ) J. N. Chatterjee, J . Indian Ckem. Soc., 93, 175 (1956). (6) See F. K. Beilstein, "Handbuch der organischen Chemie," 4 t h Ed., Vol. 19, p. 176; Sui.ipl. I , p , 6 8 8 ; Suppl. 11. p. 192. (7) E. .Idor. Ann., 164, 229 (1872): P . Karrer, et ai., Helu. Cham. Acta, 11, 233 (1928); M. Pailer, H. Woerther, a n d A. Meller, Monatsh., 02, 1037 (1961).

OCTOBER,1964

NOTES

9

307 1

F1II COLORLESS BIPHTHALIOE _.. . BIPHTHALIDE TRANSPHTHALANYLIDENEPHTHPLIDE N ACETONITRILEI

VI

VI1

thesis from trans-bifurandione was accomplished.8 More recently a very convenient new synthesis of (yellow) biphthalide having a melting point of 352354" was announced.y This product had been found to be identical with the biphthalide for which the trans structure had been proved. We have prepared the yellow (ie., trans) biphthalide according to the reported9 procedure and compared it with our colorless compound. The mixture melting point was between 335 and 345", not conclusive evidence for the presence of two different compounds. However, the infrared spectrum of the yellow biphthalide was definitely different from that of the colorless biphthalide. While trans-biphthalide exhibits only one carbonyl peak a t 1780 cm.-' (in KBr), the colorless biphthalide shows a split carbonyl absorption a t 1720 and 1740 cm.-l (in KBr). The ultraviolet spectra (Fig. 1) of the known transbiphthalide and the colorless biphthalide are similar, but show significant differences. However, the differences in the ultraviolet spectra of the two biphthalides are equivalent to those between the spectra of 295 mp (emax 27,000 in ethanol)] trans-stilbene [A,, and cis-stilbene [A,, 280 nip (emax 13,500 in ethanol) 1. Therefore, one can assign the cis structure (VII) to the colorless biphthalide. The generic relationship between the yellow and the colorless biphthalide was demonstrated by conversion of the trans-biphthalide into its colorless isomer by treatment with warm, concentrated sulfuric acid. Attempts to isomerize the cis-biphthalide into its trans isomer have not been successful." This is in agreement with the findings on the isomerization of transbifurandione into cis-bifurandione, which itself could not be isomerized.* Concerning the mechanism for the formation of biphthalide, the steps involving carbanion formation (l), displacement of the methoxy group (2), and elimination (3) involving a second carbanion intermediate VI11 appear to represent a reasonable route. The stereochemistry of the carbanion VI11 probably is responsible for the formation of the cis rather than the trans isomer. The original older literature on biphthalide actually discloses that "completely colorless" biphthalide had been obtained earlier.'* "Pure biphthalide" ( i e . , colorless) had been obtained by treatment of 3-ethoxy(8) J. C. Sauer, R. D. Cramer. V. A. Engelhardt, T. A. Ford, H. E* Holmquist, and B. W .Howk, J. A m . Chem. Soc., 81,3677 (1959). (9) F. Ramirez, H.Y a m a n a k a , and 0. H. Basedow, i b i d . , 88, 173 (1961). (10) ( a ) A. E. Gillam and E. S. S t e r n , "An Introduction to Electronic Absorption Spectroscopy in Organic Chemistry," Edward Arnold, Publishers, Ltd., London, 1955, PP. 233,234; (b) H. Suzuki, Bull. Chem. Soc. Japan, 33, 379 (1960). (11) T r e a t m e n t of colorless biphthalide with warm, concentrated sulfuric acid led to quantitative recovery of the starting material. (12) J. Wislicenus, Ber.. 17, 11, 2178 (1884).

. > U

'200

220

U

240

ZEO

260

300

-

320

310

'

' I 360

'

3bOYk

'

A,mp.

k

0

'

Figure 1.

phthalide with potassium cyanide in 90% ethanol,l8 and in absolute ethan01.l~ It was later pointed out that this reaction yields a completely colorless biphthalide (yield IS%), while other procedures give 'hiore or less yellow biphthalide which is very difficult to decolorize completely."'6 These findings apparently had been overlooked whenever the yellow form of biphthalide was reported later.

II 0

0

0 II

- RO-

0

8

b

a

y

-RO-

(3)

VI11 Experimental Dimethyl sulfoxide was dried over calcium hydride and distilled a t about 1 mm. Potassium t-butoxide was a commercial product. All condensation experiments were carried out by agitating the reaction mixture with a stream of nitrogen. Melting points were taken on a Fisher-Johns apparatus and are uncorrected. Analyses were carried out by Schwarskopf Microanalytical Laboratory, Woodside 77, N . Y . Self-condensation of Phthalide (IV).-Potassium t-butoxide (2.8 g., 23 mmoles) was added to a solution of 2.68 g. of phthalide (20 mmoles) in 100 ml. of dimethyl sulfoxide. The reacation mixture was kept a t room temperature by means of an ice hath, and stirred by a stream of nitrogen. After 1 hr., 50 ml. of ice(13) C. Graebe and P. Juillard, A n n . , 1149, 219 (1887). (14) G. Goldschmidt and L. Egger, Monalah., l a , 49 (18gi). (15) C. Graebe and A. Landriset, Eer., a4, 2296 (1891).

NOTES

3072

water was added to the dark red solution. Acidification with 50 ml. of 0.5 N HC1 gave a fine crystalline, colorless precipitate, m.p. l55', yield 2 g. (75%). Recrystallization from warm chloroform raised the melting point to 163-165'. The wbetance turned yellow upon melting. Anal. Calcd. for CL6H1204(268.26): C, 71.63; H , 4.51. Found: C, 71.41; H, 4.37. 3-Phthalanylidenephthalide (VI.-Phthalide (2.68 g., 20 mmoles) in 10 ml. of dimet.hy1siilfoxide was added to a solution of 2.16 g. of sodium methoxide (40 mmoles) in 35 ml. of dimethyl sulfoxide a t 6C-65". The solution turned orange, and after a few minutes brown-red. The reaction mixture was kept a t 65" for 25 min. Then about half of the solvent was removed by distillationat about 1 mm. and 65' bath temperature. Addition of 25 ml. of ice-water and 20 ml. of concentrated hydrochloric acid gave a cloudy solution that was twice extracted with 50 ml. of chloroform. The organic layer was separat'ed and the solvent was removed by distillation. To remove traces of water and dimethyl sulfoxide, the oily residue was subjected to vacuum distillation a t 1 mm. and 70' bath temperature. Treatment of the yellow, oily residue with methanol gave yellow crystals that exhibit a greenish fluorescence, m.p. 225-227*, yield 1.06 g. (&?yo). The substance can be recrystallized from hot chloroform. It was only slightly soluble in methanol. Anal. Calcd. for C16HLa03 (250.24): C, 76.79; H , 4.03. Found: C, 76.90; H , 4.18; mol. wt., 287 (thermoelectric measurement in dioxane). V from 1V.-IV, 250 mg., was placed in a test tube and heated for 15 min. a t 185". The colorless substance turned yellow with evolution of gas. The solid yellow residue was washed with a mixture of chloroform and methanol, yielding 210 mg. of V (goyo), m.p. 225-227'; mixture melting point with V obtained by direct self-condensation of phthalide showed no depression. cis-Biphthalide (VII).-Potassium t-butoxide (2.4 g., 21 mmoles) was added to a solution of 3.3 g. of phthalaldehydic acid pseudo methyl ester (3-methoxyphthalide, 20 mmoles) in 50 ml. of dimet,hylformamide. The deep red solution was kept under nitrogen for 20 min. Addition of 50 ml. of ice-water and slow addition of 100 ml. of 0.5 N HCl gave a colorless precipitate (needles) in 400-mg. yield (15y0). The substance can be recrystallized from boiling acetic acid; m.p. 334435'. It exhibited a bright, blue fluorescence. (trans-Biphthalide fluoresces yellow.) Anal. Calcd. for C1~Hs04(264.22): C, 72.73; H, 3.05. Found: C, 72.84; H , 3.08; mol. w t . 286 (osmometric in benzene). Extraction of the aqueous filtrate with chloroform yielded 250 mg. of yellow needles besides an unidentifiable resin. The infrared spectrum of the yellow crystalline material indicated a mixture of cis- and trans-biphthalide. Conversion of trans-Biphthalide into cis-Biphtha1ide.-transBiphthalide (264 mg., 1 mmole) was dissolved in 30 ml. of concentrated sulfuric acid a t 110" and kept a t this temperature under nitrogen for 1 hr. Addition of about 80 g. of ice gave an almost colorless precipitate. It was recrystallized from hot acetic acid, giving 240 mg. of very faintly yellow needles, m.p. 320'. According t o ultraviolet and infrared spectra the product was cisbiphthalide contaminated with about 20y0 trans-biphthalide.

Dianhydrophenylosazone Acetates from Isomaltose and Gentiobiose H. EL KHADEM' Department of Chemistry, The Ohio State University, Columbus, Ohio 43210 Received April 3, 1964

Boiling acetic anhydride convert^^,^ monosaccharide phenylosazones into dianhydrophenylosazone acetates (I) Chemistry Department, Faculty of Science, Alexandria University, Alexandria. Egypt, U. A. R . (2) H . El Khadem and M. hl. Mohammed-Ali, J . Chen. Soc., 4929 (1963).

VOL.29

which have a pyrazole ring structure. The structure I was established for the product from any D-hexose precursor by transhydrazonation experiments and, after 0-deacetylation, by oxidation to give known pyrazole derivatives. Similarly, the dianhydrophenylosazone acetates from the 6-deoxy-~-hexosesand the pentoses were assigned the structures I1 and 111, respectively. This work describes theunambiguous characterization of I, 11, and I11 by n.m.r. spectroscopy and also reports the conversion of the disaccharides isomaltose and gentiobiose into the corresponding dianhydrophenylosazone acetates (IV and V, respectively). CH=N -N(Ac)-Ph

CH=N -N( Ac)- Ph

L=N

C=N,

1

I

>N-Ph

cH=e

I

,N-Ph

CH=C'

-0Ac

H-

2I%!(

CH=N -N(Ac)-Ph

I

-0Ac I

CHzOAc I11

I1

CH=N-N(Ac)

-Ph

CH=N-N( I

Ac) -Ph

C=N,

1

,N-Ph

CH=C

I

H-C-OAc

I

I Ac

A OAc IV

v

The n.m.r. spectra of the compounds I, 11, and I11 (Fig. 1-3) permit detailed structural analysis. Aromatic resonances in the T 2.5-2.7 region may be assigned to the phenyl group protons and the proton of the pyrazole ring. The singlet of unit-proton intensity a t T 3.08, 2.98, and 3.05 in the three spectra respectively may be assigned to the strongly deshielded methine hydrogen of the hydrazone function. A singlet of threeproton intensity appears at T 7.42, 7.40, and 7.44 in each spectrum, respectively, and may be assigned to the methyl protons of the N-acetyl group. The other singlet a t T 8.01, 8.10, and 7.89, respectively, may be assigned to the methyl protons of the 0-acetyl groups; for I1 and I11 this peak has intensity of three protons, while for I the peak is of six-proton intensity. The spectrum of I11 shows a two-proton singlet, T 4.90, attributable to the methylene group from C-5 of the pentose precursor. The inethyl group of 11, corresponding to C-6 of the B-deoxy-~-hexoseprecursor, appears as a three-proton doublet, T 8.45, being split by the adjacent proton, with a spin-spin coupling constant, J = 6.5 C.P.S. The latter proton, from C-5 of the 6-deoxy-~hexose, appears as the expected quartet, T 4.08, of unitproton intensity, spin-spin coupling constant, J = 6.5 C.P.S. The D-hexose derivative I showed the methylene group from C-6 of the hexose as a two-proton doublet, (3) H. El Khadem, Z. iA4 El-Shafei, and M , M Mohammed-Ail. .I Or@ Chem., 39, 1565 (1964).

NOTES

OCTOBER, 1964

3073

I

ACO

Fig. 2.-X

-C -H

.m.r. spectrum of 5-( ~-gZycero-1-acetoxyethyl)-3-formyl-1-phenylpyrazole N-acetylphenylhydrazone

5.70, split by the adjacent proton, J = 6.0 C.P.S. The latter proton, from C-5 of the hexose, gives the expected unit proton triplet a t r 3.95 with J = 6.0 C.P.S. The dianhydrophenylosazone acetate structure (I111) would be expected from reducing disaccharides possessing (1-5) or (1-.6) links, the (1-2) link would prevent osazone formation and the (1-3) and (1-4) links would stop the formation of a pyrazole ring. Our findings indicate that the phenylosazones of a-~-g~ucopyranosy~-(~+3)-~-arabinose,~ maltose, cellobiose, and lactose were indeed simply acetylated in the presence of refluxing acetic anhydride while the (1 + 6)linked disaccharides, isomaltose and gentiobiose, were converted readily to the expected dianhydrophenylosazone acetates, I V and V, respectively. Experimental6 Dianhydrophenylosazone Hexaacetate (IV) from Isomaltose Pheny1osazone.-The osazoneB (I .2 9.) wae refluxed with acetic

(XI).

anhydride (50 ml.) for a period of 1 hr., then poured into crushed ice (0.5 kg.). The aqueous solution was decanted and the oily residue was washed twice with cold water, then crystallized by dissolving in hot 50% ethanol (25 ml.) and filtering the cooling solution upon turbidity. 2-(3-Formyl-l-phenylpyraeol-5-y1)2S-2-acetoxyethyl2,3,4,6-tetra-0-acety~-a-~-glucopyranoside acetylphenylhydrazone crystallized in colorless needles: yield 1 g. (60%); m.p. 190’; [ ~ ] * O D +118’ (c 0.55, chloroform); XFzH282 mp (log e 4.46); Y:”, 1740 (0-Ac), 1690 (N-Ac), 1600 cm.-’ (C=N); X-ray powder diffraction pattern’: 14.25 w, 10.78 s, 7.89 m, 5.06 w, 4.77 w, 4.46 w, 4.25 w, 4.09 m, 3.91 w, 3.60 w, 3.45 m, and 3.03 w. (4) G . Zemplen. Chem. Ber., 60, 1555 (1927).

(5) Infrared spectra were measured with a Perkin-Elmer Infracord spectrophotometer: ultraviolet spectra were measured on a Bausch and Lomb Model 505 spectrophotometer: n.m.r. spectra ne:e obtained in deuteriochloroform soluticns on a Varian Associates apparatus, Model A60. Microanalytical determinations were made by W . N. Rond, The Ohio State University. (6) A . Thompson and M. L. Wolfrom, J . A m . Chem. Soe., 7 6 , 5173 (1954). (7) Interplanar spacing, A., Cu Kar radiation. Relative intensities est,. mated visually: 8 . strong: m , medium: w, weak; v, very.

NOTES

3074

VOL.

29

TMS

5 N\Na I

Hc-N.:a

=

HC=C’

.- ,’

I I

C-N,

o ~ c

HC-C’

NQ

I Ch-OAe

N AC I

CH, - 0 A C

I

HC=N-N-

I

I

,

& . _..

&, -_ _ _ _ - . . . _ I

I

I

I

ilnal. Calcd. for CzeH40N4013: C, 58.69; H , 5.47; N, 7.60; total Ac, 35.06; 0-Ac, 29.21. Found: C, 59.19; H, 5.93; N, 8.17; total Ac, 34.18; OAc, 28.34. Dianhydrophenylosazone Hexaacetate (111) from Gentiobiose Pheny1osazone.-The osazone8 (2.3 g.) was refluxed with acetic anhydride (50 ml.) for a period for 1hr., then poured onto crushed ice (0.5 kg.). The aqueous solution was decanted and the oily residue crystallized on trituration with ethanol. 2-(3-Formyl-1pheny1pyrazol-5-yl)-2S-2-acetoxyethyl 2,3,4,6-tetra-O-acetyl-ao-glucopyranoside acetylphenylhydrazone recrystallized from 50% aqueous ethanol in colorless needles: yield 2.2 g. (68%); m.p. 178”; [ a ] 2 0f31.6” ~ (c 0.475, chloroform); : : :A 282 mp (log 6 4.46);:::Y 1790 (0-Ac), 1685 (N-he), 1595 cm.-1 (C=N); X-ray powder diffraction pattern?: 11.79 m, 10.78 s, 10.04 m, 7.76 s, 5.86 w, 5.40 w, 5.18 w, 4.82 vs, 4.17 m, 4.02 8 , 3.86 m, 3.66 w, 3.48 m , and 3.23 m. Anal. Calcd. for C3SH40N4O13: C, 58.69; H, 5.47; N,7.60; total Ac, 35.06; 0-Ac, 29.21. Found: C, 58.63; H , 5.26; N, 8.01; total Ac, 34.38; 0 - A c , 29.64.

Acknowledgment.-The author is indebted to the Educational and Cultural Exchange Program for a Fulbright Grant to visit the Ohio State University; to Professor 31. L. Wolfrom, Dr. D. Horton, and Mr. F. Koniitsky for their counsel; to The Ohio State University for the laboratory facilities provided; to Dr. A. Sat0 who kindly made available the isomaltose and gentiobiose used; to Dr. R. D. Nelson of Chemical Abstracts who helped in the nomenclature of the compounds. (8) H. Berlin, J . A m . Chem. Soc., 48, 1107 (1928).

Amidino and Carbamoyl Osazones of Sugars M. L. WOLFROM,~ H. EL KHADEM, AND H. ALFES Department of Chemistry, T h e Ohio State University, Columbus, Ohio .&3910 Received M a y 99,196’4

1,2-Bis(amidinohydrazone) salts2v3 are known4 to be active against some forms of leukemia but have the (1) T o whom inquiries should be addressed.

2

.~ 1

~

I

I

I

I

disadvantage of producing toxic effects. It was considered that saccharide bis(amidinohydraz0ne) salts might be less toxic. These compounds can be considered as amidinoosazones but attempts to synthesize them according to the method used by Fischer5 to prepare sugar osazones proved unsuccessful. We have therefore prepared them from aldosuloses (“osones”) and 1-aminoguanidine sulfate. Thus, D-arabino-hexosulose bis(amidinohydras0ne) manosulfate (Ia) was obtained from D-arabino-hexosulose (“D-glucosone”), L-xylo-hexosulose bis(aniidinohydrazone) monosulfate (IC) from L-sylo-hexosulose (from L-sorbose), and Dthrso-pentosulose bis(aniidinohydrazone) monosulfate (Id) from D-threo-pentosulose (from D-xylose). These compounds exist as slightly yellow crystals, soluble in hot water, from which they crystallize on cooling. Although sugar disemicarbazones can be considered as carbamylosazones, they cannot be prepared by treating aldoses with semicarbazide as the reaction does not proceed beyond the semicarbazone stage; we have therefore prepared these from aldosuloses (“osones”) and semicarbazide also. Thus, D-arabino-hexosulose disemicarbazone (IIa) was obtained from ~-arabinohexosulose (“D-glucosone”) , D-lyxo-hexosulose disemicarbazone (IIb) from D-lyxo-hexosulose (“D-galactosone”), L-xylo-hexosulose disemicarbazone (IIc) from L-xylo-hexosulose (“L-gulosone”) , and D-threo-pentosulose disemicarbazone (IId) from D-threo-pentosulose (“D-xylosone”) . Unlike arylosazones these compounds were colorless and their ultraviolet spectra were characterized by a single maximum at 292 mp instead of consecutive maxima stretching from 256 to 399 mp in the case of the phenylosazones.6 Furthermore, they possessed greater solubility in water than in organic solvents, excluding pyridine. (2) J. Thiele and E. Dralle, Ann., 801, 275 (1898). (3) E. G. Podrebarac, W . H. Nyberg, F. A. French, and C. C. Cheng, J. Med. Chem., 6, 283 (1963); F. Baiocchi, ef ol.,i b i d . , 6, 431 (1983). (4) B. L. Freedlander and F. A . French, Cancer Rea., 18, 360, 1288 (1958). (5) E. Fischer, Ber., 17, 579 (1884). (6) V. C. Barry, J. E. MoCormick, a n d P . W . D. Mitchell. J . Chem. Soc., 222 (1955); G. Henseke and M.Winter, Ber., 88, 45 (1960).

NOTES

OCTOBER,1964 0

NH

HC=N-NH-~-NH~

HC=N--SH-~-NH~ LN-NH--C-NHz 1 I1

R

NH I

Experimental la

I/

/I

.H2S04

I

C=N-NH-C-NHz

II

I

R

0

I1

Like the sugar semicarbazones studied earlier,' the disemicarbazones formed green complexes with copper salts; these, however, were deeper in color and possessed greater stability, probably owing t o the contribution of the second semicarbazone residue in the formation of tetradentate ligands as in 111. Nitrous acid is known to react with seniicarbazoness and osazones,g liberating the free sugar or aldosulose ("osone"). Nitrous acid readily reacted with Darabino-hexosulose disemicarbazone t o yield D-arabinohexosulose ("D-glucosone1'), which was characterized by conversion to the quinoxaline derivative. Acetylation of saccharide seniicarbazones has been showna,lot o result in the formation of several isomeric acetates; the same difficulty was encountered with disemicarbazones where eight isomers could be detected on thin layer chromatograms. However, it was possible to isolate a chromatographically homogeneous compound from acetylated D-arabino-hexosulose disemicarbazone after treatment with nitrogen trioxide. This compound now possessed two nitrogen atoms and four acetyl groups and its infrared spectrum showed the 0-acetyl band a t 1755 and the C=N band a t 1590 cm.-'. Furthermore, it reacted with phenylhydrazine bisyielding 3,4,5,6-tetra-0-acety~-~-arabino-hexosu~ose (phenylhydrazone) l 1 and the latter was converted to Percival's anhydroosazone.12 It was therefore tentatively ascribed the structure I V of tetra-o-acetylD-arabino-hexosulose diimine. (7) H. El Khadem, M. F. Iskander, and S. E. Zayan, Z. Anoro. Allpem. Chem., 316,72(1963). (8) M . L. Wolfrom, L. W. Georges, and S. Soltzberg, J . Am. Chem. Soc., 68, 1794 (1934). (9) H. Ohle, G. Henseke, and A. Czyzewski. Ber., 88, 316 (1953). (10) M . L. Wolfrom and 9. Soltzberg, J . A m . Chem. Soc., 68, 1783 (1936). (11) K . Maurer and B. Schiedt, Ber., 68, 2187 (1935). (12) E . G. V. Percival, J . Chem. Soe., 1770 (1936): 1384 (1938): G . Henseke, 0. Muller, and G. Badicke, Ber., 91. 2270 (1958).

3075

D-arabino-HeXOSUlOSe Bis(amidinohydraz0ne) Monosulfate (Ia) .-To a suspension of aminoguanidine bicarbonate in water sulfuric acid (diluted 1 : 1) was added slowly until the pH was between 4 and 5. This solution was diluted with water so that l ml of the diluted solution contained 360 mg. of I-aminoguanidine sulfate. Dry ~-arabino-hexosulose~~ powder (5 g ) was added to a solution of 9 g. of 1-aminoguanidine sulfate (25 ml. of the above prepared solution, 30y0 excess) and 7.5 ml. of glacial acetic acid. The mixture was heated a t 100" for 10 min. and cooled to room temperature. Crystallization began after 20 min. On standing for 2 days a t room temperature and 1 day in the refrigerator, the product was filtered and washed with water, methanol, and ether; yield 3.5 g. The substance was recrystallized twice from boiling water to yield light yellow needles: m.p. 220" dec.; [ a I z 4+9" ~ (c 0.5, water, no mutarotation observed in 48 hr.);" : :A 289 mp (log c 4.56); ;::1 3.0 (OH, XH), 5.9, 6.9-(HN:CNHz), 6.3 (C=N), 8.8-9.4 p (S04-2); X-ray powder diffraction datal5: 10.50 w, 7.80 m, 6.81 w, 6.15 m, 5.75 s, 5.30 s (2), 4.97 s, 4.25 m, 4.05 w, 3.87 w, 3.75 w, 3.54 s (3), and 3.39 vs (1). Anal. Calcd. for CsHloNaOaS: C, 24.74; H, 5.19; N, 28.86; S, 8.26. Found: C, 24.61; H , 5.63; N , 28.89; S, 8.21. L-zylo-Hexosulose Bis( amidinohydrazone) Monosulfate (IC). -Attempts were made to prepare the title compound according to the procedure mentioned above. A few crystals formed after 4 months. These were filtered and recrystallized from hot water to be used as nuclei. The procedure was repeated. Dry L-zylo-hexosulose powder (5 g.) was added to a solution of 1amidinoguanidine sulfate (9 9.) and 7.5 ml. of glacial acetic acid. After standing 1 day a t room temperature the solution was seeded with the above nuclei. After 24 hr., crystallization began. The solution was allowed to stand a t room temperature for 5 days, then in the refrigerator for 1 day. The crystals were filtered and washed with water, methanol, and ether; yield 2 g. Two recrystallizations from boiling water gave slightly yellow crystals: m.p. 214" dec.; [ C X ] ~ ~-46" D (cO.41, water); 289 1 mp (log E 4.42); 3.0 (OH, NH), 5.9, 6.0 (HN:CNHz), 6.3 (C=N), 8.8-9.4 p (Sod-'); X-ray powder diffraction datals: 7.73 m, 6.68 w, 5.96 vw, 5.70 m, 5.44 w, 5.17 s, 4.84 8, 4.58 w, 4.27 8, 4.01 8 (3), 3.72 m, 3.62 w, 3.48 vs ( l ) , 3.37 vs (2), and 3.20 m. Anal. Calcd. for CaH20N808S: C, 24.74; H, 5.19; N , 28.86; S,8.26. Found: C,24.84; H , 5.17; N,28.60; S , 8 . 2 6 . D-threo-Pentosulose Bis( amidinohydrazone) Monosulfate (Id). -D-threo-Pentosulose ( 5 g.) was treated as above with 10.7 g. of I-aminoguanidine sulfate; yield, 3 g. of Id. Pure material was obtained as light yellow needles on two recrystallizations from boiling water: m.p. 211-212" dec.; [ a I z 1-17' ~ ( e 0.34, water, 289 mp (log e 4.57); no mutarotation observed in 4 days); 3.0 (OH, PU'H), 5.9, 6.0 (HN:kNHz), 6.3 (C=N), 8.8-9.4 (S04-9; X-ray powder diffraction datal5: 10.28 m, 7.53 m, 7.03 m, 5.42 m, 4.93 w, 4.63 m, 4.43 VI ( l ) , 3.88 m, 3.75 m,

::A;

p

3.56 s (3), 3.46 s (Z),3.28 w, 3.18 w, 3.05 w, and 2.99 s. AnaE. Calcd. for G H I ~ N ~ O ~C, S : 23.46; H, 5.06; N, 31.28; S, 8.95. Found: C, 23.19; H , 5.28; N,31.46; S, 8.87. D-arabino-Hexosulose Disemicarbazone.lB-The sirupy Darabino-hexos~lose~~ prepared from 10 g. of phenylosazone was taken up in 10 ml. of water, and semicarbazide hydrochloride (3 9.) and sodium acetate (3 9.) in the least amount of water were added. The mixture was allowed to stand a t 0" for 5 days and the crystalline product formed was filtered, waehed with water and ethanol, and recrystallized from 50y0 ethanol: yield 1.2 g.; m.p. 215" dec.; [ a I z 4$55" ~ ( c 0.34, water); 292 mp (log e 4.31); At: 3.0 (OH), 5.9, 6.0 (CONH), 6.3 p (C=N); (13) Infrared spectra were obtained on a Perkin-Elmer Infracord spectrometer and ultraviolet spectra on a Bausch and Lomb model 505 spectrophotometer. Microanalytical determinations were made by W. N. Rand. (14) "D-Glucosone." E. Fischer and E. F. Armstrong, Ber., 35, 3141 (1902); S. Bayne, T. A. Collie, and G. A . Fewster, J . Chem. Soc., 2766 (1952). (15) Interplanar spacing, A,, Cu K a radiation. Relative intensity, estimated visually: s, strong; m. medium: w, weak v, very: three strongest linesnumbered (1. strongest). (16) Experimental work by E. G. Wallace.

3076

NOTES

VOL. 29

X-ray powder diffraction datal5: 7.80 m, 6.89 m, 5,97 w, 5.57 w, of Health, Education, and Welfare, Bethesda, Maryland. 5.21 m, 4.92 vs ( l ) , 4.18 m, 3.86 m, 3.46 s (2), 3.30 s, 3.06 s (The Ohio State University Research Foundation (3), 2.78 w, and 2.69 w. 759F). H. El Khadem is indebted to the EduProject A n d . Calcd. for CsHleN~oe: c, 32.87; H , 5.52; N, 28.75. cational and Cultural Exchange Program €or a FullFound: C, 32.84; H , 5.65; N, 28.79. Absorption Spectrum of the Copper Complex.-A solution of bright Grant. E. G. WallacelGis indebted to the Allied D-arabino-hexosulose disemicarbazone (292 mg.) in water (10 Chemical and Dye Corporation of New York, h'ew ml.) was treated with a solution of cupric chloride (133 mg.) in York, for a fellowship. ethanol (10 ml.) and the volume was brought to 25 ml. with ethanol. The absorbance of the green solution ":A:[ 740 mp, (log e 1.6911 remained constant for 24 hr. D-1 yzo-Hexosulose Disemicarbazone .-D-lyzo-Hexosulose sirup Reduction Potential and Effect of ortho (1.0 g.) was treated as above and the product was isolated in the same manner: yield, 0.25 g. of disemicarbazone which was reSubstituents on Dimerization of Aromatic Nitroso crystallized from water; m.p. 235" dec.; [ C Y ] ~ ' D -2" (c 1.9, Compounds water); AEz 292 mp; ::A: 3.1 (OH), 5.9, 6.0 (CONHz), 6.3 p (C=N); X-ray powder diffraction data's: 4.78 m, 4.25 s (3), 3.99~(2),3.65m,3.24vs(1),3.15w,2.85~.,2.48w,and2.33w. RICHARD R. HOLMES Anal. Calcd. for C&N&a: C, 32.87; H, 5.52. Found: C, 32.35; H , 5.47. Department o j Chemistry, Hojstra University, Hempstead, L-zylo-Hexosulose Disemicarbazone .-L-zylo-Hexosulose sirup Keur York (10 9.) was treated as above: yield, 0.75 g. of disemicarbazone which was recrystallized from water; m.p. 225" dec.; [ c Y I Z O D - 16" Received February 6, 1964 (c, 0.45, water); A",t."," 292 mp (log B 4.69); A",: 3.0 (OH), 5.95, 6.05 (CONHz), 6.33 p (C=?J); X-ray powder diffraction data's: 7.38 m (31, 5.91 w, 5.32 w, 4.86 m, 4.23 w, 3.67 s (2), 3.45 vs ( l ) , I t has been known for many years that ortho groups 3.14 m, 3.02 vw, 2.50 w, and 2.29 w. larger than hydrogen favor dimerization of aromatic Anal. Calcd. for CsH1GN600: C, 32.87; H , 5.52; N, 28.75. nitroso compound^.'-^ For example, in a 0.1 M benFound: C, 32.86; H, 5.98; N, 28.66. D-threo-Pentosulose Disemicarbazone.-D-threo-Pentosulose zene or chloroform solution a t 20°, nitrosomesitylene sirup (1.0 g.) yielded 0.15 g. of disemicarbazone which was reis4t5 about 70% dimer, whereas p-bromonitrosobenD (c 1.5, crystallized from water: m.p. 232" dec.; [ c Y ] ~ ~4-3" zene, 4,5 p-methylnitro~obenzene,~and nitrosobenzene water);" : :A 292 mp; ::A: 2.9 (OH), 5.9, 6.0, (CONH2), 6.33 p itself4s5 are close to 100% monomer. Various explana(C=N); X-ray powder diffraction data's: 10.92 m, 7.34 m, 5.47 tions have been offered.3,6-8 Luttke' has presented a w, 5.07m,4.21s,4.05w,3.88s(2),3.55s(3),3.46s,3.30vs(l), 2.99 m, and 2.97 m. detailed theory in which he attributes increased diAnal. Calcd. for C7Hl4N6OS: C, 32.05; H, 5.38. Found: merization of ortho-substituted compounds to steric C, 31.62; H , 5.61. inhibition of resonance in the dimers. It is the purpose Action of Nitrous Acid on D-arabino-Hexosulose Disemicarbaof the present Note to show that di-ortho-substituted zone.-A solution of the disemicarbazone (0.5 g.) in water (25 ml.) was treated a t 50" with sodium nitrite (5.5g . ) in 25 ml. of nitroso monomers are subject to a sizeable steric effect water followed by the dropwise addition of 15% hydrochloric and to suggest that steric inhibition of resonance in the acid (12 ml.). After 30 min. the hexosulose solution was neumonomers is the cause of the greatly increased dimerizatralized and refluxed for 10 min. with o-phenylenediamine (0.2 tion of these compounds. 9.) and 1 ml. of acetic acid. The quinoxaline derivative separated Some years ago, Lutz and Lyttong measured the on cooling and was recrystallized from ethanol: m.p. 188", [ c Y ] ~ ~-76.6' D (c 1.84, 5 N hydrochloric acid) (1it.l' m.p. 187reduction potentials, EO, of a number of monosubsti188", [ a ] D -75.2'). tuted nitrosobenzenes. The effect of a substituent in Acetylation of D-arabino-Hexosulose Disemicarbazone and meta or para position varied with the electronic nature Reaction with Nitrogen Trioxide.-The disemicarbazone (3.5 g.) of the group, but every substituent in the ortho posit'ion was acetylated by stirring with a mixture of 15 ml. of acetic made reduction take place more readily. This obseranhydride and 30 ml. of pyridine until complete dissolution occurred (4 days). Since the acetylated disemicarbazone vation suggests a steric effect. It was of interest, was water soluble, the reaction mixture was evaporated to dryness therefore, to measure polarographically the reduction and the acetate, which could not be obtained in a crystalline potentials of some di-ortho-substituted nitrosobenzenes form, was subjected to thin layer chromatography [silica gel G, and, for comparison, of related compounds with one methanol-benzene (1 : 9)] t o reveal eight spots on spraying with sulfuric acid. The crude acetate was taken up in acetic acid and substituent in the para position only. The results are a stream of nitrogen trioxide was passed into the solution for 5 presented in Table I. hr. a t 5' after which the mixture was left overnight a t room The second column lists observed half-wave potentemperature. The reaction mixture was then evaporated, tials, the third the difference between the potential taken up in chloroform, washed with water and then with aqueous and that for nitrosobenzene itself. A more positive sodium hydrogen carbonate, and evaporated to dryness, yielding a nearly colorless, hygroscopic sirup which distilled a t 140" (0.5 potential means that the compound is more easily remm.): it appeared as a single spot ( RF 0.88) on a paper chromaduced. Although the absolute values of the reduction togram developed with 1-butanol-ethanol-water (4:1 : 5); [CY] 2 2 ~ potentials listed are not directly comparable to those f 2 8 " (c 3.8, chloroform); ::A: 5.7 (0-Ac), 6.29 p (C=N). reported by Lutz and L y t t ~ nsince , ~ the latter authors' Anal. Calcd. for (C1,HzJV,Ol)z.HzO: C, 47.59; H , 5.99; N , 7.93; CHSCO, 48.73. Found: C, 47.13; H , 5.63; N, values refer to acidic solutions and were obtained by 7.82; CH,CO, 48.67. the more accurate potentiometric titration method, The above substance yielded with phenylhydrazine and ace(1) E.Bamberger and A . Rising, Ber., 34, 3878 (1901). phenylosazonell tic acid 3,4,5,6-tetra-0-acety~-~-arabino-hexose (2) C. K.I n g o l d a n d H. A . Piggott, J . Chem. S O C . 125, , 168 (1924). which was converted with sodium hydroxide in acetone to Perci(3) D.L. Hammick. J . Chem. S O C .134, , 3105 (1931). Val's dianhydrophenylosazone,l2m.p. and m.m.p. 238".

Acknowledgment.-We are pleased to acknowledge the support of Grant CY3232 (C3) from the National Institutes of Health, Public Health Service, Department (17) H. Ohle, Ber., 67,155 (1934)

(4) V. Keussler a n d W. Luttke. Z. Elektrochem., 63, 614 (1959). (5) W. 3. Ylijs, S. E. Hoekstra, R. M . Ulmann, a n d E. Havinga, Rec. trau. chim., 7 7 , 746 (1958). (6) K. Nakamoto a n d R. E. Rundle, J . A m . Chem. S O C .7, 8 , 1113 (1956). (7) W.Luttke, Z.Elektrochem., 61, 982 (1957). (8) J. W. Smith, J . Chem. Soc., 1124 (1957). (9) R . E.Lutz and M R . Lytton, J . O w . Chem., I,71 (1937).

OCTOBER,1964

SOTES TABLE I AEl/t

El/1

(s.c.e.),

AEljZ,

steric,

V.a

V.

V.

AAF' steric, kcel./ mole

Nitrosobenzene Unsubstituted

-0

20

0 00

0

0

443-

-0

19

t o 0 1

0

0

4-CHa2.6-Di-C12,4,6-Tri-C1-

-0.24

-0

- 0.08' -0.07'

+O +O

04 12 13

+ O 10 + O 10

-4 -4

6' 6'

2,4,6-Tri-CHa-

-0.21
160-170°), the nickel-catalyzed hydrogenation of I to I1 results in the formation of I11 by the following reaction. 211 +I11

+ NHs

(5)

I n addition to the formation of secondary amine, an important side reaction a t very high temperatures is the production of cyclohexane. This reaction is H2

I1 * CeHiz

+ NHI

(6)

noticeable below 300' but not important below 325' with nickel and cobalt catalyst^.^ Cyclohexane formation becomes excessive with the more active ruthenium catalyst at about 200° and is not diminished by the addition of I11 or ammonia." Experimental Each experiment on the hydrogenation of I w m run in a 1-l., stainless steel, rocking autoclave with 186 g. (2.0 moles) of I and, unless otherwise specified, 1.86 g. of a dry 5% metal-oncarbon catalyst (Engelhard Industries, Inc.) at 800-1000 p.s.i.g. The effect of adding each of the following substances was based on I ) , concenstudied: I11 (18.0 g., 9.10 mole, 5 mole trated aqueous ammonia (14 ml., 0.2 mole, 10 mole "/b based on I ) , and ammonia (5 ml. as liquid ammonia, ca. 0.2 mole, 10 mole (5) C. F. Winans, I n d . Eng. Chem., 32, 1215 (1940). (6) C. F. Winans, U. S. P a t e n t 2,129,631 (Sept. 6. 1938). (7) A. I. Naumov, 2. G . Lapteva, and M . M. Shumilina. U.S.S.R. Patent 114.260 (July 30, 1958); Chem. Abstr., 63,14.026e (1958). (8) P. Richter, .J. Pasek, V . Ruzioka, and L. Jarkovsky. Czechoslovakian Patent 98,263 (appl. Nov. 17, 1959); Chem. Abstr., 66, 12,771i (1962). (9) J. yon Braun, G. Hlessing, a n d F, Zobel. Ber., 66B,1988 11923). (IO) C. F. Winans and H. Adkins, J . A m . Chem. Soc.. 6 4 , 306 (1932). (11) G . M . Illich, Jr., and R. XI. Robinson, U. S.Patent 2,822.392 (Feb. 4, 1958); U. S. Patent 2,955,926 (Oct. 11. 1960); British P a t e n t 836,951 (June 9, 1960).

OCTOBER,1964

KOTES

TABLEI HYDROGENATION OF A N I L I N E ~ Temp., Added

Catalyst

(

OC.

Time, hr.

Mole % yieldb I1 I11

71 5 15 5 120 11 120 11 5 69 5 15 5 145 2 73 15 5 140 3 72 16 5 3 145 2 25 75. 82 6 5 145 18 5 75 63 25 .. 120 65 25 120 12 Rh 145 1 5 61 28 h 145 4 5 5 a 145 Pd 145 Pt 5 j a Cnless otherwise specified, each experiment was run in a 1-1, rocking autoclave with 186 g. (2.0 moles) of I and-1.86 g. of a 5% metal-on-carbon catalyst a t 800-1000 p.s.i.g. until the reaction was complet,ed. Determined by gas-liquid chromatography. Reactions that went to completion resulted in about 18 g. 0.5 mole % yields of cyclohexane and traces of benzene. 14 ml. (0.2 mole). e 756 yield of cyclohexanol (0.10 mole). Liquid NH3, ca. also was obtained. Used 3.72 g. of catalyst. 5 ml. (ca. 0.2 mole). Little or no reaction a t 1250 p.s.i.g. Ca. one-third complete. j Little or no reaction at 1200 p.s.i.g.

111.

'

yo based on I ) . The liquid ammonia was added to the autoclave after the vessel, containing I and catalyst, had been cooled in a Dry Iceacetone bath. The autoclave then was sealed and purged with nit.rogen and then with hydrogen, and hydrogen was added while the vessel still was cold. At the end of each reaction the autoclave was cooled and depressurized, and its contents were filt'ered to remove the catalyst. Methanol was used as a solvent to assist in removing the reaction mixture from the autoclave and for washing t,he filter cake. The filtered reaction mixture, now containing methanol, was separated by distillation into several fractions that then were analyzed by gas-liquid chromatography. The results are summarized in Table I . Except for the experiments with added liquid ammonia, duplicate experiments gave excellent reproducibility both of react,ion rates and product yields. Duplication of the experiment with the ruthenium catalyst and added liquid ammonia gave very good reproducibility of product, yields; however, the reaction rates were variable due, no doubt, to variable losses of the volatile ammonia during loading and purging operations. In one experiment, 198 g. (2.0 moles) of I1 was treated with 1.86 g. of 5% ruthenium on carbon and hydrogen for 5.75 hr.'at 145' and 1075 p.s.i.g. After removal of the catalyst by filtration, the reaction mixture was analyzed by gas-liquid chromatography. About 3'3, of I1 had been converted into 111. No other reaction was detected. The material balancesfor the experiments described in Table I are significantly less than quantitative. It is likely that this is due to mechanical losses, since no other products were detected.

Results and Discussion The palladium catalyst was severely inhibited and the platinum catalyst almost completely poisoned, presumably by the strongly adsorbed I1 initially formed. No further work was done with these catalysts. Similar inhibitions of platinum catalysts have been reported.12 The use of an acid iiiediuni with platinum catalysts to convert toxic free bases to nontoxic animoniuni ions is well known. Such techniques mere not used in this study. Ammonia suppresses the formation of secondary amines during the hydrogenation of nitriles,13 and it (12) J . M Devereux. K. R. Payne, and E. R. A . Peeling, J . Chem. S o c . , 2845 (1957). (13) E. J. Schwoegler and H. Adkins, J . Am. Chem. SOC.,61, 3499 (1939).

3083

has been postulated that this effect is due to the reverse of eq. 4.13 The formation of an N-substituted imine and ammonia froni an iniine and a primary amine (eq. 2 and 4) proceeds at room temperature, is not catalyzed, and should be suppressed by amnionia.l0 Indeed, it has been demonstrated that the reaction sequence in eq. 2 and 4 is in fact re~ersib1e.l~ The use of dry ammonia resulted in considerable inhibition with ruthenium and coiiiplete inhibition with rhodium. The experinlent with ruthenium was carried to completion; although the reaction was impractically slow, the ammonia significantly decreased the formation of 111. Water is known to eliminate the catalyst toxicity of ammonia by converting it to the nontoxic ammonium cation.1 5 , 1 6 Predictably, the use of aqueous ammonia with the ruthenium catalyst did not cause any inhibition and still decreased the formation of 111. However, there resulted the expected formation of cyclohexanone (and its subsequent reduction to cyclohexanol) by hydrolysis of enamine intermediate~'~,'* (eq. 7). Thus, the use of either dry or

aqueous ammonia to decrease the formation of I11 is not feasible. The addition of 111retards the formation of additional secondary amine in the n i ~ k e l - , ~~olsa!t-,~q~ .~ and rutheniuni-cat'alyzedll hydrogenations of I a t high temperatures (>ZOO'). This effect of added I11 probably is due to the demonstrated rever~ibility'~ of the reaction shown in eq. 5 . KO such effect was observed in our experiments atj lomer temperatures with ruthenium and rhodium catalysts ; results of the t,reatment of I1 with a ruthenium catalyst at 145' in the presence of hydrogen indicated that the reaction shown in eq. 5 is of minor importance under our experimental conditions. Furthermore, t'he added I11 retarded the rate of the rhodium-catalyzed hydrogenation, although the rate with the ruthenium catalyst was little affected. An inhibitory effect of alkyl amines on a, rhodium catalyst previously has been reported.20 There was very little formation of cyclohexane (I. Hart, The Pennsylvania. Sta.tc Cniversity, 1963. ( 2 ) National Science Foundation Cooperative Fellow, 1962-1963. (3) 31, Talnres and S. Searles, J r . , J . Am. Chem. Sac., 81,2100 (1959). ( I ) H . J . Cainpbell and J. T. Edward, Can. J . Chem., 38, 2109 (1960). ( 5 ) S. Spades. 11.Tamres, and G. hI. Barrow, J. Am. Chem. Sac., T6, 71 ( 19.53). (6) R . Huis.gen. H. Brade, H. Wala, and I. Glogger, Chem. B e r . , 90, 1 4 3 7 il9.57). ( 7 ) . I . T . E d w a r d n n d k1. Stollar, Can. J . Chem., 41, 721 (1963). 1 8 ) T. Graitisml, Spectrorhzm. Acta, 19, 497 (1963). ( 9 ) .\I. I ) . .roesten n n d R. 5. Drago, J . .4m.Chem. S a c . , 84, 2037 (1962). f l o i 11. I ) . Joesten and R . S.Drazo, i b i d . . 84, 2696 (1962). 11) K . 11,I h i i g e r and S.H. I h i i e r , J . Chem. P h y s . . 5 , 839 (1937). ( 1 2 ) I). 1 ., Glrisker. H , 17. Thotnpson, and R . S. hIulliken, ibid., 21, (13) r. 1). S~liiiiiilhi\rltand R . S. Drago. J . .4m. Chem. Sac.. 84, 4484 (ISfiO). r l i ) R 3 , I>ra.zo a n d 11. A . K e n z . ihzd.. 84, 526 (1962).

V OL. 29

data taken at 1418 and a t 285 mp provide strong evidence for the existence of l : l complexes. The enthalpy and entropy changes accompanying complexation were determined by a least-squares treatment of data plotted for eq. 1.

A summary of thermodynamic constants is given in Table 11. All of the S-methyl lactams possess larger equilibrium constants for 1:1 complex formation than do the X , N - d i n i e t h y l a n i i d e ~ ~(see ~ ~ ~Table 11). If in fact the decrease in the equilibrium constant that accompanies substitution of the propionaniide for the acetamide is due to an unfavorable entropy effect arising from the methyl group on the a-carbon, suggested by Joesten and Drago, then the magnitude of the equilibrium constants involving amides with bulkier alkyl groups on the a-carbon would not be expected to exceed 107 ]./mole, the value reported for N,N-diniethylpropionaniide. From the liniited equilibrium constant data it appears that an increase in basicity toward phenol results from cyclization of amides. Based on the magnitude of the equilibrium constants, the order of basicity for the N-methyl lactams is six> seven- = five-membered ring. The greater basicity of the six-membered ring indicated by previous pKa measurenients6 has been confirmed. An explanation of this order is not readily apparent from our data. Certain variations in therniodynamic constants are significant however. Table I1 shows that the enthalpy and entropy changes for complex formation with N,N-dimethylaniides and the seven-membered ring lactani are comparable. Assuming that solvent effects are relatively small or that these effects tend to cancel for a series of related compounds it appears that the seven-membered ring has sufficient flexibility to behave like the open-chain amides in adjusting to the orientational and spacial restrictions iniposed by complex formation. The increased rigidity of the sixand five-membered rings restricts the capacity of these lactams to make similar adjustments. As a result the enthalpy and entropy changes are lowered. A comparison of the thermodynamic data of the Nmethyl-E-caprolactam-phenol system with that of the E-caprolactam-phenol system shows that the values of the equilibrium constants are the same within experimental error. The magnitude of the enthalpy term is increased, however, upon substitution of a methyl group on nitrogen. Such a trend would be predicted on the basis of an inductive effect. The decrease in entropy accompanying adduct forniation is considerably greater for the N-methyl lactam. Qualitatively the larger entropy term for the N-methyl derivative arises from a more restricted configuration of the atoms as the complex becomes more stable ( A H becomes a larger negative value). Apparently no adverse steric effect is caused by the introduction of the methyl group on the nitrogen. Experimental Reagents .-The carbon tetrachloride used in these studies was either Fisher Spectranalyzed certified reagent grade or Fisher certified reagent grade. It was used without further purification.

NOTES

OCTOBER,1964

3123

TABLE I EQUILIBRIUM CONSTAXTS FOR 1: 1 PHENOL-LACTAM ADDUCTSI N CARBON TETRACHLORIDE" t, oc.

K,b ]./mole

ac

- aab

K,' l./mole

N-Methyl- y-butyrolactam 16.0 25.0 35.0 45.0

197 f 7 150 f 3 106 f 7 7 7 . 3 f2 . 9

1520 f 10 1450 f 40 1420 f 30 1390 f 20

149 f 9 ...

1660 f20 1580 f 10 1450 f 20 1430 f 60

166 f 17 ...

...

N-Methyl-Svalerolactam 25.0 35.0 45.0

260 f 12 185 f 3 145 f 4 104 f 5

16.0 25.0 35.0 45.0

216 f 9 149 f 3 104 rt 2 78.9 f 3 . 6

15.5 25.0 35.0 45.0

205 f 10 154 f 5 114 f 9 81.8 f 5 . 0

16.0

N-Methyl-e-caprolactam 1560 f 40 1590 f 20 1550 f 20 1510 f 50

146 f 9 ..,

e-Caprolactam 1500 f 30 1430 f 20 1380 f 50 1380 f 30

152 f 11

e-Caprolactamd 12.5 f 0 . 7 1 1001 3z 14e 679 f 14f 12.3 f 0 . 4 e 15.0 660 f 20f 9 . 6 f 0.6' 979 f 32e 1 0 . 1 f0 . 4 e 25.0 8.2 f 0 . 3 f 996 f 12e 652 f 14f 8.1 f 0 . 4 e 34.9 6.5 *0.2f 981 f 20e 625 f 8 f 6 . 5 f0 . 3 e 45.0 a All errors are given in terms of standard deviations for a series of no less than four samples. * Calculated from data obtained a t 285 mfi, c Calculated from data obtained a t 1418 mp. d Data for iodine-e-caprolactam 1: 1 adduct. e Calculated from data obtained Calcubted from data obtained a t 515 mp. a t 450 mp.

TABLE I1 THERMODYNAMIC DATAFOR 1: 1 ADDUCT FORMATION BETWEENPHENOL AND LACTAMS OR AMIDES IN CARBON TETRACHLORIDE" K (25O),

- AHt kcal./mole

- A S 7 e.u. N-Methyl- 7-butyrolactam 150 f 3 5 . 9 5 f 0.15 10.04f0.49 N-Methyl-6-valerolactam 186 f 3 5.62 f 0 . 1 2 8.42 f 0 . 4 0 N-Methyl-e-caprolactam 149 f 3 6.41 f 0 . 1 1 11.52 f 0 . 3 8 +Caprolactam 154 f 5 5.67 f 0.12 9.05 f 0.49 8.50 f 0 . 2 7 s-Caprolactamc 10.1 f 0 . 4 3.98 1 0 . 0 8 6 . 4 f0 . 3 11.7 f 0 . 6 K,N-Dimethylacetamided 134 & 3 N,N-Dimethylpropionamided 107 f 2 6 . 4 f0 . 2 12.1f 0 . 6 N,N-Dimethylformamided 64 f 1 6.1 f 0 . 4 12.1 f 0 . 7 From data collected a t 285 mfi. Errors for our data are given in standard deviations. * Value of In K obtained line of In K us. l / T plot. Data for iodine-e-caprolactam 1: 1 adduct provided for comparison only. d See ref. 10. Lactam or .amide

l./mole

Baker and Adamson reagent grade phenol was distilled under atmospheric pressure to give white crystals, m.p. 40-41'. The reported melting point is 40.9".'6 The compound N-methyl-y-butyrolactam (General Aniline and Film Corp.) was twice distilled under nitrogen atmosphere. The constant boiling middle fraction, n25~1.4678, was collected. Reported values of n Z 5are ~ 1.466618and 1.469." The n-methyl-e-caprolactam was obtained by one of the procedures described by Benson and Cairns.Ig By this method ecaprolactam (Eastman Kodak) was first converted to O-methylcaprolactam with dimethyl sulfate. The 0-methyl derivative was converted to N-methyl-c-caprolactam by retreatment with dimethyl sulfate. The N-methyl-e-caprolactam obtained in this manner gave a value for n z 6of~ 1.4819. The reported values for ~ 2 2 range 5 ~ from 1.4812 to 1.4833.'@ The ecaprolactam was twice recrystallized from anhydrous ether. After drying for 3 hr. on a sintered-glass funnel under a (15) R. R . Drsisbach and R. A. Martin, Ind. Eng. Chem., 41, 2875 (1949). (16) "Dictionary of Organic Compounds," Vol. 11, I. M . Heilbron, Ed. Oxford University Press, London, 1946, p. 806. (17) E . Fisher, J . Chem. Sac., 1382 (1955). (18) "Dictionary of Organic Compounds," Vol. 11, I. M . Heilbron, E d . Oxford University Press. London, 1946, p. 796. (19) R . E . Benson and T. L. Cairns, J . A m . Chem. Soc., 70, 2116 (1948).

- AFb(250), kcal./mole

2.95 f 0 . 1 5 3.11 f 0 . 1 2 2.98 f 0.11 2.97 f 0 . 1 2 1.36 f O . 0 8

from least-squares

steam of dry nitrogen, it gave a melting point of 68-69", lit.20 68-70 Baker and Adamson resublimed iodine was ground together with 25% by weight of potassium iodide and 10% by weight of calcium oxide. The mixture was then sublimed. The iodine was collected and stored in a desiccator over phosphorus pentoxide. Apparatus.-All spectrophotometric measurements were mad e on a Cary Model 14 recording spectrophotometer. The temperature of the cell compartment was controlled to within f 0 . 2 " with water from a constant-temperature bath circulating through the thermostated cell compartment and cell jacket supplied with the instrument. The temperature of the cell compartment was measured directly with a thermometer graduated in units of 0.1". Procedure for Ultraviolet Spectral Measurements.-Stock solutions of phenol were prepared by direct weighing of the phenol into a volumetric flask on an analytical balance and adding carbon tetrachloride to the mark. The stock solutions were used immediately. Solutions of known lactam concentration were prepared by direct weighing of the lactam into clean, dry volumetric flasks that had been flushed with nitrogen. Aliquots of

'.

(20) C. E . Dalgliesh, A. W. Johnson, and L. J . Haynes, "Chemistry of Carbon Compounds," Vol. I B , E. H. Rodd, Ed., Elsevier Publishing C o . , New York, N. Y., 1952, p. 840.

3124

NOTES

the phenol stock solution, measured to f-0.02 ml., were added from a 10-ml. buret t o the flasks containing the lactam, and carbon tetrachloride was added nearly to the mark. Carbon tetrachloride was added to the mark after the sample had come to thermal equilibrium a t the temperature spectral measurements were to be run. This procedure was used to eliminate concentration corrections arising from temperature-dependent volume changes of the solution. The spectra were measured with carbon tetrachloride in the reference cell. Cells of 1-cm. path length fitted with ground-glass stoppers were used. The absorbance of the solution containing only the Lewis acid was also obtained. The phenol-lactam complexes absorb in the same region as free phenol, but the former have larger extinction coefficients than free phenol. Spectra of phenol-lactam solutions were recorded at a constant wave length of 285 mp. This wave length was chosen, after studying numerous spectra of lactam-phenol mixtures between the wave length of 300 and 270 mp, because it was the wave length at which the difference between the extinction coefficient of the phenol and the complex was the greatest. Similar findings have been reported in connection with amide complexes of phen~l.~,lOExperiments on solutions which contained only lactam indicated that absorption due to free lactam was negligible at 285 mp; thus, the following (eq. 2) applies.21 K-1 = ab cc A - aaCa(a, - a,) - C, -

c b

- aaCa + A___ a, - a.

(2)

Here, K is the equilibrium constant for 1: 1 complex formation, c, is the initial concentration of the Lewis acid, c b is the initial concentration of the Lewis base, A is the absorbance a t a specified wave length, aa is the extinction coefficient of the Lewis acid, and a, is the extinction coefficient of the 1: 1 adduct a t the specified wave length. I n the derivation of eq. 2, Rose and Drago have assumed that the sole product species is a 1:1 complex, and that the Beer-Lambert law is obeyed by all absorbing species. The graphical method described by Rose and Drago was replaced by a more rapid and precise analytical solution adaptable t o computer programming. The simultaneous solution of eq. 2 for two sets of experimental data, denoted by subscripts 1 and 2, results in the quadratic expression that follows (eq. 3),

where D = a, - a,. The two values of D obtained from the solution of the quadratic equation are substituted into eq. 2 to solve for K-l. The true values for K-1 and D for the system are obtained by inspection. Ideally, one true value of K-1 and corresponding value of D should be obtained. Practically, it is necessary to take the average of the K-1 values obtained from each pair of experimental values. Data for the e-caprolactamiodine adduct were obtained a t 515 and 450 mp, the wave lengths for the maxima of the free and the complexed iodine bands, respectively. Values of K were obtained by the use of eq. 2. Procedure for Near-Infrared Spectral Measurements .-As a check on the ultraviolet measurements, equilibrium constants for the phenol-lactam systems were also computed from spectral data on the first overtone band of the free 0-H stretching frequency a t 1418 mp.2* Since this band is due to free phenol, it is possible to measure the equilibrium phenol concentration and, by assuming 1 : 1 complex formation, compute the concentrations of free lactam and the 1 : 1 complex as well. Two restrictions on the measurement of equilibrium constants by this procedure reduced the precision of the equilibrium constants to 7-10% of the mean value (in terms of standard deviation). The small extinction coefficient of 3.45 required as large a concentration of phenol as possible, as well as a path length of 10 cm. The upper limit of phenol concentration was set a t 0.006 M . Above this concentration, polymeric phenol species are reported to be present in appreviable amounts.22 The combination of these factors required spectral measurements at low absorbance, generally no greater than 0.150.

____ N . J. Rose and R. S. Drago. J . A m . Chem. S o c . . 81, 6138 (1959). ( 2 2 ) G . Alisnes and T . Gramstad. Acta Chem. Scand., 14, 1485 (1960)

(21)

VOL.29 The Replacement of Phenolic Hydroxyl Groups by Hydrogen W. H. PIRKLE'A N D JOHK L. ZABRISKIE~ Department of Chemistry, C'niversity of Rochester, Rochester, N e w York Received J u n e 15, 1964

We recently required a method for the replacement of phenolic hydroxyl groups by hydrogen. There existed at the time no good general method for this simple transformation, and none of the then reported3 procedures regularly gives high yields, although an excellent method has since been reported by Sawad for renioving the phenolic hydroxyl of various S-niethoxy4-hydroxymorphinan derivatives. We have indrpendently developed a very siniilar method5 and our modification, which gives excellent yields in selected cases, notably in the morphine series6 as does Sawa's, appears to offer soiiie practical advantages over that of Sawa. This report deals with an atteinpt to define the scope and applicability of the method. We may say at the outset that, although in many cases excellent yields are obtained, the reaction is not so general as we had hoped. It appears to be most successful when inethoxyl or phenyl groups are adjacent to the hydroxyl. Our modification of the method consists of the preparation of the 2,4-dinitrophenyl ether of the phenol, its catalytic hydrogenation to the corresponding 2,4-dianiinophenyl ether, and cleavage of this ether with sodium in liquid ammonia. The w e of the dinitrophenyl ether offers several advantages over that of the simple phenyl ether used by S a ~ a .Perhaps ~ the most important is ease of preparation. Sawa's phenyl ethers were prepared by the Ulliiiann reaction, considerably more cumbersome than the method of Reinheinier' or arylation with 2,4-dinitrofluorobenzene and sodiuni hydride in benzene-dimethylforniainide, both used in this work. The latter in particular proceeds smoothly to give excellent yields, even with highly

r

N\H~

1

(1) National Science Foundation Predoctoral Fellow, 1961-1963. (2) Union Carbide Fellow, 1963-1964. (3) G. W. Kenner and M . A. M u r r a y , J . Chem. Soc., S 178 (1949); G. W. Kenner a n d N. R. Williams, i b i d . , 522 (1955); S. W. Pelletier a n d D. M. Locke, J . Org. Chem., 29, 131 (1958). T h e classical zinc d u s t distillation is of course n o t generally useful except for the exhaustive degradation of hydroxylated substances t o hydrocarbons. (4) Y . K. S a w a , N. Tsuji, and S. Maeda, Tetrahedron. 1 6 , 144, 154 (1961). We are greatly indebted t o Dr. S a a a for making details of his procedure available t o us considerably in advance of publication. (5) Both methods are based on the cleavage of diphenyl ethers b y sodium in liquid ammonia. a reaction which has been studied a t length b y F. J . Sowa a n d his co-workers [ J . A m , Chem. Soc.. 69, 603, 1488 ( 1 9 3 7 ) : 6 0 , 94 (1938) I. (6) M. G a t e s a n d R. H. Pirkle, to be published. See also A I . Gates and T. A. Montzka. J . Med. Chem., 7 , 127 (1964). (7) J. D. Reinheimer, J. P. Douglass, H. Leister, and &I. B. Voelkel, J . Org. Chem., 22, 1743 (1957).

OCTOBER,1964

NOTES

3125 TABLE I1

TABLE I 2,4-DIiVITROPHENYL ETHERSO F Phenol

Dinitrophenyl ether, m.p., O C .

PHENOLSn

Yield, %

HYDROGENATION A N D CLEAVAGE O F 2,4-DINITROPHENYL ETHERS OF PHENOLS" Phenol

Cleavage Product

Guiacol Resorcinol monomethyl ether Hydroquinone monomethyl ether 2,6-Dimethoxyphenol

hindered cryptophenols.6j8 Table I lists the ethers prepared. Another reason for preferring the 2,4-dinitrophenyl ether is that the sense of cleavage is clearly defined, inasmuch as catalytic hydrogenation to the corresponding dianiino ether provides the structural features we deenied essential to specific c l e a ~ a g e . ~ Sowa's early work5 mggested that the sense of cleavage was dependent on the electron density a t the carbon atoms linked to the ether oxygen. It now appears that this is an oversimplified view, since the influence on cleavage of an ortho niethoxyl group is very different from that of a pura methoxy group,lo whereas their effect on the electron density of the neutml molecule should be much the same. Other advantages are greater crystallinity and ease of purification of the dinitrophenyl ethers and the higher solubility in the cleavage solvent of the corresponding dianiino conipounds. The reduction by this method of a number of phenols is suniniarized in Table 11. Those phenols with niethoxyl or aryl groups in the ortho position appear to give acceptable to good yields of cleavage products and

this generalization is in accord with our experience with the method in the 3-methoxy-4-hydroxyniorphinan series.8

(8) h notable exception is 2,6-di-t-butylphenol which gives in 78% yield only C-arylated product, m.p. 168.5-169.5', probably 2',4'-dinitro-3,5-diL-butyl-4-hydroxybiphenyl. ( A n a l . F o u n d : C, 64.50; H, 6.53.) (9) T h e specifity of cleavage observed by Sawa4 with 4-phenyl ethers in the morphinan series would nut have been predicted on t h e basis of Sowa's . . results.5 (10) A . J . Birch [ J . Chem. Soc., 102 (1947)l has found t h a t among the dimethoxy benzenes, cleavage to monomethoxy phenols occurs a s follows under comparable conditions: orlho, 89%: meta, 7 1 % ; para, 2.5%. These results are in sharp contrast t o those obtained during t h e cleavage diaryl 99% hydroquinone monoethers, e . g . , 2,4,-dimethoxydiphenyl ether methyl ether and 1% guiacol [ P . A. Sartoretto and F. J. Sowa, J. A m . Chem. S o r . , 69, 603 (1937)1, and it appears to us t h a t Zimmerman's v i e a s [Tetrahedron, 16, 169 (196111 on t h e effect of substituents on t h e cleavage of ethers may require modification a t least insofar as they apply to diaryl ethers. (11) These are not isolated as such hut used without purification owing to their facile autoxidation.

-

Anisole Anisole

Yield,* %

Guiacol 92.5-94 95 Resorcinol monomethyl ether 87-88,s 95 Hydroquinone monomethyl ether 112 94 2,6-Diniethoxyphenolb 167-168 95 64.5-66.5 84 Thymol Carvacrol Oil 93 2-Phenylphenol 114.5-115.5 85 1-Kaphthol 127-1 28 80 2-Saphthol 94-95 94 5,6,7,8-Tetrahydro2-naphthop 116-117.2 94 p-Benz ylphenol 74 5-76 96 p-Tritylphenold 242-245 70 a With the exception of 2,6-dimethoxyphenol, 5,6,7,8-tetrahydro-2-naphtho1, and p-tritylphenol, the 2,4-dinitrophenyl ethers of all of these phenols have been prepared by Reinheimer and his eo-workers' and some also by earlier workers. Our melting points agree with those reported by Reinheimer. * Calcd. for C,1H12NZ07:C, 52.50; H, 3.78. Found: C, 52.80; H, 4.00. c Calcd. for CI6H14X2O5: C, 61.14; H, 4.49. Found: C, 61.07; H, 4.42. Calcd. for CslH2;1N20a: C, 74.09; H 4.41. Found: C, 74.27; H , 4.57.

60 31 ...

0

1,3-Dimethoxybenzene

100 ... 0 Thymol Carvacrol ... 0 2-Phenylphenol Biphenyl 85 1-Naphthol Naphthalene 50" 2-Naphthol Naphthalened 18 5,6,7,8-Tetrahydro-2naphthol Tetralin 1 p-Benzylphenol Diphenylmethane 20 p-Tritylphenol Tetrap henylmethanee 4' a Cleavage products were analyzed by gas-liquid chromatography using appropriate internal standards, or by ultraviolet spectroscopy. Product identifications were made by comparison of retention times and infrared spectra or by melting points and mixture melting points. * Based on ether hydrogenated. Lithium used for cleavage. Longer cleavage times or more drastic conditions result in isolation of tetralin. e ;\Io& of the product (83%) appeared to consist of the product of Birch reduction, neutral and chromatographically homogeneous which consumed bromine without liberating hydrogen bromide. f Isolated yield.

Experimental Infrared spectra were taken on a Perkin-Elmer Model 421 infrared spectrophotometer, ultraviolet specti-a on a Cary Model 11MS recording spectrometer. Gas-liquid chromatography was carried out on a 200-em. column packedwith 307, Apiezon L on Chromosorb. Preparation of Ethers.-The following description is typical. A solution of 5.80 g. (34 mmoles) of 2-phenylphenol in 50 ml. of dry benzene were stirred under nitrogen with dry sodium hydride (1.50 g., 62.6 mmoles) until evolution of hydrogen ceased. 2,4Dinitrofluorobenzene (50 mmoles) and 25 ml. of benzene were then added, followed by 20 ml. of dimethylformamide, the latter added over 10 min. The reaction mixture was stirred for 0.5 h r . , refluxed for 0.5 hr., cooled, and extracted with dilute alkali and then with brine. Removal of solvent and crystallization of the residue from alcohol gave 9.70 g. (85%) of ether, m.p. 114.,5 115.5'. Hydrogenation and Cleavage of Ethers.-The following description is typical. A suspension of 3.47 g. (10 mnioles) of the 2,4-dinitrophenol ether of 2-phenylphenol in 500 ml. of methanol was hydrogenated a t atmospheric pressure over 100 mg. of Adam's catalyst. Uptake ceased when B molar equiv. of hydrogen had been consumed. The colorless solution was filtered and concentrated to dryness under diminished pressure. The residue was dissolved in 500 ml. of 9 : l liquid ammoniaether and treated with small pieces of clean sodium until the solution acquired a permanent blue color. After a brief standing, enough absolute ethanol was added to destroy the blue color of the solution. The ammonia was evaporated under an air stream and the dark residue was t,aken up in 250 ml. of water and extracted with three 50-ml. portions of ether. The combined ether extracts were dried with brine .and anhydrous sodium sulfate and divided into two equal portions. To one portion was added 770 mg. of naphthalene to serve as an internal standard during g.1.p.c. This analysis showed the yield of biphenyl to be 8574. Processing the remaining portion of ethereal solution gave crystalline biphenyl identified by comparison with an authentic sample.

NOTES

3126

Acknowledgment.-It is a pleasure to acknowledge the generous support of the National Science Foundation under Grant NSFG10474,Marshall Gates, principal investigator.

VOL. 29 SCHEMEI

HsC-N I

A Search for New Drugs in the Group of Xanthine Derivatives. XXII. Chemical Properties of 1,3-Dimethyl-6H,7H-ozazolo[2,3-f]xanthine System' The Chair of Pharmaceutical Chemistry and the Laboratory of Chemical Technology of Drugs, Medical Academy, Cracow. Poland

CH3 I X , Y = B r ; Z = H XVII,Y=Cl; Z=C1 X , Y = B r ; Z=C1 XVIII,Y=Cl; Z = J X I , Y = B r ; Z = B r XIX,Y=Cl; Z = B r XVI,Y=Cl; Z-H

X,xr,XVII-XIX

MARIANECKSTEIN, MARIAGORCZYCA, A N D ALFREDZEJC

Received February 26, 1964

HsC-N

By the action of alkali, 7-~-hydroxy-y-chloropropyl derivatives of 8-chloro- and 8-bromotheophylline cyI clized to 1,3-diniethyl-7-chloromethyl-6H,7H-oxazoloCH3 V,X=Br [2,3-f]xanthine. We have found that the 7-P-hydroxyVI,X=I y-bromo- and -y-iodo-8-halotheophyllines (I, 111 and VIII, x = c1 11, IV) undergo cyclization with alkali to give the 8H~C-~~~-CHz-CHz-CH3 bromoniethyl- and 8-iodomethyl-6H,7H-oxazolo [2,3-f]xanthines (V and VI) in high yield; no oxazirane prodO N N O H NaOH I ucts were obtained. The reaction of these oxazolidine CH3 derivatives with acids was then examined. xx Compounds V-VI11 dissolved readily in concentrated Q hydrobromic acid a t room temperature and, after brief H3C- ~ ~ ~ - C H ~ - ~ H - C H Z - X boiling, acidic products separated from the hot soluO Y N X I OH tion. The acid obtained from VI1 was transformed CH3 back to VI1 on heating with base and is assigned the I,X1=C1; X = B r I I I , X l = X = B r uric acid structure IX. Similarly VI11 gave the 7-(@II,X1=C1;X=I IV,XI=Br; X = I bromo-7-chloropropy1)-1 ,&dimethyluric acid (X),which was not identical with the isomeric 7-(P-hydroxy-ychloropropyl)-8-bromotheophylline obtained from 8bromic acid, giving the chloro-substituted uric acids bromotheophylline and epichlorohydrin.2 The same XVI, XVII, and XIX (Scheme I). The acid XVI was product X I was obtained from both the bromo comconverted back to VI1 with alkali, but much less readily pound V and iodo compound VI (Scheme I ) ; in the than bromide I X . Halogen exchange did not occur latter case iodine was liberated. with VI in this case, and the P-chloro-y-iodo derivative Like compounds IX and X, compound XI is soluble in was obtained. Attempts to reduce the halogen in posicold alkali and is precipitated unchanged on acidification 7 of compounds V and VI by means of zinc and tion. Recyclization by heating witfhsodium hydroxide acetic or hydrochloric acid to obtain V I I , gave only solution gave 1,3-diniethyl-7-bromomethyl-6H,7H-oxa- product XX, containing no halogen. The crystalline zolo [B,X-j]xanthine (V). The dibromopropyluric acid form, melting point, and analysis of compound X X (see structure X I was confirmed by an alternative syntheTable I) conformed to those of VII, but the mixture sis shown in Scheme 11. melting point of the two substances was depressed. By analogy to the work of Fischer3on 8-chlorocaffeine, Moreover compound X X , in contrast to VII, was the 7-allyl derivative Xl14gave XI11 which was readily readily soluble in alkali and separated unchanged after hydrolyzed to 1,3-dimethyl-7-allyluricacid (XIV). acidification, indicating that it is a uric acid derivative. Addition of bromine gave XI. In contrast to the uric The structure of compound X X as 1,3-dimethy1-7-nacid derivatives IX and X, compound XI is readily propyluric acid was confirmed by conversion with phosmethylated by dimethyl sulfate in the presence of phorus oxychloride and dimethylaniline to 7-n-propylsodium hydroxide, giving 1,3,9-trimethyl-7-(P,y-dibro- 8-chlorotheophylline (XXI) , identical with a sample mopropy1)urib acid (XV). prepared by alkylation of 8-chlorotheophylline with nThe cleavage of the oxazolidines V-VI11 with hydropropyl bromide. On heating with sodium ethoxide chloric acid was analogous to that observed with hydrosolution compound X X I was converted to 1,3-diniethyl7-n-propyl-8-ethyluric acid (XXII) which with hydro(1) This work was partially supported by Polish Academy of Sciencechloric acid gave XX. 1,3,7-Triniethyl-6H,7H-oxazoCommittee of Pharmaceutical Sciences. (2) lf. Eckstein, Dissertationes pharm.. 14, 425 (1962). lo [2,3-f]xanthine (VII) is not altered by heating with (3) E. Fischer and F. Ach, Ber., 39,430 (1906). zinc and acetic acid, indicating that reductions of V and ( 4 j hI. Eckstein a n d Lf. Gorcsyca, A. Zejc. Disserfafiones pharm., 16, 61 VI with zinc and acetic acid described above involve (1964).

\

\

NOTES

OCTOBER,1964

3127

of uric acid with alkali furnishes new possibilities of obtaining 7-substituted derivatives of 1,3-dimethyl-GH,7H-oxazolo [2,3-f]xanthine. Studies on the chemical properties of oxazoloxanthines and related imidazolidinexanthiness and hexahydropyrimidinexanthines8 are under way.

\

Experimental !

CH3 XI11

0 II

H,c-J~,.-cH~-cH=cH~ O

N

I

N

CH3

H3C-N

O

H

XIV

'$--J

Hf

CHrCHz-CH3

0

N

C1 C

CH3 XXI

s

H

L

0

H3C-SJ&N-CH2-CHz-CHs I

CH3

XXII

initial loss of hydrogen halide which then cleaves the oxazolidine. The reaction of VI11 with sodium iodide in methyl ethyl ketone also gave XX rather than the expected product VI.5 The transformations produced by hydrochloric and hydrobromic acid on the 7-substituted l13-dimethyl6H17H-oxazolo[2,3-f]xanthines (V-VIII) are in agreement with previous observations on the instability of oxazolidines6 and oxazolines' in acid media. These ring-opening reactions of the theophylline-oxazolidines give good yields and represent a convenient method for obtaining 7-substituted derivatives of uric acid. The above-described reaction of 7-/3-chloroalkyl derivatives ( 5 ) Examples of reduction b y heating with sodium iodide are known in t h e literature: e.g., sulfonyl chlorides are reduced t o the sodium salts of sulfinic acids, with collateral formation of certain a m o u n t of sulfoxide and sulfone [R.O t t o a n d J. Troeger, Ber., 24, 482, 486, 488, 494 (1891); S. Gebauer-Fulnegg a n d E. Riesz, Monatsh., 49, 41 (1928)). (6) E. D. Bergmann. Chem. Rev., SS, 309 (1953). (7) J. D. Loudon in "Chemistry of Carbon Compounds," Vol. IV, E. H. R o d d , Ed., Elsevier Publishing Co., Amsterdam, 1957, pp. 363, 357, 376.

The starting products, 7-( p-hydroxy-p-bromopropyl)-8-chlorotheophylline ( I ) , 7-(p-hydroxy-p-iodopropyl)-8-chloroetheophylline (TI), 7-(~-hydroxy-p-bromopropyl)-8-bromotheophyIline (111), 7-(p-hydroxy-p-iodopropyl)-8-bromotheophylline (IV), and 7-allyl-8-chlorotheophylline ( X I I ) , were obtained according t o methods described previously.' 1,3-Dimethyl-7-bromomethyl-6H,7H-oxazolo [2,3-f]xanthine (V). Method A.-To a solution of 3.51 g. (0.01 mole) of I or 3.96 g. (0.01 mole) of I11 in 20 ml. of 60% ethanol (or 50 ml. of water), a solution of 0.4 g. (0.01 mole) of sodium hydroxide in 10 ml. of water was added, and the mixture was heated in the water bath for 0.5-1 hr. (until neutral reaction was reached). The product separated on cooling and crystallized from 96Yc ethanol, m.p. 254-256' dec. The product was slightly soluble in water, benzene, and chloroform, and somewhat more soluble in ethanol and acetone; yield 88%. Method B.-From the cooled reaction mixture obtained by hydroxybromination of 0.01 mole of 7-allyl-8-chloro- or -8-bromotheophylline,' after making basic with sodium hydroxide a t room temperature, a product was obtained in 85Yc yield, m.p. 254296". It was not depressed on admixture with the compound prepared by method A . Method C.-A solution of 0.39 g. (0.001 mole) of SI in 5 ml. of 507, ethanol containing 0.04 g. (0.001 mole) of sodium hydroxide was heated for 10 min. on the water bath. A product separated from the hot solution, which after crystallization from ethanol had m.p. 254-256" which was not depressed on admixture r i t h the compound, obtained by method A or 13. Anal. Calcd. for CI0HllBrN4O3: C, 38.13; H, 3.55; N, 17.79. Found: C, 38.48; H, 3.77; N, 18.10. 1,3-Dimethyl-7-iodomethyl-6H,7H-oxazolo [2,3-f]xanthine (VI). Method A.-Under conditions analogous with those described by method A to obtain V, 3.98 g. (0.01 mole) of I1 or 4.43 g. (0.01 mole) of IV gave a product in 8570 yield which, when crystallized from 70% ethanol, melted a t 214-215' dec. The compound was slightly soluble in water, ethanol, benzene, acetone, and chloroform. Method B.-From the reaction mixture after hydroxyiodination of 7-allyl-8-chloro- or -8-bromotheophylline,' described by method B for compound V, a substance was obtained in 80% yield, m.p. 214-215" dec., which did not depress the mixture melting point with the product that was obtained by method A above. Anal. Calcd. for CloH,JN,O,: C, 33.18; H , 3.06; N , 15.48. Found: C,33.19; H,3.11; X , 15.78. 1,3,7-Trimethy1-6H,7H-oxazolo [2,3;f] xanthine (VII) .--A solution of 0.32 g. (0.001 mole) of IX or 0.27 g. (0.001 mole) of X T I in 5 ml. of 50% ethanol containing 0.04 g. (0.001 mole) of sodium hydroxide was heated on a water bath until the alkaline disappeared. The product separated after cooling and, after reA crystallization from 207, ethanol, melted at 243.5-244.5'. mixture melting point with the compound obtained by another method2 showed no depression, Reaction Products of 7-Substituted 1,3-Dimethyl-6H,7H-oxa2010 [2,3-f]xanthine with Hydrobromic Acid. 1,3-Dimethyl-7(8-bromopropy1)uric Acid (IX).-After dissolving 2.36 g. (0.01 mole) of VI1 in 25 ml. of hydrobromic acid ( d 1.48), t,he solution was refluxed for 6 hr. On cooling, dilut,ing with water, and crystallizing from 50% ethanol, a product with m.p. 250-251' was obtained which was moderately soluble in water and ethanol and readily soluble in solutions of alkali, from which the unaltered product was separated (yield 7 0 5 ) after acidification. 1,3-Dimethyl-7-(p-bromo-~-chloropropyl)uricAcid (X), Method A.-A solution of 2.7 g. (0.01 mole) of 1 - 1 1 1 2 in 25 ml. of hydrobromic acid ( d 1.48) was refluxed for 1 hr. Itecrystallization from 807, acetic acid gave the product, m.p. 272-272.5", which was insoluble in water and ethanol and readily soluble in ( 8 ) M. Eckstein, Dissertationes phorm., 14, 435 (1962).

NOTES

3128

VOL. 29

TABLE I 7,8-DISUBSTITUTED lj3-DIMETHYLXAXTHINEs

a

Compd.

RI

Ra

IX X XI XI11 XIV XVI XSTI XT'III XIX XX XXI XXII

CH2CH(Br)CH3 CHzCH(Br)CHzC1 CHzCH(Br)CHzBr CHzCH=CHz CHzCH=CHz CHzCH( Cl)CH3 CHzCH(CI)CH,CI CHzCH( C1)CHzI CHzCH(C1)CH2Br CHpCHzCH3 CHzCHpCH3 CH,CHzCH,

OH OH OH OCzK OH OH OH OH OH OH

c1

OCzHs

%-

% '-

0c.a

-Calcd., C

H

N

C

n

N

250-251 272-272.5 265-266 93-94 257-259 255-256 253-254 231-231.5 255-256 2 55-2 5 6 123.5-124 80.5-81

37.79 34.12 30.33 54.59 50.90 43.99 39.12 30.10 34.18 50.46 46.79 54.18

4.13 3.44 3.05 6.11 5.13 4.80 3.94 3.03 3.44 5.93 5.10 6.82

17.68 15.92 14.15 21.22 23.74 20.56 18.25 14.04 15.92 23.54 21.80 21.06

37.50 33.91 30.50 54.54 51.07 44.15 39.44 30.30 34.48 50.21 47.45 54.50

4.14 3.44 3.13 6.03 5.03 4.78 3.92 2.98 3 55 5.59' 5.25 7.02

17.41 15.66 14.55 21.15 24.04 20.20 18.01 13.92 16.21 23.18 21.84 21.19

M.P.,

-Found

All melting points are corrected; for recrystallization solvents and yields, see Experimental.

alkali, from which the unaltered product was separated on acidification. Method B.-To a solution of 2.7 g. (0.01 mole) of T'II12 in 50 ml. of chloroform cooled to 5", 1.75 ml. of hydrobromic acid ( d 1.48) was gradually dropped in 30 min. The reacbion mixture was allowed to stand for 12 hr. at approximately 0". Removal of the solvent and purification as described above gave a product with m.p. 272" and m.m.p. (with compound obtained in A) 272". 1,3-Dimethyl-7-(8,-pdibromopropyl)uric Acid (XI). Method A. -A solution of 3.1 g. (0.01 mole) of V in 25 ml. of hydrobromic acid ( d 1.41) was refluxed for 1 hr. A product, which separated from the hot solution and crystallized from 80% acetic acid, melted a t 265-266"; yield 837,. The compound was very slightly soluble in water and ethanol, slightly in acetic acid, and readily in alkali. Method B.-A solution of 3.62 g. (0.01 mole) of VI in 25 ml. of hydrobromic acid ( d 1.41) was refluxed for 4 hr. (the sublimation of iodine occurred). The product, which separated from t,he hot solution and crystallized from 807, acetic acid, melted at 2 6 5 266" and did not^ depress the melting point on admixture with compound obtained in A . Method C.-To a solution of 2.36 g. (0.01 mole) of XI\' in 100 ml. of chloroform was slowly dropped 2.4 g. of bromine in 10 ml. of chloroform at' room temperature, and the mixture was allowed to stand for 24 hr. The solvent was removed and the residue was crystallized from 80% acetic acid yielding 70% of X , m.p. 265". The melting point was not depressed on admixture with samples obtained in A and B , 1,3-Dimethyl-7-ally1-8-ethyluric Acid (XIII).-To a solution of 0.46 g. (0.02 g.-atom) of metallic sodium in 100 ml. of absolute ethanol was added 5.08 g. (0.02 mole) of 7-allyl-8-chlorot.heophy1line ( X I I ) and the mixture was refluxed for 1 hr. (until appearance of neutral reaction). The solution was then evaporated under reduced pressure; the residue crystallized from water had m.p. 93-94"; yield 90%. 1,3-Dimethyl-7-allyluric Acid (XIV).-A solution of 5.2 g. (0.02 mole) of XI11 in 15 ml. of concentrated hydrochloric acid was heat,ed for 3 min., then cooled, diluted with water, filtered, and washed several times with m-ater , Recrystallization from water and then from absolute et,hanol gave the sample, m.p. 257-259', yield 91yc. 1,3,9-Trimethyl-7-(~,~-dibromopropyl)uric Acid (XV).-Into a solution of 0.4 g . (0.001 mole) of XI in 5 ml. of water and 0.04 g. of sodium hydroxide \vas dropped 0.1 ml. of dimethyl sulfate (about l o o ) , and t'he mixture was then heated on the water bath for 0.5 hr. The product, separated while hot, was washed with dilute sodium hydroside and crystallized from ethanol, yielding 7OCi; of a compound of m.p. 210-211.5° dec. dnal. Calcd. for CI1HI4BrzNIOB: C, 32.22; H, 3.44. Found: C, 32.16; H , 3.65. Reaction Products of 7-Substituted 1,3-Dimethyl-6H,7H-oxa2010 [2,3-f]xanthine with Hydrochloric Acid. 1,3-Dimethyl-7-

(6-chloropropy1)uric Acid (XVI).-A solution of 2.36 g. (0.01 mole) of VI1 in 20 ml. of concentrated hydrochloric acid was refluxed for 2 hr. The product,, separated while hot, was washed several times with water, and after crystallization from ethanol It was very slightly soluble in xater and melted a t 255-256'. readily soluble in alkali, from which the unaltered product was separated on acidification; yield 65%. 1,3-Dimethyl-7- (P-chloro-ychloro-, -?-bromo-, or -r-iodopropyl-) uric Acids (XVII, XVIII, and XIX).-From t,he compounds T'III, VI, and T' in a manner described for compound XT'I, but with prolongation of the reaction time to 4 hr., were obtained XVII, m.p. 253-254" (from 50% et,hanol); XF'III, m.p. 231231.5' (from aceic acid); and XIX, m.p. 255256' (from 707, ethanol). 1,3-Dimethy1-7-propyluricAcid (XX). Method A.-A mixture of 0.002 mole of S7 or XI, 2 g. of zinc dust, 1 ml. of glacial acetic acid, and 60 ml. of ethanol was refluxed for 4 hr. From the filtrate a product, m.p. 255-256" (from water or ethanol), was obtained; yield 60-65%. Method B.-A solution of 0.27 g. (0.001 mole) of T'III and 0.9 g. (0.006 mole) of sodum iodide in 20 ml. of methyl ethyl ketone was refluxed for 30 hr. The solvent was removed under reduced pressure. The residue was crystallized from water and ethanol and gave X X , m.p. 255-256", yield 587,, which did not depress the melting point on admixture with compound obtained in A . Method C.-A solution of 0.52 g. (0,002 mole) of X X I I in 2 ml. of concentrated hydrochloric acid was boiled for 3 min. After cooling, diluting with water, and recrystallizing, a product, m.p. and m.m.p. 255256', was obtained. 7-Propyl-8-chlorotheophylline (XXI). Method A.-A mixture of 2.36 g. (0.01 mole) of X X , 1.2 ml. of dimethyfaniline, and 5.6 ml. of phosphorous oxychloride was refluxed for 20 hr. The solution was poured into ice and the product, after crystallization from water, melted at 123.5-124"; yield 49Y0. Method B.-To a solution of 2.14 g. (0.01 mole) of 8-chlorotheophylline and 0.4 g. (0.01 mole) of sodium hydroxide in 15 ml. of 60% ethanol was added 1 ml. of n-propyl bromide and heated until disappearance of the alkaline reaction (approximately 3 hr.). After partial evaporation of the solvent under reduced pressure, the solid was fibered, washed with water, and crystallized from anhydrous ethanol to yield 85% of X X I , m.p. 123.5124'. A mixture melting point with compound obtained in A showed no depression. 1,3-Dimethyl-7-propyl-8-ethyluric Acid (XXII).-.k solution of 2.57 g. (0.01 mole) of X X I in 100 ml. of absolute ethanol containing 0.23 g. (0.01 g.-atom) of metallic sodium was refluxed for 1 hr. The solvent was evaporated under reduced pressure, and the residue crystallized from anhydrous ethanol had m.p. 80.j-81'; yield 857,. The properties of campounds IX-XI, X I I I , XIV, and XVIX X I I and elemental analyses are shown in Table I.

3129

NOTES

OCTOBER,1964 A Convenient Preparation of Benzo[k]fluoranthene

Reactions of Cyclooctanone Oxime Tosylate with Organolithium Reagentsla

H. W. WHITLOCK, JR.

NORMAN W. GAB EL^^

Department of Chemistry, University of Wisconsin, Madison, Wisconsin

Department of Chemistry, University o f Chicago, Chicago, Illinois

Received March 97, 1.964

Receivcd M a y 28, 1964

A useful modification of the Wittig reaction is that employing alkylphosphonate in place of alkylidinetriphenylphosphoranes. I n the presence of sodium methoxide, acenaphthenequinone and phosphonate I 5 condense smoothly to afford a moderate yield of benzo [klfluoranthene.

The recent disclosure by Pelosi and Lyle2 on the reactions of oxime tosylates with Grignard reagents has prompted the reporting of a similar investigation involving cyclooctanone oxime tosylate 1 and organolithium reagents. The purpose of these experiments was to determine whether an electron-deficient nitrogen could be produced by the attack of a n organolithium reagent on the C=S bond of a n oxime tosylate with concomitant loss of the tosyl group.

NaOCH?

""1 I

4

The yield in this condensation, 20-2770, is relatively insensitive to the order of addition of the reagents, the nature of the nietal alkoxide (LiOCH, or NaOCH3), or the presence of methanol in the reaction n ~ i x t u r e , ~ but drops to less than 1% when the reaction is conducted a t 105'. Experimental

A solution of 2.0 g. (37 mequiv.) of sodium methoxide in 20 ml. of anhydrous methanol was added over a period of 0.5 hr. to a slurry of 1.82 g . (10 mmoles) of acenaphthenequinone and 4.0 g. (10.5 mmoles) of tetraethyl o-xylylenediphosphonate in 25 mi. of dry IIMF which was stirred a t 0" under a nitrogen atmosphere. On completion of additisn of the sodium methoxide solution the dark reaction mixture was allowed to stir a t 0" for 1 hr. and a t 25' for 2 hr. and then poured into 150 ml. of water. The aqueous layer was extracted with five 50-ml. portions of chloroform and the combined chloroform extracts were evaporated. The dark residue was taken up in 150 ml. benzene and the benzene solution was washed with 10-ml. portions of water until the aqueous extracts were colorless followed by saturated salt solution, dried over anhydrous sodium sulfate, and evaporated. The dark residue was chromatographed on 100 g. of alumina (Fisher Scientific Co.). Elution with 5% benzene in hexane afforded, after recrystallization from benzene, 554 mg. (22% yield) of benzo[k]fluoranthene as faintly yellow flakes, m.p. 215-216" (lit.6 m.p. 2 1 i 0 ) , whose ultraviolet spectrum agreed with the published spectrum.'. Anal. Calcd. for C Z ~ H ~C, Z : 95.21; H, 4.79. Found: C, 95.18; H, 4.78. Large-scale reactions using nonchromatographic work-ups were not attempted.

Acknowledgment.-Partial support by the National Science Foundation is gratefully acknowledged. (1) W. S. W a d s a o r t h , J r . , and W. P. Emmons, J . A m . Chem. Soc., 83, 1733 (1961). ( 2 ) E. J. Seus a n d C. V. Wilson, J . Org. Chem., 16, 5243 (1961). (3) H. 0. Huisman. Chem. Weekblad, 69, 133 (1963). (4) S. T r i p p e t t , Quart. Reu. (London), 17, 406 (1963). ( 5 ) W . Stilz and H . Pommer. German patent 1,124,949 (1962); Chem. Abstr., 6 7 , 3358c (1962). (6) E. Clar. "hromatische Kohlenwasserstoffe," Springer Verlag, 1952, p. 408. (7) R. A . Friedel and M. Orchin, "Ultraviolet Spectra of Aromatic Compounds." J o h n Wiley and Sons, Ino., New York, N. Y . , 1951, spectrum 543.

2

5

6

a, R1 = CH,; Rz = H

b , Rl C , R1 d , R1

= = =

Rz = H CH,; Rz = C&,SOzC&;

C6H5;

Rz = CsH5SOz-

I n view of the ease with which cyclooctane systems undergo transannular reactions, it seemed plausible that, if an electron-deficient nitrogen were generated, a transannular insertion reaction might produce an 9-azabicyclo [3.3.l]nonane derivative 2. Since it was recognized that a Stieglitz rearrangement3 could occur rather than, or in addition to, the desired reaction, the possibility that the resultant crude product would be a mixture of 2, 3 , 4 , 5 , and 6 was considered. To facilitate the separation of these possible products, the material obtained from the reaction mixtures was treated with aqueous acid in order to hydrolyze any Schiff's bases 3 and 5. After extraction with ether the acidic solution was made alkaline, and the resulting basic fraction was treated with benzenesulfonyl chloride in aqueous sodium hydroxide. The reaction of methyllithium with 1 a t -30" was highly exothermic and yielded only one isolable product 4a (71y0 as the benzenesulfonamide 4c). The infrared spectra of the crude and purified material were essentially the same. The structural assignment was based primarily on spectral evidence. The presence of a gerndimethyl group was clearly indicated by infrared absorptions at 1385 and 1370 cm.-'. The n.m.r. spec(1) (a) .4bstracted f r o m p a r t of the P h . D . Dissertation of N. \V. Gahel, Deo., 1961; ( b ) Biological Research Laboratories, Department of Psychia t r y , College of Medicine. University of Illinois, C h i c a g o Ill. (2) S. S. Pelosi, J r . , and R. E. Lyle, Ahstracts. l 4 i t h National Meeting of t h e American Chemical Society, Philadelphia, P a . , April, 1964, p. lis. (3) (a) J. Stieglitz and P. N. Leech, Jr.. J . Am. Chem. Soc., 36, 272 (1914): ( b ) A I . L. Cohn. Ph.D. Dissertation. Cniversity of Chicago, 1929.

NOTES

3130

trum disclosed a multiplet with center a t r 2.41 (phenyl hydrogens), a small broad peak at 7.35 (hydrogens of methylene group adjacent to heterocyclic nitrogen), a sharp resonance peak at 8.30 (methyl hydrogens /3 t o Sulfonamide nitrogen),* and a peak a t 8.56 (methylene hydrogens). The ratio of aliphatic to aroniatic hydrogens was 4. Although methyllithium appears to have added t o the C=S double bond of 1 with concomitant loss of the tosyl group, the absence of a n insertion product 2a which might have resulted from the formation of an electron-deficient nitrogen intermediate suggests that loss of the tosyl group did not proceed concertedly with addition of a nucleophile. Futhermore, since there is probably little difference between the intrinsic migratory aptitudes of methylene and methyl groups, the absence of any product 6a resulting from migration of the methyl group seems to indicate an intermediate similar to the one which is depicted for the Beckmann rearrangement. The tosyl group is removed via the electrophilic attack of the lithium from the methyllithium ion pair. The intermediate 7 then reacts with methyllithium to give 3a and then finally 4a.

VOL. 29

I

I

I

H

H

H

tPhLi 1

Ph

I

LJ

I

Ph I

-OH I

A 11

I

H

H

9

10

Ph

I

(@HZ)f ?-OH \C-NHSO*Ph I H 8

Yeber rearrangement' the mechanism of which has been previously proposed by Cram and Hatch.7c As corroboration for the structural assignment MeLi of 8, its n.m.r. spectrum in deuteriochloroforni dist closed two sharp resonance peaks at 7 2.81 and 3.11 --+ N , d 3a +4a I1 ' . (phenyl hydrogens), a diffuse doublet with center a t --CH*-C . . CHy5.00 (sulfonamide hydrogen), a diffuse peak with center 7 a t 6.25 (hydrogen CY- to amide nitrogen), a broad peak a t 7.76 (hydroxyl hydrogen), and a broad peak with When phenyllithium and 1 were stirred together for center a t 8.64 (methylene hydrogen). After the deu5 hr. a t room temperature, treatment of the reaction teriochloroform solution was shaken with deuterium mixture in the previously stated manner again yielded oxide, the peaks at 5.00 and 7.67 disappeared demononly one isolable product (62% as the benzenesulfonstrating that both were due to exchangeable protons. amide) having a n empirical formula of C20H2,1V03S. Since there are two asymetric carbon atoms in 8, This immediately excluded 2b, 4b, Sb, and 6b as prodthere is a possibility of two pairs of enantiomorphs ucts of the reaction. Although the benxenesulfonyl arising from this reaction sequence. However, during derivative of the amine resulting from hydrolysis of 3b the s N 1 acid-catalyzed hydrolysis of the aziridine 10, has the same empirical formula, the carbonyl region of the most thermodynamically stable isomer with both the infrared spectrum was blank. The infrared specphenyl and amino groups in quasi-equitorial positions trum of a dilute chloroform solution showed sharp should be formed. This would result in the isolation of absorption peaks a t 3610 and 3385 cm.-' which can be only one pair of enantiomorphs as the product. attributed to nonbonded 0-H stretching and K-H An explanation for the difference between the reacstretching of a nionosubstituted benzenesulfonamide.5 tions that 1 underwent with methyllithium and phenylThe only structure which can be written to accomolithium can be found in their differing behavior as elecdate the foregoing evidence and can also be rationalized trophilic reagents. Alkyllithium compounds are known on the basis of the probable course of the reaction is 8. to act as electrophilic reagents owing to coordination of What apparently occurs is a base-induced ?-elimination the lithium with centers of electron density in such reacof the tosyl group followed by ring closure to an intertions as nietalations, halogen-metal interchange, and mediate azirine 9 which is then attacked by a second substitution of aromatic rings.* Phenyllithium, in mole of phenyllithium to give l-ph'enyl-g-azabicyclocontrast, is a very poor electrophile, evidenced by its [ G . 1.O]nonane 10. Acid-catalyzed hydrolysis of the greatly decreased reactivity compared with alkyllithiums aziridine ring gave 1-phenyl-1-hydroxy-2-an1inocyclo- in the aforementioned reactions.8 Therefore, it did not octane 11 which was isolated as its benzenesulfonanlide bring about a Beckmann type of rearrangement but in8. Campbell, et ~ l . have , ~ shown that acid-catalyzed stead served as a base for the 7-elimination reaction. hydrolysis of 2-arylaziridines only yields amino alcohols with hydroxy groups and aryl groups on the same carbon Experimental atom. It should be aoted that reaction 1 -t 9 is a

,+

(4) The resonance peak due t o t h e methyl hydrogens of t h e t-butyl group

Reaction of Cyclooctanone Oxime Tosylate with Methyllithium. -To a solution of 10.0 g. (0.034 mole) of cyclooctanone oxime

of S-1-butyl-p-toluenesulfonamide a a s found to occur a t 7 8.82. I t is possible t h a t in 40 the phenyl ring is allowed to exert a greater deshielding effect. (5) J. N . Baxter. J. Cymerman-Craig, and J. B. Willis, J . Chem. Soc., 1966, 669. (6) K. ?i. Campbell, 9. K. C+mphell, L. G. Hess, and I. J . Schaffner, J . O r g . Chem.. 9 , 184 (1944).

( 7 ) (a) P. W. Neher and A . yon Friedolsheim, Ann.. 449, 109 (1926); (b) P. W. Neber, .4. Burgard. and W. Thier, ibrd., 636, 277 (1936); (e) D. J. Cram and M. J. Hatch, J . Am. Chem. Soc., 76, 33 (1953): (d) M. J. Hatch and D. J. Cram, ibid., 7 6 , 38 (1953); ( e ) H. 0. House and W. F. Berkowite, J. O r g . Chem., 2R, 307, 2271 (1963). (8) H. Gilman and J. W. Morton, Jr.. O r g . Reactions,8,258

(1964).

OCTOBER,

NOTES

1964

toeylateg in 300 ml. of dry 5 : 1 ether-benzene a t -30" was added dropwise 70 ml. of 1.5 M methyllithium (0.1 mole) in ether. A vigorous, exothermic reaction took place while the addition was cariied out. A t the end of the addition period the reaction mixture was allowed to warm up to room temperature and was then poured into 250 ml. of ice-water. The organic layer was separated, dried over Wa2SOa,and evaporated a t reduced pressure to a dark oil which was then stirred with 25 ml. of 1 N HC1 for 1 hr. The acidic solution was extracted with ether to remove neutral material and was then made strongly basic and reextracted with ether. S o cyclooctanone was found in the first extract. The second extract was evaporated to a dark oil which was treated with benzenesulfonyl chloride in aqueous sodium hydroxide. The basic aqueous solution was separated from insoluble material and upon acidification gave only a small amount of benzenesulfonic acid. The insoluble material yielded, after recrystallization from ethanol, 7.2 grams (71%) of a benzenesulfonamide of a secondary amine, m.p. 84". An,al. Calcd. for CL6H25N02S (295.44): C, 65.04; H, 8.41; N, 4.74; S, 10.85. Found: C, 64.70; H, 7.79; N, 5.08; S, 10.69. Reaction of Cyclooctanone Oxime Tosylate with Phenyllithium. -A solution of 5.0 g. (0.017 mole) of cyclooctanone oxime tosylate in 150 ml. of dry ether and 50 ml. of dry benzene was cooled to -30". With stirring under nitrogen 35 ml. of 1.5 'iA phenyllit,hium (0.05 mole) in ether was added dropwise over 15 min. KOevolu$on of heat occurred when the reagent wae added. The mixture was very slowly warmed to room temperature and stirred for 5 hr. The crude product was processed in the same manner as in the preceeding experiment and yielded 3.7 g. (62%) of a benzenesulfonamide, m.p. 140" (ethanol). Anal. Calcd. for C2~H2,N03S (359.48): C, 66.82; H , 7.01; N, 3.90. Found: C, 66.85; H, 7.04; N, 4.00.

Acknowledgment.-The author would like to thank Professor G. L. Closs for suggesting this problem. (9) (a) P. W. Neher and G. H u h , A n n . , 616, 283 (1935); (b) P. Oxley and W. F. Short, J . Chem. Soc., 1514 (1948); (c) W. 2. Heldt, J . A m . Chem. Soc., 80,5880 (1958).

A New Bicyclic Thiete Sulfone' DONALD C. DITTMERA N D FRANKLIN A. DAVIS

Department of Chemistry, Syracuse University, Syracuse, New York 15210 Received M a y 12, 1964

Only four reports of the synthesis of thiete sulfone (thiasyclobutene 1,l-dioxide) derivatives, all monocyclic, are found in the l i t e r a t ~ r e . ~ -We ~ were interested in the preparation of substituted thiete sulfones since these conipounds might not undergo ring opening on reduction with lithium aluminum hydride as do several of the more simple derivative^.^,^ Reports by Stork and Borowitz' and by Opitz and Adolfs of the synthesis of aminothietane sulfones by reaction of enamines with niethanesulfonyl chloride in the presence of a base prompted us to apply this method to the synthesis of the bicyclic thiete sulfone, 7-thiabicyclo[4.2.0]-1(8)-octene 7,7-dioxide (1). (1) This work was supported b y G r a n t GP-726 of the National Science Foundation. (2) D. C. Dittmer and hl. E. Christy, J . Org. Chem., 26, 1324 (1961). (3) W. E. Truce, J. R. Norell, J. E. Richman, and J. P. Walsh, Tetrahedron Letlers, 1677 (1963). (4) W . E. Truce and J . R . Norell, J . A m . Chem. Soc., 86,3236 (1963). (5) R. H . Hasek, P. G. G o t t , R . H. Meen, and J. C. M a r t i n , J . Orp. Chem., 28, 2496 (1963). (6) D. C. Dittmer and M. E. Christy, J . A m . Chem. Soc., 84, 399 (1962). (7) G. Stork and I. J. Boroaita. ( b i d . , 84, 313 (1962). (8) G. Opitz a n d €1. Adolph, Angew. Chem., '74, 27 (1962).

3131

:m02 2

1

The reaction of methanesulfonyl chloride with cyclohexanone pyrrolidine enamineg in benzene in the presence of triethylamine gave an 80% yield of the cyclic sulfone 2. Treatment of 2 with niethyl iodide in

I-

l

I

2 heat

niethyl ethyl ketone gave the quaternary salt (30y0 yield) which was stirred overnight with freshly prepared silver oxide to yield, upon heating to 40-50°, the thiete sulfone 1, in 3540% yield. A better procedure is to convert the intermediate amine 2 to the amine oxide which on heating is converted to the unsaturated sulfone in 55-65% yield. The structure of the bicyclic thiete sulfone 1 was supported by its infrared spectruni, proton magnetic resonance spectruni, elemental analysis, and niolecular weight. Its infrared spectrum showed absorption at an olefinic C-H stretching frequency of 3003 c1n-l and at a C-C double bond stretching frequency of 1608 cm. -l. The proton n.1ii.r. spectrum provides further proof that the double bond is ex0 to the six-inembered ring. Absorption (at 60 hlc. relative to tetraniethylsilane) occurs at 6 2.0 (relative area 8, protons of cyclohexane ring), 4.5 (relative area 1, 6-proton), and 6.2 (relative area 1, 8-proton). Had the double bond been between carbons 1 and 6, only two principal proton absorptions would have been observed in the ratio 8:2. The e m product is the one expected on the basis of a trans elimination of N-inethylpyrrolidine. Reduction of 1 with lithium aluminum hydride gives an oil (lacking infrared absorption for a sulfone or an olefin) which can be oxidized to a sulfone which is identical with the product obtained on catalytic reduction of the double bond of 1. This result strongly suggests that the oil has structure 3, and that this reduction of a thiete sulfone with lithium aluniinuni hydride is the first one which occurs without opening of the four-ineiiibered ring. The fused six-nien~bcred ring may hinder attack on the 6-position by hydridc ion

(9) G . Stork, A . Brizaolara, H . Landesman, J. Samuszkovicz, and R Terrell, J . A m . Chem. Soc., 86, 207 (1963).

3132

NOTES

(complexed with aluminunl) which had been suggested to account for the ring opening.6 This reduction is under study. Further work is in progress on the chemlstry of this new bicyclic unsaturated sulfone. Experimental Cyclohexanone pyrrolidine enamine is ,prepared from pyrrolidine and cyclohexanone.9 Triethylamine is dried over potassium hydroxide, and methanesulfonyl chloride and methyl iodide are used as obtained. Silver oxide is prepared as described in Organic Reactions.1o Commercial 40% peracetic is used as obtained, Cyclohexanone Pyrrolidine Enamine Sulfone ( 2 ) .--Methanesulfonyl chloride (114.0 g., 1.00 mole) in 300 ml. of dry benzene is added dropa-ise with stirring to triethylamine (151.5 g., 1.5 moles) and cyclohexanone pyrrolidine enamine (151 .0 g., 1.0 mole) in 1000 ml. of dry benzene. The reaction mixture is cooled in an ice bath during the addition and is done in an atmosphere of dry nitrogen. After the addition of methanesulfonyl chloride the reaction mixture is stirred overnight in the ice bath. The triethylamine hydrochloride is removed by filtration and washed with benzene. These washings are combined with the filtrate and the benzene is removed on a rotary evaporator a t a bath temperature below 40". The last traces of solvent are removed under high vacuum. The light yellow sirup is dissolved in absolute ethanol and allowed to crystallize at 0" t'o give 172 g. (0.750 mole, i S 7 ; ) of colorless crystals, m.p. 42-43". Anal. Calcd. for CI1H,,NO,S: C , 57.59; H, 8.35; X , 6.12; S, 13.98. Found: C, 57.49; H,8.26; S , 6 . 1 2 ; S,13.85. Quaternary Salt of 2.-A slight excess of methyl iodide is Bdded to cyclohexanone pyrrolidine enamine sulfone (5.0 g., 0.022 mole) in 100 ml. of methyl ethyl ketone. Use of ether or 1,2-dimethoxyethane as solvent is less satisfactory. After 2 days at room temperature, the salt is removed by filtration and recrystallized from ethanol-ether to give 2.5 g. (0.0067 mole, 30%) of colorless crystals, m.p. 168-170'. Anal. Calcd. for C12H2JSS02: C, 38.32; H , 5.97; N, 3.78; S,8.63. Found: (2,3859; H,6.07; N,3.82; S,8.84. 7-Thiabicyclo[4.2.0]-1(8)-octene 7,7-Dioxide ( 1 ) . A.-Freshly prepared silver oxidel0 (13.3 g., 0.058 mole) is added with stirring to the quaternary iodide (10.0 g., 0.027 mole) of 2 in 200 ml. of water through Tvhich nit,rogen has been passed. The reaction mixture is stirred for 12 hr. and the precipitated salts are removed by filtration, heated under vacuum (water pump) a t 50" for 1 hr., cooled, and extracted with chloroform. The chloroform extracts are dried over calcium sulfate and evaporated to a clear sirup which is recrystallized from ether to yield 1.7 g. (0.011 mole, 41%) of white crystals of the bicyclic sulfone, m.p. 88-89'. The infrared spect,rum of the sulfone showed absorption a t 3003 (s), 2933 ( m ) , 2841 (w),1608 ( w m ) , 1443-1422 (m), 1328 (m), 1292 (vs), 1225-1200 (vs), 1170 (vs), 1140 (s), 1070 (m), 1042 ( w ) , and 927 ( m ) em.-'. The proton n.m.r. spectrum in CDC13 (10) A . C. Cope and E. R. Trumbull, Org. Reactions, 11, 317 (1960).

VOL. 29

(tetramethylsilane reference) showed complex absorption from 60 to 175, complex.absorption a t 268, and a doublet ( J S 6= 2 c.P.s.) a t 377 C.P.S. The relative areas were 8: 1: 1 respectively. ~ 2 c.P.s., is about that expected in a The coupling constant, J s = thiete sulfone for a l&interaction.Z Anal. Calcd. for C~H,OO%S: C, 53.14; H, 6.37; S, 20.27; mol. wt. (boiling point elevation in bromoform), 158. Found: C, 53.28; H , 6.48; S, 20.16; mol. wt., 165. B.-Peracetic acid (20 ml., 40y0) is added dropwise during 30 min. to cyclohexanone pyrrolidine enamine sulfone 2 (10 g . , 0.044 mole) in a 125-m1. erlenmeyer flask cooled in a n ice-salt bath. After addition of the peracetic acid, the mixture is kept in the bath for an additional 30 min. and a t room temperature for 4 hr. Glacial acetic acid (10 ml.) is added, and the mixture is cooled and neutralized with saturated sodium carbonate solution. The solution is heated a t 60" under vacuum (water pump) for 2 hr., cooled, saturated with sodium chloride, and extracted with chloroform or ether. The extracts are washed with 10% hydrochloric acid and dried over sodium carbonate. Removal of the solvent leaves 4.0 g. (0.025 mole, 574%) of white needles, m.p. 85-88". The infrared spectrum was identical with the product obtained by the Hofmann elimination procedure. Reduction of the Bicyclic Thiete Sulfone (1) with Lithium Aluminum Hydride.-The thiete sulfone 1 (3.0 g., 0.019 mole) in 100 ml. of dry ether is added to lithium aluminum hydride (2.3 g., 0.058 mole) in 50 ml. of dry ether. The hydride and ether have been refluxed for 30 min. prior to the addition. The reaction mixture is stirred for 3 hr. at room temperature followed by destruction of the excess hydride by successive additions of 20 ml. of ammonium chloride (20% solution) and 2 ml. of concentrated ammonium hydroxide in 5 ml. of the ammonium chloride solution. Evaporation of the ether layer, after it is dried over calcium chloride, gives a yellow oil ( 3 , 2.1 g . ) . The infrared spectrum of the oil shows bands a t 2940 (s), 1450 ( m ) , 1310 (w),1270 ( w ) , 1160 (w), and 1140 (w) cm.-'. Oxidation of the Lithium Aluminum Hydride Reduction Product.-Hydrogen peroxide (30%, 1.62 g., 0.0143 mole) is added to 10 ml. of a solution of acetic acid and 0.6 g . of the oil obtained on reduction of the bicyclic thiete sulfone with lithium aluminum hydride. The mixture is heated for 1 hr. at about 100" and allowed to stir a t room temperature overnight. The reaction mixture is poured into water, made basic with 20% sodium hydroxide, and extracted with ether. The ether extract is dried and evaporated to give 0.7 g. of a pale yellow liquid which solidifies at about 10". The infrared spectrum of the oil shows absorption a t 2941 (s), 2865 (m), 1449 (m), 1414 (m), 1309 (s), 1294 (s), 1212 (m), 1193 (m), 1160 (s), 1139 (s), 1105 (m), and 762 ( m ) cm.-l. Reduction of 1 With Hydrogen.-Hydrogenation of 1 (3.0 g., 0.019 mole) in 50 ml. of absolute ethanol a t 40 p.s.i. over 50 mg. of 10% palladium on charcoal for 4 days yields a clear oil which solidifies a t 10" and crystallizes from ether a t -10". The infrared spectrum is identical with that of the product of the oxidation of the material obtained on reduction of 1 with lithium aluminum hydride. Anal. Calcd. for C,HIZOZS:C , 52.48; H , 7.55. Found: C, 52.50; H , 7.58.