DECEMBER
1957
Arapahoe Chemicals, Inc., Boulder, Colo. Some of the Nbromoglutarimide was prepared as described below. Glutarimide. A solution of 70 g. (0.53 mole) of glutaric acid in 150 ml. of concentrated aqueous ammonia was heated under a steam-heated reflux condenser for 7 hr. as the temperature of the mixture rose from 90-180". The temperature was held a t 170-180' until the evolution of ammonia ceased (about 1.5 hr.). The reaction mixture solidified on cooling. Crystallization from acetone yielded two crops totalling 37.4 g. (63%) of glutarimide, m.p. 145-146' which was somewhat lower than 151-152' r e p ~ r t e d . ~ ~ ~ N-Bromoglutarimide. The general methodlo involving bromination of the silver salt of the imide in trifluoroacetic acid was used. To a mixture of 32.0 g. (0.14 mole) of silver oxide in 200 ml. of trifluoroacetic acid was added 32.1 g. (0.28 mole) of glutarimide. After an hour of stirring 45.6 g. (0.29 mole) of bromine was added over a 3-hr. period. The silver bromide was removed by filtration, and the solution was concentrated a t reduced pressure. Absolute ether (150 ml.) precipitated 45 g. (84%) of N-bromoglutarimide, m.p. 135-140'. Reactions of N-bromo compounds with styrene. The reactions were carried out on each N-bromo compound with half the equivalent amount of styrene (stabilized with p-tertbutylcatechol) based on available bromine in the N-bromo compound. I n each case the solvent was 50 ml. of alcoholfree chloroform. Each mixture was heated under reflux until a negative test with acidified potassium iodide solution was obtained or until a t least 10 hr. had elapsed. At the end of the reaction time the solution was filtered hot to remove any insoluble products, then cooled in an ice-salt bath and filtered again if further precipitate was produced. After filtration the chloroform was removed under reduced pressure and the resulting residue recrystallized from 95% ethyl alcohol in order to obtain any styrene dibromide formed during the reaction. The results are summarized in Table I. Reaction of N-bromogluhrimide with cyclohexene. A mixture of 9.7 g. (0.118 mole) of cyclohexene, 5.0 g. (0.026 mole) of N-bromoglutarimide, and 50 ml. of carbon tetrachloride was heated under reflux for 2.5 hr. under an ultraviolet lamp. The reaction mixture was cooled and filtered to give 2.72 g. (92%) of glutarimide. Distillation of the filtrate yielded 2.75 g. (66%) of 3-bromocyclohexene, b.p. 76-78' (3 mm.); d:' 1.365;n?f1.5285. Reaction of N,N-di6romobenzenesulfo~~idewith styrene. A mixture of 1.3 g. (0.0125 mole) of styrene and 4.0 g. (0.0127 mole) of N,N-dibromobenzenesulfonamide in 60 ml. of chloroform was boiled under reflux for 20 min. The solution was washed twice with 10% sodium hydroxide and was then dried over calcium sulfate. Evaporation of the chloroform yielded a residue which was crystallized from 95% ethyl alcohol t o give 1.54 g. (48%) of N-benzenesulfonylstyreneimine, m.p. 75-76". Anal. Calcd. for ClrHrsNOzS: C, 64.8; H, 5.05; N, 5.40. Found: C, 64.1; H, 4.96; N, 5.35. In a similar experiment twice as much styrene was used; the reaction mixture was heated under reflux for 25 hr.; and no sodium hydroxide solution was used. The product was crystallized from alcohol to give 6.0 g. (90%) of N-(2-bromol-phenylethy1)-N - (1- bromo-2-phenylethy1)benzenesulfonamide, m.p. 136-137'. Anal. Calcd. for CzzHzlNOaBrzS: C, 50.5; H, 4.05; N, 2.68. Found: C, 50.8;H,4.12;N,2.86. DEPARTMENT OF CHEMISTRY STATEU N I ~ R S I TOF Y IOWA IOWA CITY,IOWA
(8) S. S.G.Sircar, J. C h m . Sac., 600 (1927).
(9) 0. Bernheimer, Gam. chim. ital., 12, 281 (1882). (10) A. L. Henne and W. F. Zimmer, J . Am. C h m . Soc., 73,1103 (1951).
1729
NOTES
Regeneration of Alkaloids from Their Picrates with an Anion-Exchange Resin' JAMESM. BOBBITT Received M a y 16, 1957
The characterization of organic nitrogen bases as salts of the polynitro phenols, picric acid (2,4,Gtrinitrophenol) and styphnic acid (2,4,6-tr initr oresorcinol) is almost a universal practice and has been the subject of much work.2 These salts are easily prepared and are generally well defined, nonhygroscopic, crystalline compounds amenable to analysis. The picrates can often be purified by recrystallization and then decomposed with aqueous inorganic bases to yield the pure, free amines. The decomposition of picrates by aqueous base is complicated by the low solubility of sodium, potassium, and ammonium picrates in ~ a t e rThis . ~ ~complica~ tion can be alleviated with lithium hydroxide3 or ethanolamine4J' producing the more soluble lithium and ethanolamine picrates. Difficulty has also been encountered in the decomposition, by base, of the picrates of organic nitrogen compounds which are base-labile,6 easily oxidized,6 or contain phenol groups conferring solubility in aqueous base.' The use of lamb's wool6 (giving good yields, 80-97%, of the base-hydrochlorides) or hot aqueous potassium chloride' (poor yields) has partially overcome these obstacles. The decomposition of inorganic salts of alkaloids and similar compounds on anion-exchange resins and subsequent titration of alkaloid in the eluate has been developed as an analytical tool, and has been critically reviewed.8 Tropone and G-hydroxytropinone have been obtained from their picrates with the weakly basic resin Amberlite IR-45 and strongly basic Dowex 2, respe~tively.~ This paper suggests the use of the bicarbonate salt of the strongly basic resin, Amberlite IRA400 [IRA4OO(HCOa)] for the decomposition of alkaloids by a "columnar" technique. This technique
(1) First conceived and tested in Professor Carl Djerassi's laboratory a t Wayne State University. (2) (a) L. Kofler and F. A. Miiller, Mikrochemie, 22, 43 (1937); (b) A. Oliverio and F. S. Trucco, Atti accad. Gioenia sci. nut. Cataniu, [6] 4, (1939); Chem. Abstr., 35,7115 (1941); (0) C. Massatsch, Pharnz. Ztg., 83,210, (1947). (3) A. Burger, J. Am. Chem. Soc., 67, 1615 (1945). (4)I. A. Kaye, I. C. Kogon, and W. Burlant, J . Am. Chem. Sac., 72,5752 (1950). (5) N. Weiner and I. A. Kaye, J . Org. Chem., 14, 868 (1949). (6) H. MuIler, Hoppe-Seyter's 2. physio2. Chem., 209, 207 (1932). (7) D. Taber and A. I. Kosak, J . Org. Chem., 21, 257 (1956). (8) L. Saunders, J . Pharm. Phamacol., 5, 569 (1953). (9) E. E. van Tamelen, Patricia Barth, and F. Lornitzo, J . Am. C h .SOC.,78,5442 (1956).
1730
NOTES
VOL.
22
TABLE 'I RECOVERY OF ALKALOIDS FROM PICRATE SALTSWITH IRA-W( HCOI) Picrate m.p., "C.
Alkaloid
PKt
Cinchonidined
Codeine
5.80 10.03 5.85 9.92 6.05
Morphine
6.13
Reserpine
7.401a
Narcotine
7.80
Papaverine
8.07
Cinchonined
207- 209 O (208- 209') 222-225" (212-215')pb*' 196-197' (196-197°)11 164167' (163-165°)11 180-183' (183- 186")'* 174-170' (175') 186-189' (183')
Recovery,
%
95. ge 96.9 100. OQ 86. 4h 98.8 98.0 10O.O8
Recovered Base M.P., 'C? 205-206' (210") 261-262' (265") 154-157" (154- 156')" 249-253 O (254") 256-258" (277 " )I* 176-178' (176') 14% 150" (147")
Resin Needed, GJM. Eq. BaseC 5.4 5.8 5.1 7.2h 5.9 4.6 3.6
a Melting points were taken on a Kofler apparatus. Recorded melting points are given in parentheses. Unless noted, physical constants are taken from The Merck Index, 6th Ed., Merck & Co. Inc., Rahway, 1952. Melting points of recovered alkaloids were identical with those of starting material. Mixture melting points were undepressed. Estimated from amount of red "spent" resin formed. Density of wet resin was 0.52. Form di-picrates. e Salted out with aqueous sodium chloride. f Also rec. 194' and 198-210°.~0 Isolated by "freeze-drying." Flow rate, 100 ml. in 20 min. Only 75% recovery a t rate of 100 ml. in 60 min.
has been briefly described'O and used for the deSatisfactory results were obtained with all of the composition of the picrates of 6-methoxy-, 6,7-di- alkaloids except morphine and no decomposition methoxy-, 6-methoxy-7ethoxy-, 6,7-dimethoxy-8- was noted in any case. The relatively low recovery hydroxy-, and 6,7,8-trimethoxy-l-isobutyl-2-meth- (8601,) of morphine is presumably due to partial yi-1,2,3,4-tetrahydroisoquinolinesin reference to retention on the resin through its phenolic group. the structure proof of the cactus alkaloid, pilocer- Lower yields of morphine were obtained at slower eine. This method has now been applied to the pic- flow rates, however, in other cases the flow rate over rates of seven alkaloids of varying structures and the resin was not crucial. At least 2-7% of water in base strengths (pKb, 5.8-10.03).Results are given the solvent was mandatory, The nearly neutral in Table I. A study of flow rate over the resin, sol- character of Amberlite IRA-400(HCOa)allowed the vent ratios, and the amount of resin used yielded the use of acetone, an excellent solvent for picrates. On strongly basic resins, acetone condenses with itdata in Table 11. self,l4 A "batch" technique was not successful beTABLE I1 cause of the hydrolysis of Amberlite IRA400 (picEFFECT O F FLOW RATEAND SOLVENT VARIATION O N AMOUNT rate). The resin could be effectively regenerated OF RESINUSEDO after use. Flow Rate Solvent, Ml./Hr. % Water in Acetone 160 75 15 9 40 40
10 10 10 10
OC
40
1 5
40
10
Resin Needed,
G./M. Eq. Base 16.0 7.4 5.8 2.9 No picrate decomposition 28.9 7.4 6.6
a Decomposition of 0.2 g. of codeine picrate in 40 ml. of solvent. Resin was pre-washed with the solvent used. Resin was washed with acetone and air-dried before use.
(10) C. Djerassi, S. K. Figdor, J. M. Bobbitt, and F. X. Markley, J. Am. Chem. Soc., 7 8 , 3861 (1956); 79, 2203
(1957). i l l ) K. W. Bentley, The Chemistry of the Morphine Alkaloids, Oxford University Press, London, 1954, pp. 81, 30. (12) N. Neuss, H. E. Boaz, and J. W. Forbes, J . Am. Chem. Soc., 75,4870 (1953). (13) L. Dorfman, A. Furlenmeier, C. F. Huebner, R. Lucas, H. B. MacPhillamy, J. M. Mueller, E. Schlittler, R. Schwyzer, and A, F. St. Andr6, Helv. Chim. Ada, 37, 69 (1954).
EXPERIMENTAL
Alkaloids. Except for reserpine which was donated by the Upjohn Co. of Kalamazoo, Mich., the alkaloids used were commercial samples obtained through the School of Pharmacy of The University of Connecticut. Preparation of picrates. One g. of alkaloid in 50 ml. of hot 95% ethanol was combined with a slight excess of picric acid in 50 ml. of hot 95% ethanol. The solution was allowed to cool and the precipitate was filtered and dried. Preparation of resin.16 Two hundred g . of Amberlite IRA400(C1)16in an appropriate column were washed with 3 4 1. of 10% sodium bicarbonate solution until the eluate gave no chloride test (silver nitrate). It was then washed with 4 5 1. of distilled water and stored under distilled water. Decomposition of p i c r d s . (Table I ) A column of resin (28 X 1.2 cm.) waa washed With 100 ml. of 10% aqueous acetone
(14) R. Pa,llaud and 6. V. Austerweil, Compt. rend., 240, 1218 (1955); C. J. Schmidle and R. C. Mansfield, Ind. Eng. Chem.,44,1388 (1952). (15) Different from and much superior to the method previously presented.10 (16) Provided by the Rohm & Hass Co., Philadelphia, Pa.
DECEMBER
1957
1731
NOTES
and agitated by inversion to remove gas bubbles (10% resin shrinkage occurred). One half g. of picrate in 100 ml. of 10% aqueous acetone was allowed to flow over the resin a t a rate of 30-60 ml./hr. The column was washed with 100 ml. of solvent, the washings and eluate were combined, and the acetone was removed under vacuum. The resulting aqueous phase was filtered, salted out (sodium chloride) and filtered, or ((freeze-dried" to obtain pure crystalline alkaloid. Regeneration of resin. The bright red "spent" resin was transformed immediately after use back into Amberlite IRA400(C1) by washing with a solution of 50 ml. of concentrated hydrochloric acid in 200 ml. of acetone until no more yellow color was eluted.
Acknowledgments. The author is indebted to Dr. Roy J. Gritter of this department and Professor Carl Djerassi for much helpful advice.
?-
CH3
IV
diazomethane. Acetic anhydride gave an acetyl derivative, V, which from the infrared spectrum had the indicated structure rather than the alternate (Vaj. OCOCHS
0
DEPARTMENT OF CHEMISTRY THEUNIVERSITY OF CONNECTICUT STORRS,CONN. EXPERIMENTAL
I-Hydroxy-2,4-Di- t-butylphenazine TODVC'.
CAMP BELL^
Received M a y 17, 1967
It has been proposed2 that the oxidation of 4,6di-t-butylpyrogallol in alkaline solution proceeds via a highly colored intermediate hydroxy-o-quinone :
Evidence for such an intermediate o-quinone has been obtained by oxidation of di-t-butylpyrogalol to an intermediate purple compound, either by air oxidation in alkali or by bromine oxidation in buffered acetic acid solution. The intennediate purple compound coupled wit,h o-phenylenediamine t o give l-hydroxy-2,4-di-t-butylphen~izine (111).
This phmazine could not be methylated with dimethylsulfate t o zi p y q a n i n e :;ne (9 V) derivst h e 3 4 It was recovered unch ,nged from ethereal (1) I'resent! address: E. pany, Wilmngtor,, Del.
I. d u Pont d e Kernours & Corn.-
(2) T.W. CampbeYi, J ~ iim. . Chern. SOC.,73, 4!W (1.95::. (3: A. R. Surrey, Org. Syn~he (4) Wred- and Strack, R a v . Chcnz.. 74, 181 l84,is5 !192V ,
d,4-Di-t-butyl-l-hydroxyphenazine. Ten grams of dibutylpyrogallol2was dissolved in 150 ml. of acetic acid containing 15 g. of sodium acetate. To this solution was added 6.5 g. of bromine in 50 ml. of acetic acid. The solution at once developed a brilliant purple color. The mixture was allowed to stand for 15 min. after which 5 g. (10% excess) of ophenylenediamine dissolved in 50 m:. of acetic acid was added. The purple coior changed instantly to a brown-orange color. This mixture was allowed to stand for 48 hr. and the acetic acid was evaporated under nitrogen on a steam bath. The solid residue was extracted with 50 ml. of methyl alcohol containing a little concentrated hydrochloric scid. This dissolved most of the brownish orange solid. The product which remained undissolved was then treated with water to remove potassium bromide and the bright orange crystalline residue was dried. This crystalline residue was dissolved in chloroform and the chloroform solution was washed repeatedly with water. The bright orange chloroform layer was retained and the aqueous washirgs were discarded. Evaporation of the dried chloroform solution gave an orange crystalline solid which was recrystallized from chloroformmethanol mixture. The product was obtained in s yield of 3.2 g. (21%) with a melting point of 170.5'. A deep blue solution resulted when the product was dissolved in alcoholic alkali. Anal. Calcd. for CsoIiz4Nz0:C, 77.9; H, 7.9; N, 9.1. Found: C, 77.4,77.5; H, 7.76,7.8;N, 9.1, 9.4. The same product was obtained by the air oxidation of a mixture of di-t-butylpyiogallol and o-?henylenediamine. However, the yield was lower and the product difficuk to isolate. d,4-Di-t-butyE-l-hydroxyphenazineacetate. Twc hundred milligrams of 2,4-di-i-but,yl-l-hy~r~~~phenazin~mixed neth 3 cc. of acetic xiihydride plus 5 cc. of p mixture was boiied 0.5 hr., then cooled a.nd al!owed to stand for 48 hr. Ne crystallization had occiirred so the mix;Lire was diluted witk water. The gunmy materia; v5ich precipitated was rubbed with a stirring rod to indune crystailization. The solid 80 obtained was filtere6 and reerystailizes horn aqueous methanol. T h e pale yeiiom crj-stallinr po:\-der had a, melting point of lfili.5-162". Anal. Calcd. for C22H2H02X2: K >S.9. S'orin5: ?T, 8.3, In aLka!ine methano:, a blue coior dsveloped s!o indicating siow hvdroiysis to the p Infrared spectra qf ;il,e phenazine trum of the parmi: :hecazine shon, -OM band, shiftad :;2 2.95 p, This bonding to the e,djxcent nitroger:. showed a,n ester band at 5.65 fi, no
