4772
COMMUNICATIONS
VOL.
26
in each the strong electron donor properties of the pyrrole Ir-system6 is utilized. Tri-2-pyrrylphosphine oxide (IV) has an intense band in the ultraviolet a t 237.5 mp (log e 4.06) and the N-methyl analog (V) has a similar band at sir: 248 mp (log e 4.11).6The parent structures pyrrole’ The occurrence of Ir-bonding involving overlap and N-methyl pyrrolea show no such intense of a phosphorus &orbital with a porbital of an bands; their absorption drops smoothly from 225 adjacent atom has been firmly established for a mp to 300 mp (pyrrole’ has two weak maxima a t number of tetracovalent phosphoryl compounds 262,268 mp; log e 0.92,0.92). A substituent acting (I).I This bonding is conventionally expressed in only inductively would be expected to produce no terms of canonical structures such as I1 and 111. change in the general appearance of the spectrum, though slight bathochromic shifts and hyperchromic effects are expected.* The appearance of a new and intense band in the spectra of IV and V is indicative of a conjugative interaction between the phosphoryl group and the ~relectrons of the pyrrole ring. For pyrroles containing substituents capable of Such dpIr bonding has only been observed in in- conjugation in the 2- position (e.g. carbonyl and stances in which anti-bonding electrons of elec- phenyl) two intense bands are observed a t 228tronegative atoms, primarily oxygen, nitrogen, 252 mp (log e 3.57-3.70) and 263-2895 mp (log E and sulfur, are donated to the empty phosphorus 4.10-4.22),s~10designated as bands A and B, d-orbital. In no instance has a comparable inter- respectively, by Andrisano and Pappalardo.g For action utilizing a ?r-bond system such aa benzene example, pyrrole-2-carboxylic acid has maxima been demonstrated. In an intensive study, Jaffe, a t 228, 258 mp (log e 3.65, 4.10).O Since only a Freedman, and Doak2ea examined the ultraviolet single band is observed in the spectra of IV and V, absorption spectra of triphenylphosphine oxide it is probable that this band is related to band and a number of phenylphosphonates and found B; the hypsochromic shift observed for IV and V evidence for only very weak &-pa bond formation, relative to the a-carbonyl substituted pyrroles ini.e., bathochromic shifts of uents cont,ained powerful (5) A. R. Katritsky, Qziatt. Revs. (London), 13, 353 electron donatsinggroups.‘ We now wish to report (1959). the first two cases in which such bonding occurs; (6)The spectra of these compounds were determined 88 dilute solutiolls (C = 5 X 10-6 M)in 95% ethanol amploy-
Ultraviolet Absorption Spectra of Tri-2-pyrrylphosphine Oxides : Evidence for dwpr Bonding
( 1 ) J. R. Van Wazer, “Phosphorus and Its Comporinds,” Vol. 1, Interscience, New York, 1958;H. Finegold, J . Am. Chem. Soc., 82,2641(1960);L.Lamon, Svensk Kem. Tidskr., 71, 336 (1959). (2) H.H. Jaffe, J . Chem. Phya., 22, 1430 (1954). (3) H. H.Jaffe and L. I).Freedman, J . Am. Chem. Soc., 74, 1069 (1952);L. D. Freedman and G . 0.Doak, J . Am. Chem. Soc., 77,6221 (1955). (4)K. D. Berlin and G . B. Butler, Chem. Reus., 60, 243 (1960).
ing a Cary Model 14 spectrophotometer. Satisfactory analytical data and sukmtantiation of structure have been ohtained for IV and V (R. A. Polsky, M. S. thesk, University of Pittsburzh, 1961)and will be reported elsewhere. (7) K. Ladenburn and C. R. Noller, J . Am. Chem. Soc., 63; 1240 (1941\. ( 8 ) L. C. Craig. L. Michaelis, S. Granich, and W. A. Jah0.L. J . B i d . C k . , 154, 293 (1944). (9) R. Andrisano and C. Pappalardo, Gazz. chim. ital., 85, 1430 (1955). (10)B. Elpern and F. C. Nachod, J . Am. Chem. Soc., 72,3379 (1950).
NOVEMBER
1961
and a number of 2-pyrrylphosphonate structures are in progress and will be reported in detail. This study was supported in part by a research grant (CY-5338) from the National Cancer Institute, Public Health Service and by an American Cancer Society Institutional Grant to the University of Pittsburgh. E. GB" DEPARTMENT OF C ~ M X S T B YCLAIBOUELNE ROSALYN A. Poxany UNIYER~ITY OF PITISBTJB~E PI'l"SBURQH
13, P A .
Steroids and Related Natural Products. VI. Dihorane Reduction of Lactones to Cyclic Hemiacetalsk'
sir: Several carbohydrate lactones have been reduced, under carefully controlled conditions, to cyclic hemiacetals by sodium or potassium boroh ~ d r i d e .A ~ few examples of lactone -* cyclic hemiacetal conversion using lithium aluminum hydride have also been reported.' We now wish to report a new and convenient procedure for reducing lactones to cyclic hemiacetals using diborane.6.6 This useful reduction re-
v; G. R. Pettit, B.
Green, and
action has been applied to synthesis of several unusual oxasteroids. In a typical experiment, the diborane prepared from sodium borohydride (0.2 g.) and boron trifluoride etherate (1 g.), in diglyme, was passed (over 3 hr.) into a solution of 3.0~04 oxa-5a-cholestane (0.39 g.)7 in tetrahydrofuran a t room temperature. Following addition of ethyl ether and water, the crude product was isolated. Chromatographic separation led to 38-hydroxy4 om-5a-cholestane (Ia, 0.28 g.) ; m.p. 197-199O, [ L L ] ~ , 106O (chloroform). Anal. Calcd. for Cw HMO*: C, 79.94; H, 11.87; 0, 8.19. Found: C, 79.61; H, 11.52; 0, 8.58. The cyclic hemiacetal structure (I) was confirmed by chromic acid oxidation to 3-oxo-l-oxadc~-cholestane and transformation to 3~-methoxy4oxa-5acholestane(Ib, m.p. 106-107°, A w l . Calcd. for G7Ha02: C, 80.14; H, 11.96; 0, 7.91. Found: C, 80.45; H, 11.85; 0, 7.47.) by hydrobromic acid-methanol.
+
Received JdyZ4, 1961
(1) Part
4773
COMWONICATIONS
W. J.
7
R
& H
R&l
IIa. R = R I = 0
Ia. R = OH b. R=OCHZ c.
H
b. R 0, R,
R=H
='"I 'S
Bowyer,
J.Org. C l m . , 26,2879 (1961).
(2) This investigation was supported by PHS Reaearch Grant CY-4074(C2) from the National Cancer Institute, Public Health Service, and National Science Foundation Research Grant G-9585. (3) These studies have been reviewed by E. Sehenker, Angm. Chem., 73,81(1961). (4) Cf.;G. E. Arth, J . Am. Chem.SOC.,75,2413 (1963); J. Schmidlin, G. Anner, J. R. Billeter, IC. Heusler, H. Ueberwaeaer, P. Wieland, and A. Wettstein, Relv. Chim. Acta, 40, 1034 (1957); and H. Obara, Nippon Kagaku Zmehi, 82,58(1961). ( 5 ) While reaction between dihrane and rbutyrolactone has been obaerved, the product was not described: H. C . Brown and B. C. Subba Rao, J. Org. Chem., 22, 1135 (1957). Interestingly, bis3-methyl-2-butylbrane haa been reported to reduce both 7-butyrolactone and 7-valerolactone to hydroxyaldehydes: H. C. Brown and D. B. Bigley, J. Am. Chem. Soc., 83,486 (1961). (6) Several aspects of diborane chemistry and toxicology have been described by J. Cueilleron and P. Guillot, Bull. soc. chim. Fmnce, 2044 (1960). The potential value of this interesting substance in organic synthesis is illustrated by a number of recent investigations. For example, see: J. Kollonitsch, J . Am. Chem. Soc., 83, 1515 (1961); H. C. Brown and G. Zweifel, J. Am. Chem. Soc., 83, 1241 (1961); H. C. Brown and C. H. Snyder, J . Am. Chern. SOL,83,1002 (1961); H. C. Brown, C. Verbrugge, and C. H. Snyder, J . Am. Chem. SOC.,83, 1001 (1961); W. Jeffera,Ch. and I d . (London), 431 (1961); R. KWer and G. Griaznov, Angew. Chem., 73, 171 (1961); N. Noth and G. Mikulaschek, Chem. Rer., 94, 634 (1961); H. C. Brown and G. Zweifel, J . Am. Chem.Soc., 83,486 (1961); M. F. Hawthorne, J . Am. C h . Soc., 8 3 , 367 (1961); ref. 3; and a review prepared by H. C. Brown, OrganmnetaUic Chemistry, H. Zeiss, Ed., Reinhold Publishing Corp., New York, 1960, p. 150.
/dY
H0
Conversion of 3,20dioxo-4-oxa-5a-pregnsne (IIa)8 to 20ethylene-thioketal derivative IIb (m.p. 237239O, A d . Calcd. for CBHNOS~:C, 66.96; H, 8.69; S, 16.30. Found: C, 66.70; H, 8.58;S, 16.54.), followed by Raney nickel desulfurization gave 3-oxo-l-oxa-5a-pregnane (IIc) ; m.p. 185-190°, [a]5, 107' (chloroform), (Anal. Calcd. for C22H3202: C, 78.89; H, 10.59. Found: C, 78.38; H, 10.33.). Diborane reduction of lactone IIc (0.3 g.) provided hemiacetal IId (0.2 g.); m.p. 161-163', CY]^, 69' (chloroform). Anal. Calcd. for CzoHa,02: C, 78.38; H, 11.18; 0, 10.44; Active H, 0.33. Found: C, 77.93; H, 11.11; 0, 10.79; Active H, 0.25. Diborane reduction of 3j3-hydroxy-17-oxo-17a-
+
+
(7) A. Salamon, 2.physiol. Ghem., 272, 61 (1941). Conclusive evidence for the structure and stereochemistry of the lactone prepared by this procedure will be presented in a subsequent communication. ( 8 ) 6. R. Pettit and T. R. Krtsturi, J . Org. Chem., 26, 986(1961).
4774
COMMUNICATIONS
oxa-5a-androstane (1.25 g.)gto hemiacetal I11 (0.32 g.) was also easily accomplished. A pure specimen melted a t 226-228'. AWE. Calcd. for C19Ha03: C, 73.98; H, 10.46; 0, 15.56. Found: C, 74.14; H, 10.27; 0, 15.73. We also wish to report that diborane reduction of certain lactones in the presence of boron trifluoride etherate yields the corresponding ether derivative.1° For example, when reduction of 3oxo-4-oxa-5a-cholestane (0.39 g.) was repeated employing diborane-boron trifluoride etherate the product was 4-0xa-5a-cholestane~~ (IC,0.07 g.). The stereochemistry assigned hemiacetals Ia, IId, and I11 received substantial support when each was recovered following equilibration in acidified (hydrochloric acid) tetrahydrofuran or methanol solution. Several other facets of this unusually mild route to hemiacetals are now under investigation. '
GEORGE R. PETTIT T. R. KASTURI BRIANGREEN JOHN C. KNIGHT Received August 9, 1961
DEPARTMENT OF CHEMISTRY UNIVERSITY OF MAINE ORONO,ME.
(9) M. F. Murray, B. A. Johnson, R. L. Pederson, and A. C. Ott, J. Am. Chem. SOC.,78,981 (1956). (10) Cf., G. R. Pettit, U. R. Ghatak, B. Green, T. R. Kasturi, and D. M. Piatak, J. Org.Chem., 26,1685 (1961). (11) The 4-oxasteroid (IC) was identical with an authentic specimen kindly supplied by Dr. J. T. Edward. See, J. T. Edward and P. F. Morand, Can. J.Chem., 38,1325 (1960).
VOL.
26
of these compounds and find them to be one of the most potent series of orally active progestogens known to date. A toluene solution of 5,6a-epoxy-16-pregnene3/3+1-20-0ne acetate (I) was added to an excess of ethereal methyl magnesium bromide containing powdered cuprous chloride. The magnesium enolate so formed was treated in situ with acetic anhydride3 for periods of up to 3 days to form GP,lGa-dimethyl15(20)-pregnene-3~,5a,20-triol triacetate (11). The crude compound I1 on treatment with peracetic acid gave the 17(20)-epoxide which without purification was saponified with methanolic potassium carbonate. The crude 60,16a-dimethylpregnane-3&5a,l7a-triol-2O-one 5-acetate (IIIa) was purified as the 3,bdiacetate (IIIb), m.p. 206.5208.5', [a]? -27.5°.4 Saponification of IIIb with potassium bicarbonate in aqueous methanol gave IIIa as a hydrate, map. 186-190°, [a]? -27.3', which on oxidation with 8N chromic acid-sulfuric acid reagent in acetone gave G&lGa-dimethylpregnane-5a,17a-diol-3,20-dione5-acetate (IIIc), m.p. 161-163' (hydrated form), [a]? -26.9'. Refluxing a sohtion of IIIc in ethanol containing hydrochloric acid effected 0-elimination together with isomerization at C-6 producing 17a-hydroxyGa,l6a-dimethylprogesterone (IVa).S The 17aacetate was prepared in the usual manner.' Dehydrogenation of IVa with chloranil in tbutanol8 gave 17a-hydroxy-A6-dehydro-6,16a-dimethylprogesterone (Va) which was also converted to its 17a-acetate (Vb). The A'-dehydro com-
l7a-Acetoxy-6,16a-diniethylprogesterones
(3) K. Heusler, J. Kebrle, C. Meystre, H. Ueberwasser, P. Wieland, G. Anner, and A. Wettatein, Helv. Chim. Acta, 42, 2043 (1959). sir: (4) All melting points determined on a micro hot stage; In a recent communication' we reported the syn- all rotations in chloroform. Sataifactory elemental analyses thesis of a series of 6,16a-dimethylproge~terones.~obtained for all new compounds described herein. (5) See ref. 2b. Reported m.p. 200-202", [ a ] D f 51.4', We have now prepared the 17a-acetoxy derivatives X?2OH 240-242 u, e 16,600. 71.1", (6) See ref. 2b. Reported m.p. 170-172", [ a ] D (1) R. P. Graber and M. B. Meyers, Chem. & Id., 1478, X$:60H 240 mp, e 13,800. (1960). (7) Cf.C. G. Bergstrom, P. B. Sollman, R. T. Nicholson, (2) 6a,l6a-Dimethylprogesteronehas been reported recently by (a) s. Bernstein, E. W. Cantrall, and J. P. Dusza, and R. M.Dodaon, J. Am. Chem. Soc., 82,2322 (1960). (8) E. J. Agnello and G. D. Laubach, J. Am. Chenz. SOC., J. Org. Chem., 26, 269 (1961) and (b) J. Iriarte and M. L. 82,4293 (1960). Franco, J. Org. Chem.,26,2047, (1961).
+
TABLE I Compound M.P. 17a-Hydroxy-6a,l6a-dimethylprogesterone( IVa) 203-207 " 17a-Acetoxy-6a,l6a-dimethylprogeaterone(IVb) 169-171' 17a-Hydroxy-A~-dehydro-6,16a-dimethylprogesterone (Va) 220.5-229" 17a-Acetoxy-A6-dehydro-6, 16a-dimethylprogeekrone (Vb) 189.5-195' 17a-Acetoxy-A1-dehydro-6a,16~-dimethylprogeskrone (VI) 168-173' 17a-Acetoxy-A~~~-bisdehydro-6,16~-dimethylprogesterone 161.5-163' (VII)
a
[ffID'
A,,CSHIOH -(e)
Oral Activityb
$53.5' $69.0'
242(15,400) 242( 15,800)
0.25 55
+27.9"
290(23,200)
-
$25.6'
288( 24,900)
130
$21.7"
245( 15,800)
40
-24.2'
228( 12,000) 256(8470) 302( 11,540)
120
Temperature ca. 25". Clauberg assay; ethinyl testosterone = 1. Assays by Endocrine Laboratories, Madison, Wis.
1961
NOVEMBER
4775
COMMUNICATIONS
pounds, VI and VII, derived from IVb and Vb product was warmed, its infrared spectrum changed were readily obtained by use of 2,3-dichloro-5,6- radically; the band a t 5.65 p disappeared and three dicyanobenzoquinone (DDQ) in refluxing b e n ~ e n e . ~new bands a t 6.05, 6.25, and 6.38 p appeared. Table I gives the physical constants of the series Distillation gave a 93% yield of l-dimethylaminoof 17-oxygenated-6,16a-dimethylprogesteronesand 4-methyl-1-pentene-3-one (IV) , b.p. 105-107 ' (2 their respective oral progestational activities. mm)., nio 1.5301. Anal. Calcd. for CsHISNO:C, CENTRAL RESEARCH LABORATORIES ROBERTP. GRABER 68.0; H, 10.6; N, 9.9. Found: C, 68.0; H, 10.9; N, GENERAL MILLS,INC. MARTINB. MEYERS 9.7. The structure of IV was confirmed by indeMINNEAPOLIS 13, MINN. pendent synthesis from 3-methyl-2-butanone, methyl formate, and dimethylamine.' Received August 17, 1961 (9) D.Burn, D.N. Kirk, and V. Petrow, Proc. Chem. SOC., 14 (1960).
CHa C \'
CHI
0
II
C
+
e + II
Cycloaddition of Ketenes to Enamines
Sir: We wish to report some new reactions involving cycloaddition of ketenes to enamines derived from aldehydes. Ketene and dialkylketenes react with enamines having either one or no 0-hydrogen to give 3dialkylaminocyclobutanones. If the resulting cyclobutanones have one or more a-hydrogens (111, V, VII), they undergo an irreversible, ring-opening reaction; cyclobutanones that have no a-hydrogens (I) are quite stable. Dimethylketene and N,N-dimethylisobutenylamine, when mixed in isopropyl acetate a t room temperature, reacted to give a 64% yield of 3-dimethylamino-2,2,4,4-tetramethylcyclobutanone (I), b.p. 83-85' (24 mm.), ng 1.4439, infrared maximum a t 5.65 1.1 (cyclobutanone). The NMR spectrum of I was in complete agreement with the proposed structure. Anal. Calcd. for CloH1gNO: C, 71.1; H, 11.2; N, 8.3. Found: C, 71.3; H, 11.2; N, 8.1. Quaternization of I with methyl tosylate followed by treatment with aqueous potassium hydroxide solution gave, after acidification, a 71% yield of 2,2,4-trimethyl-3-pentenoic acid (11), b.p. 86' (2 mm.), ng 1.4472. Anal. Calcd. for CsH1402: C, 67.6; H, 9.9; neut. equiv., 142. Found: C, 67.6; H, 10.1; neut. equiv., 142.
0
H2CC=O (CH3)zN-(!d(CH& H
CHd b H 3
\H
( CHs)zC-C=O
CHI
I
I1
I
I
CH3
Ketene reacted with N ,N-dimethylisobut,enylamine in hexane at 0' to give the intermediate 3-dimethylamino-2,2-dimethylcyclobutanone (111), identified by the characteristic infrared absorption of cyclobutanones at 5.65 1.1. When this reaction
I11
0
0
11
il + CH3OCH
(CH3)zCH CH,
IC:
CH( CH3)2
IV
I
TI
NaOCHa
(CH:)zNH.HCl
0
!I
( CH~)~CH&CH=CHON~
Berchtold, Harvey, and Wilson have isolated a cycloaddition product similar to 111 (N-morpholino in place of the dimethylamino group) and noted its thermal rearrangement to the acyclic aminovinyl ketone.2 It has also been brought to our attention that a keto base is found in the acylation of an enamine with acetyl chloride, apparently by abstraction of hydrogen chloride to form ketene, and subsequent cycloaddition to the enaminee3 Dimethylketene and N - (l-butenyl)piperidine4 in
u
CHf
c
'CH3
n
H
(CH3)zK'
+ (CHg)zNCH=CH
V
VI
hexane reacted a t -20' to give 2-ethyl-4,4-dimethyl-3-piperidinocyclobutanone (V), which, on distillation, gave an 82% yield of 2-ethyl-4-methyl1-piperidino-1-penten-3-one (VI), b.p. 114-121 O (0.6 mm.), n$' 1.5424,infrared maxima a t 6.05, 6.25, and 6.38 p. An.aZ. Calcd. for C l 3 H d O :C, 74.6; H, 11.0; N, 6.7. Found: C, 74.6; H, 11.0; N, 6.8. (1) E.Benary, Ber., 63,1573 (1930). ( 2 ) G.A.Berchtold, G. R. Harvey, and G. E. Wilson, J . Org. Chem., 26,4776 (1961). (3) G. Opitz, private communication. (4) C . Mannich and H. Davidsen, Ber., 69,2106 (1936).
4776
VOL. 26
COMbfUNICATIONS
Ketene and N-(1-buteny1)piperidine in hexane reacted at -20' to give 2-ethyl-3-piperidinocyclobutanone (VII), as evidenced by the infrared maximum at 5.65 p. Rearrangement of VI1 occurred during distillation to give a 7Oy0 yield of a mixture of 2ethyl-1-piperidino-1-buten-3-one(VIII) and l-piperidino-l-hexen-3a1e (IX) that was not separated by fractional distillation. The mixture had a boiling point of 137' (0.3 mm.), n z 1.5544, infrared maxima at 6.05,6.25, and 6.38 p. An estimate made from the NMR spectrum showed the mixture to be about two parts of I X to one part of VIII. Anal. Calcd. for CllHlsNO: C, 72.9; H, 10.5; N, 7.8. Found: C, 72.7; H, 10.5; N, 7.7.
The thermally unstable cyclobutanones, 111, V, and VII, were treated with lithium aluminum hydride to give the corresponding cyclobutanols, which showed no tendency to undergo a ring-opening reaction. The cycloaddition of other ketenes and enamines has been observed; this work will be described in greater detail at a later date. RESEARCH LABORATORIES
TENNESSEE EASTMAN Co. I~VISIO OFNEASTMAN KODAKCo. TENN. KINGSPORT,
R. H. HASEK
a
Ia. X = CHBC b. X
I1
P
C2IIjOC
of Wiley and Smith.' 1-N-Pyrrolidinocyclohexene reacted with excess ketene to produce I1 (16%), m.p. 82.0-84.5' (C, 72.93; H, 7.21), 312 mp ( a 6800). The NMR spectra of the a-pyrones are in agreement with the proposed structures. These results are particularly interesting in view of the observations that diketene reacts with enamines to form y-pyrones.2.3 The reaction of 1-N-morpholinoisobutene with ketene follows a dBerent course. When either pentane or ether is the solvent, the product isolated is 2,2dimethyl-3-N-morpholinocyclobutanone (111) (75%), m.p. 41.142.2O (C, 65.73; H, 9.38; N, 7.71). Compound I11 rearranges readily on warming to 4 - methyl - 1 - N - morpholino - 1 - butene - 3 - one (IV), identical to a sample prepared from mor-
P@ c
r">
IV
0
V
J. C. MARTIN pholine and ethynyl isopropyl ketone.'
Received August 18, 1961
T h e Reaction of Enamines with Ketene Sir: We wish to report the results of our initial studies of the reaction of enamines with ketene. Treatment of 4-N-pyrrolidino-3-penten-2-onewith three molar equivalents of ketene in ether a t 0' and chromatography of the reaction mixture after evaporation of the solvent afforded 5-acetyl-4,Gdimethylcoumnlin (Ia) (52%), m.p. 72.8-73.4' (C, 65.10; H, 5.60), A23OE1296 mp ( E 6090). Similarly, ethyl 3-N-pyrrolidinocrotollnte reacted with ketene, to txoduce ethyl isodehvdroacstate (Ih) (35%), m.$. 18-20' (C, 61.42;" H, 6.13), hzJ:oii 247 mp ( e O )294 , mp (5240), which was idciiticnl to a sample prepsrcd by the proccdwe
Treatment of 111 with aqueous hydrogen peroxide produced a lactone (480j0), m.p. 103.4-104.3' (C, 60.41; H, 8.63; N, 7.10) which is assigned structure V. Hawk and Martin6have shown that ketene and dimethylketene react with enamines derived from aldehydes to yield the corresponding cyclobutanones. The rearrangement of these adducts has also been observed. DEPARTMENT OF CHEMISTRY MA&SlCHUSERs INSTITUTE OB TECHNOWGY CAMBRIDGE 39, MASS.
GLENNA. BERCiITOLD GEORGE It. HARVEY G . EDWIN WILSON,JR.
Received August 28, 1961
( 1 ) N. It. Smit.hand R. H. Wiley, 0 ~ ySyn., . 32,76 (1952). (2) S. Hiinig, E. Benzing, and K. Hiibner, Chem. Ber., 94, 486 (1961).
(3) B. B. Millward, J.Chem. SOL,26 (1960). (4) TI& sample was kindly supplied by Prof. N. J. Lconard of the-University of Illinois. (5) R. H. Hasek and J. C. Martin, J.Org. Chem., 26,4775 (1Nl). .
NOVEMBER
1061
4777
COMMUNICATIONS
Applications of Metal Nitrides in Organic Syntheses
Sir: To the extent that the '(ionic" nitrides contain the N-a ion,' they would be expected to react with a variety of organic compounds, showing strong basic, nucleophilic, and reducing properties. Early investigators, however, reported that no reaction occurs with alkyl halidesl5acid chlorides,' and other compound^.^.^.* Nitrides with acid anhydrides have been reported to give nitriles2 or imides.s We now report several additional reactions of ionic nitrides with organic substrates which indicate broad utility of the nitrides aa powerful basic catalysts, nucleophiles, and reducing agents. If, as seems likely, the pK,, of the NH-* ion is larger than that of the OH- ion, then the N-a ion may well be the strongest and most convenient base available under ordinary conditions and may be capable of abstracting a proton even from a saturated hydrocarbon. The nitrides are, however, almost completely insoluble in nonreacting solvents, so that the nitride ion is not readily available for reaction: this fact may account for the failure of previous investigators to observe appreciable reactivity. We have found that diglyme and certain other ethereal solvents strongly enhance the reactivity of the nitrides, possibly by partial solution of the nitrides. Fluorene, which would not react with lithium nitride in ethyl.et?her or without a solvent a t the melting point of fluorene, reacted on refluxing overnight in diglyme to produce a deep red solution; carbonation gave 31% of 9-fluorenecarboxylic acid, which on purification melted a t 205-207°,6 neut. equiv. 217 (Calcd. 210). Similarly 5 g. phenylacetylene and 1 g. lithium nitride, which would not react in mixed benzene-ether, reacted when dioxane was added; the mixture was refluxed (77') for 8 hr. and carbonated to give 5.5 g. (72%) impure phenylpropiolic acid, m.p. 120-13801 which on recrystallization from carbon tetrachloride melted a t 136-139O.' Acetaldehyde and propionaldehyde reacted very vigorously a t 15-20° with small amounts of lithium nitride to give mostly resinous products and small amounts of aldols and related products; 8 g. crotonaldehyde was obtained from 100 g. acetaldehyde in this manner. Isobutyraldehyde (100 g.) con-
taining 3 g. of lithium nitride was kept a t room temperature 3 days, then filtered and distilled to give 17.7 g. of the monoisobutyrate of 2,2,4trimethyl-1,3-pentanediol1 b.p. 103-104 (2 mm.), n? 1.4392.8 The Tischenko reaction was observed with benzaldehyde; 53 g. benzaldehyde with 3 g. lithium nitride reacted rapidly at 80' in carbon tetrachloride to give 35 g. benzyl benzoate and small amounts of benzoic acid. Fourteen grams of benzaldehyde was recovered unchanged. Although acid chlorides had been reported not to react with lithium nitride, benzoyl chloride and lithium nitride reacted in diglyme to give tribenzamide in good yield, m.p. 208-210° ( l k g m.p. 207-208'). Acetyl chloride also reacts with the nitride but thus far has given only diacetamide and none of the tertiary amide. With difunctional acid chlorides it should be possible to prepare imide polymers or related molecules containing tertiary amide linkages, which should show excellent thermal stability. Terephthaloyl chloride with either lithium or magnesium nitride gave an insoluble polymer decomposing a t 190-200°; the infrared spectrum of this material is quite similar to that of a vaguely related tertiary amide model compound, N,N'-phthaloyldiphthalimide. Various other reactions have been observed but have not yet been studied appreciably. These include reactions of lithium nitride with nitrobenzene, from which neutral products were obtained together with a gas which appeared to be nitrogen; with organotin halides, which presumably rcwted metathetically, since the lithium halide separated; with acetone, which apparently gave condensation products; and with benzyl chloride. Lithium nitride may be prepared directly from lithium and nitrogen under various reaction conditions; the nitride employed'O in this investigation was shown by analysis to be 90-95% pure. We believe the ionic nitrides constitute a class of reagents of wide applicability in synthetic organic chemistry. Investigations currently in progress extend to a variety of additions, displacements, and polymerizations, aa well aa to the use of dispersions and fused-salt solutions of the nitrides in organic reactions. GATES CEEMICAL LABOUTOEIES LOUISIANA STATE UNIVERSITY
BATONROUQE3, LA.
PAULE. EOENIQ JULESM. MORFUS EUGENE J. BUNCHARD
PEERY S. MASON
Received August 30,1961
(1) E. Masdupuy, Ann. Chim. [13], 2, 527 (1957); Chem. Abstr., 52, 2627c (1958). (2) F. Briegleb and A. Geuther, Ann., 123,237 (1862). ( 3 ) 0.Emmerling, Bet., 29, 1635 (1896). (4)hf. A. Smits, Rec. truu. cham., 12,202 (1893).
(5) G.Beck, 2.u w g . aZZp" C h . ,233, 155 (1937). (6) Beihtein's Hundbuch der Orguniachen Chemie, IX, 6!30-691. ( 7 ) T.W. Abbott, Org. Syntheses, COILVoL II, 516 (1943).
(8) F. J. Villani and F. F. Nord, J . Am. Chem. Sm., 68, 1675 (1946).
(9) C. Blacher, Ber., 28,435 (1895). (10) E. Masdupuy and F. Gallais, Z w g . Syntheses, 4,
3
I
(vnm\
4778
VOL. 26
COMMUNICATIONS
Partial Asymmetric Syntheses : The Diels-Alder Reaction'
the infrared spectrum was identical in all respects with that of an authentic sample. The rotation corresponds to 1.1% asymmetric synthesis. Repetition of the above experiment a t a temperature of Sir: 160' for six hours produced the glycol in 42% Asymmetric syntheses by the addition of various reagents to an olefinic moiety in both catalytic2 yield, [a]g- 1.7 f 0.2' (c 9.6, chloroform), 2.4% and noncatalytic3 reactions have previously been asymmetric synthesis. To a of aluminum chloride (o.025 mole) observed. We wish to report that the addition of and benzene (75 cc.) cooled to was rapidly butadiene to (--)-dimethyl fumarate results in added precooled solution of g, of a partial asymmetric synthesis, and that further(-)-dimethyl fumarate and 10 g. of butadiene the more when is used to dissolved in 50 cc. of benzene. The reaction mixture Alder r e a ~ t i o n the , ~ same product is obtained but was stirred at for one hour,6 and filtered; of opposite sign and configuration. the solvent was removed and the residue reduced by lithium aluminum hydride as described above. The glycol was isolated in 71% yield, [a12 40.4 f 0.5' (c 7.0, chloroform), 57% asymmetric synthesis. The above procedure was repeated a t 22' and for a twenty-minute reaction time to yield the glycol in 45% yield, I c Y ] ~ 19.25 f 0.42' (c 9.9, ehloroform), 26% asymmetric synthesis.' The (-) glycol I was shown to have the 1R:dR absolute configuration by relating (-)-4-cyclo~
*"''
+
+
CHZOH
9
0 - E I
%HzOH
2. I
(J-H
f R : d R (-)
"C-OH
II
(-)-Dimethyl fumarate (10 g., 0.025 mole), butadiene (10 g., 0.186 mole), and hydroquinone (0.1 g.) were dissolved in 20 cc. of benzene and heated in a sealed tube a t 67' for six hours. To avoid possible resolution the solvent was stripped from the reaction mixture, and the residue was reduced by lithium aluminum hydride to yield 4-cyclohexenetrans-1,2-dimethanol(I) i2.7 g., 72%), b.p. 100-101' a t 0.2 mm., - 0.8 i 0.2' (c 10.9, chloroform,s
IR:sR [el',"
lK:ZR
0
-70.4'
-160"
U
tR:eR
sR:iR -3.8"
-136"
[CY]?
(1) This work was supported by research grant CY 4065, National Institutes of Health, Public Health Service. ( 2 ) M. G. Vavon and B. Jakubowicz, Bull. SOC.Chim., 53, 111 (1953);D.Lipkin and T. D. Stewart, J. Am. Chem. SOC.,61, 3295 (1939);Y. Nakamura, J . Chem. SOC.Japan, 61, 1048 (1940);C. L. Arcus and D. G. Smyth, J . Chem. SOC.,35 (1955); A. Pedrazzoli, Helv. Chim. Acta, 40, 80 (1957): S. Akabori, S. Sakurai, Y . h u m , and Y . Fujii, . . biokhihiya, 22, 154(1957). (31 M.P. Balfe. J. Kenvon. and D. Y . Wadden, J . Chem. So;.,' 1367 (1954);F. J. fmpastato, L. Barash, and H. M. Walborsky, J . Am. Chem. SOC.,81, 1514 (1959);H. M. Walborsky, T. Sugita, M. Ohno, and Y. Inouye, J. Am. Chem. SOC.,82, 5255 (1960);H. C. Brown and G. Zweifel, J. Am. Chem. SOC.,83, 486 (1961). (4) P. Yates and P. Eaton, J . Am. Chem. Soc., 82, 4436 (1960);G.I. Fray and R. Robinson, J . Am. Chem. SOC.,83, 249 (1961). (5) The product was shown, by gas chromatography, not to be contaminated by (-)- menthol. The residue from the distillation had a zero rotation and showed the absence of carbonyl absorption in the infrared.
-
hexene - trans - 1,2 dimethanol to (-) - threo3,4-dimethyladipic acid whose absolute configuration has been established as SR:4R by Carnmalma8 The relationship is as follows:9 The role of the catalyst and the general scope (6) Longer reaction time caused extensive polymerization. (7) Bt our present state of knowledge we cannot assess the role that the aluminum chloride is playing in the asymmetric synthesis. Indeed, the observation that the glycol of opposite configuration is obtained may not be due to the catalyst a t all but to a temperature effect [see H. Pracejus, Ann., 624, 9 (1960)l.Another possibility is that the product from the thermal reaction is the thermodynamically controlled product whereas the AICIJ catalyzed reaction gives rise to the kinetically controlled product. (8) R. Carnmalm, Arkivf6r Khim., 15, 215 (1960). (9)All new compounds gave satisfactory elemental analyses. All rotations were taken in chloroform with the acid which exception of 4-cyclohexene-trans-1,2-dicarboxylic was taken in ethanol.
NOVEMBER
1961
4779
COMMUNICATIONS
conditions constitutes a new sulfide synthesis. The structure of 111 was readily confirmed by desulfurization with Raney nickel to 1,2-dimethylnaphINSTITUTE OF MOLECULAR BIO-PHYSICS thalene. H. M. WALBORSKY OF CHEMISTRY DEPARTMENT FLORIDA STATEUNIVERSITY L. BARASH In the solid state, the structure of Ia,b appears TALLAHASSEE, FLA. T. C. DAVIS~~to be I b as there is no band for the C=N a t 2225 cm.-' or for SH at 2555 em.-', as in the case Received September 5, 1961 for the corresponding 4-mercaptomethyl-l-naph(10) We gratefully acknowledge the technical assistance thonitrile, map. 97-98'. Anal. Calcd. for C12HONS : of Mr. D. M. Borgnaes. C, 72.31; H, 4.55; S, 16.09. Found: C, 72.19; H, (11) Ethyl Fellow, 1960-1961. 4.93;S, 15.91.Further, Ia,b forms a hydrochloride and acetyl derivative, and the latter derivative can be degraded with Raney nickel to yield the A Tautomeric Nitrile-Thiol anticipated product. Iminothiolactone System Under basic conditions, it was possible to demonstrate the acyclic tautomeric structure by other reactions. With alkaline hydrogen peroxide, I b Sir: oxidized smoothly to a disulfide (8673,m.p. 192.5In working with an intermediate, thought to be a Anal. Cabd. for C24HlBNJ32: C, 72.70; H, 1 9 3 ' . 2-mercaptomethyl-l-naphthonitrile(Ia),l an interesting type of ring-chain tautomerism has been 4.07;S, 16.17.Found: C,72.62;H, 4.15;S, 16.35. observed. When (Ia,b)2 was reduced with sodium Also I b reacted with benzyl chloride in ethanolic borohydride-aluminum chloride (3: l), its tauto- sodium hydroxide to give the anticipated sulfide, meric character became apparent in that products 2 - benzylmercaptomethyl- 1 - naphthonitrile, m.p. corresponding to both chain (11, 17%) and ring 53.5-54.5'. Anal. Calcd. for CIDHISNS:C, 78.86;H, (111, 48%) forms resulted. 2,5-Naphtho(t,2-c)- 5.23;S, 11.08.Found: C, 78.75;H, 5.32;S,11.23. Acknowledgment. This investigation was supported by a research contract with the U. s. Army CN Medical Research and Development Command dealing with antiradiation agents.
of this reaction, as it pertains to asymmetric synthesis, is currently being investigated.1O
""YS,
Ia
DEPARTMENT OF CHEMISTRY GARDNDR W. STACY WASHINGTON STATE UNIVERSITY ANTHONY J. PAPA PULLMAN, WASH. &VAS c. R A Y Received August 4, 1961
Ib
I1
111
thiophene (111) was separated from the reaction mixture after it had been poured into ice water and recrystallized from ethanol, m.p. 111-112'. Anal. Calcd. for CI?HloS: C, 77.38; H, 5.41;S, 17.21. Found: C, 77.50;H, 5.23;S,17.12.After the separation of 111,a tacky solid remained which was treated with concentrated hydrochloric acid; crystallization of the resulting material from methanol-ether gave 1-aminomethyl-2-mercaptomethylnaphthalene hydrochloride (11), m.p. 219-220.5' dec. Anal. Calcd. for C12HI4C1NS:C, 60.11; H, 5.89; N,5.84.Found: C, 59.91;H, 5.90;N, 5.87. It was found that when the proportion of sodium borohydride to aluminum chloride was increased from 3 : 1 to 4 : 1, I1 was obtained as the sole product (goyo,).The formation of I11 under the alternate (1) Compound Ia,b wm prepared from %(l-methylisothiuronium)-l-naphthonitrile bromide in 81% yield. The isothiouronium salt in turn was obtained from 2-bromomethyl-l-naphthonitrile (78%) while the latter substance wm prepared by the act.ion of N-bromosuccinimide on 2 methyl-l-naphthonitrile in a 71% yield. (2) The form Ib may also be represented as an amino tautomer, cf., A. J. Boulton and A. R. Katriteky, Tetrahedron, 12, 51 (1961).
Titanium-Catalyzed Isomerization and Olefin-Exchange Reactions of Alkylmagnesium Halides :A Novel Method for Preparation of the Grignard Reagent
Sir: George and Laddl found that when isopropylmagnesium bromide reacts with chlorosilanes in the presence of titanium tetrachloride the coupled product contains both isopropyl and n-propyl groups. ( CH3)&HMgBr
+ =Sic1
TiClr ___f
+
( CHS)~CHS~= CH&H&H2Si=
(1)
This discovery prompted us to undertake a further investigation of the effect of small amounts of titanium tetrachloride on the reactions of Grignard reagents. It has been found that the formation of rearranged products from isopropylmagnesium bromide is not peculiar to reactions of chlorosilanes, but is a general reaction, presumably due to isom(1) P. D. George and J. R. Ladd, Abstracts of Papers presented a t the 140th Meeting of the American Chemical Society, Chicago, Ill., September, 1961.
4780
VOL. 26
CO&iMWNICAl’IONS
TABLE I ISCIMERIZATION OF ~OPEOPYLM~GNESIIJMBEOXIDE Tic4 RMgBr
C&CHtCH*D TiCL
CItCII&HJ.)( CHskCHD
2.2 x lo-’ 2 . 2 x 104.4 x lo-’ 10-1
21.5’ 6.9 4.6 2.5
0.065 0.299 0.609
convenience as TiHC1, but probably a part of the solid Ziegler-type complex) and terminal addition to the olefin. The fact that more than oue molecule of rearranged product is formed for each mole of titanium tetrachloride requires that the alkyltitanium compound can exchange with the Grignard reagent.
3.1
(CH&CHMgBr
+Tim
--t
(C&)&HTiClr
-
erisation of the Grignard reagent brought about by the titanium tetrachloride. (CIt)&HMgBr
TiClc
C&CH&H*MgBr
(2)
Small amounts (10-l to moles per mole of Grignard reagent) of titauium tetrachloride were added at 0” to a solution of isopropyl magnesium bromide’ in diethyl ether and the mixture waa stirred for one hour and decomposed with deuterium oxide. Infrared analysis of the C3 hydrocarbons showed that both ldeuteropropane and Zdeuteropropane were formed. In Table I is shown the effect of varying tit.anium tetrachloride concentration on the ratjio of “rearranged” to (‘unrearranged” deuteropropanes. In all cases several moles of rearranged product were formed for each mole of titanium tetrachloride. No rearrangement could be detected when titanium tetrachloride was omitted, and n-propylmagnesium bromide did not rearrange even in the presence of titanium tetrachloride. It appears likely that the rearrangement proceeds by an elimination-addition mechanism aa in the isomerization of alkylboranes.’ The first step is the formation of an alkyltitanium compound, followed by elimination of the elements of Ti-H (shown for
(CE)2CHTiClr
+C&CH=CH,
+ TiCLH CH3CHtCHtTiCI, + RMgBr CHaCH=CHs
+ MgBrCl
+ TimH
CECHtCHzTiQ
CECH2CHtMgBr
+ RTiC4
(3) (4) (5) (6)
If an olefin other than that derived from the original alkyl group is present it should compete for the TiH in (5), displacing propylene m d producing the Grignard reagent derived from the added olefin. RCH==CH*
+ PrMgBr
TiCL
RCHSCHtMgBr
+ CH&H=CHt
(7)
This olefin exchange reaction provides a novel method for the synthesis of Grignard reagents, one which should be of preparative value, using aa it does olefhs and readily available alkyl halides. In Table I1 are shown the products and yields obtained when a number of olefins were refluxed for eighteen hours with a mixture of n-propylmagnesium bromide and titanium tetrachloride (mole ratio olefin: CaH7Br:TiCl, = 1:1 :0.03) and the product then allowed to undergo typical Grignard reactions. The expected products were obtained in 20-50% yields. These are preliminary results and undoubtedly do not represent the maximum attainable. Under the conditions employed, internal TABLE I1 double bonds do not react; thus ,i3-Aa-cyclohexenyl)R E A ~ I O NOFS GBIGNARD REAGENTSPREPARED BY OLEFIN propionic acid waa obtained from Q-vinylcycloEXCHANGE hexene : Olefin 1-Pentene 4-Methyl-1-pentene
Reagent
Product
1-Octene
I-Hexanol (45%) 6-Methyl-2heptanol (37%) Cyclohexanone 1-Hexylcyclohexanol (24%) Cot, CHsOH Methyl pelargonate
1-Octene
HC(OCA.)s
Allyl benzene
COI
Vinylcyclohexane
COl
CVinylcy clohexene
cot
4-Vin ylcyclohexene
Qa
pMethoxystyrene
CO2
1-Hexene
HCHO CHICHO
w%i
Pelargonaldehyde (19%) GPhenvlbutvric acid “(62%) fi-Cyclohexylpropionic acid (51%) 8-( A*-Cyclohexeny1)propionic acid (28%) 8-( A Wyclohexeny1)ethanol (40%) pMethoxyhydratropic acid (22y0)
(2) H. C. Brown and E. L. Subba b o , J. Org. Chem., 22, 1136 (1957); J. Am. Chem. Soc., 81, 6434 (1959).
n-PrMgBr:TiClr
co2 ~cHzcHzcozH The products from the aliphatic olefins were those corresponding to the terminal Grignard reagent, but pmethoxystyrene yielded pmethoxyhydratropic acid, rather than the expected pmethoxyhydrocinnamic acid.
Further investigation of the scope and limitations of the titanium-catalyzed olefin exchange reaction is in psogress. GENERAL ELECTRIC R.ESEARCH LAB. HERMAN L. FINKBEINER P.O. Box 1088 GLENND. COOPER SCRENECPADY, N. Y. Received September 5, 1961
NOVEMBER
1961
4781
COMft4UhTCA'ITONS
%phenyl phosphite resets with ICto give the The Reactions of 4-Bromocyclohexadienonienones tetrakyloxy phosphonium salt 111 as a white, with Esters of Phosphorous Acids intensely hygroscopic solid which reacts rapidly with water or alcohols to give 2,4,&tribromophenol Sir: and triphenyl phosphate. The reactions of quinones* and a-halocarbonyl Theoretical elementary analyses were obtained compounds2 with esters of phosphorous acid have for all new compounds. recently attracted appreciable interest. We now CHEMISTBYR"RCH DE" BERNARD MILLER' wish to report that certain features of these two A a r u c a ~ w aDrv~sxori ~ types of reactions have been combined to af!ford a h R I c A N CYANco. novel synthetic route to the biologically active aTAMFoRD, com. dialkyl aryl phosphates. Received September 8, 1961 Dialkyl phosphonates react with 2,4,&tribromo, Fmethyl-cyclohexadienone (Ia) in a manner similar to other reducing agents,s giving 2,6dibromo, 4(4) Present Address: Amencan Cyanamid Company, Agricultural Research Center, Princeton, N. J. methylphenol in high yield. stereospeci6c synthesis of 6-Methyl- tra~-5-cyclodecenone 1a.R = B r ; R ' = C & I I a . R = B r ; R ' = C&; R; f b. R P C1; R' = C& b. R C1* R' CE4; R c. R = R' = Br; R' = Et c. R = R' = Br d. R = Br; R' = C&; R' = E t Br
I11
Trialkyl phosphites, however, react with Ia-c in benzene solution to give the dialkyl aryl phosphates IIa-d in quantitative yields. The reactions a t room temperatures are apparently instantaneous, as judged by the disappearance of the yellow cyclohexadienone colors. Evaporation of the solvent leaves IIa-d in essentially pure statea, as determined by comparisons of their vapor phase chromatograms with those of distilled products. The structure of IId was confirmed by comparison of its infrared spectrum with that of a sample prepared by condensation of dry sodium 2,6dibromo, 4-methyl phenoxide with 0,Odiethylphosphorochloridate. In contrast to the very rapid reactions of the 4hromocyclohexadienones, neither Pphenylamino, 2,6-dibromo1 Pmethylcyclohexsdienone nor 2,4,6trimethyl, 6-acetoxy cyclohexadienone would react with trimethyl phosphite even on prolonged refluxing in benzene.
Sir: We wish to report the preparation of &methyltrans-kyclodecenone (III) by a sequence of reactions which exemplihs an effective approach, of wide applicability, to the Synthesis of various polyfunctions1 medium-sized carbocyclic ring systems of interest.' The key reaction of the sequence is the facile fragmentation of the 1,3401 monomesylate 11.Concomitantly, the ten-membered ring
a OMes
I
OH
0
I1
I11
(1) The simplest syatema, the cis and hm-ti-cyclodecenonea and the corresponding alcohole have been s y t h e a i d and studied. See H. L. Goering and W. D. Closson, J . Am. Chem. Soc., 83,3511 (1961), and ref. 9. Seealso D. J. Cram and L. K. Gaston, J. Am. Chem.Soc., 82,6386 (1960). The more complex systems belong to a subgroup of the sesquiterpenes which continues to grow in number at a rapid rate. For leading references see the following: J. P. devilliers, J. Chem-. Soc., 2049 (1961); J. Hochmannova, V. Herout, and F. Som, Cdl. Czcclr. Chcm. Co"., 26,1826 (1961); M. Suchf, V. Herout and F. &rm, Cdl. C d . C?. Go"., 26, 1358 (1961)! M. SouEek, V. Herout, and F. Sorm, Cdl. Czed. Chem. Comm., 26, 803 (1961); D. H. (1) (a) F. Ramirez and S. Dershowitz, J. Am. Chcm. R. Barton, 0.C. Biickman, and P. de Mayo, J. C h . Soc., 2263 (1960); M. Suchy, V. BeneSov6, V. Herout, and F. Soc., 78, 5614 (1956). (b) F. Ramirez and S. Derahowitz, J . otg. C h . , 22, 857 (1957). (c) F. Ramires and S. Der- &rm, Ber, 93, 2449 (1960); M. Suchf, V. Herout, and showita, 22, 1282 (1957). (d) F. Ramirez, E. Chen, and S. F. &rm, Cdl. C d . Chcm. Co"., 24, 1512 (1959); 6. C. Dershowitz, J. Am. C h . Sa., 81, 4338 (1959). (e) B. A. Bhattacharyya, G. R. Kelkar, and A. 8.Rao,Chem. & I d . , Arbumv, N. A. Polezhaeva, and V. S. Viiogradova, 1069 (1959); V. Herout and F. &rm, Chem. & Znd., 1067 I m t . A M . Nauk S.S.S.R. OMel. Khim. Nauk., No. 7, (1959); S. Dev, T&ah&on, 9, 1 (1960); 8. h v , Chem. & 1219 (1960). Id.,8, 171 (1860); M. Holub, V. Herout, M. Horsk, and F (2) See B. A. Arbuzov, V. S. Vinogradova, and M. A. 24,3730 (1959). For a SUE Zvemva, Izvcsl. A M . Nauk S.S.S.R. otdcl Khim. Nauk., brm, CoU. C d .C h .Comm., 9 mary of the above, see F. Sorm, Pure and Applied Chm.,2, No. 10, 1772 (1960), for leading references. 533 (1961). (3) I(.Fries and G. Oehmke, Ann., 462,1(1928).
~
4782
COMMUNICATIONS
VOL.
26
is formed from its bicylic precursor and the trans proved to be identical with the oxime prepared double bond is introduced stereospecifically.2 from authentic 6-methy1cyc1odecanone,7 m.p. 82Treatment of 8P-benzoyIoxy-9~-methyl-A4-383' (found: C, 72.10; H, 11.42). The NMR specoctalone (I)3 with alkaline hydrogen peroxide trums of the unsaturated ketone confirmed the yielded 8P-benzoyloxy-9~-methyl-4,lO~-epoxy-3presence of the trisubstituted double bond decalone (IV), m.p. 158-158.5' (found: C, 71.98; -CH=CCH3(bands with an area ratio of ca. H, 6.66). Addition of hydrazine hydrate to a 3:l centered a t 8.31 and 5.05 t, respectively), the methanolic solution of IV containing a trace of trans nature of which was evident from the photoacetic acid converted4 IV to 80-benzoyloxy-9p- desmotic transitiong observed in the ultraviolet methyl-A3-10~-octalol (V), m.p. 116-117' (found: spectrumlo; in 2,2,3,3-tetrafl~oropropanol,~~ ,A, C, 75.18; H, 7.741, and subsequent hydrogenation 227 mp ( E 2750) and in methanol, Am&= 221 mp gave 8P-benzoyloxy-9P-methyl-lO~decalol(VI), ( E 3050). m.p. 119-120' (found: C, 75.31; H, 8.15). Without Acknowledgment. The financial support of a isolation of intermediate compounds VI was conFrederick Gardner Cottrell grant (Research Corverted to I11 by successively (a) saponifying, (b) poration) is gratefully acknowledged. treating the crude decalindiol with one equivalent DEPARTMENT OF CHEMISTRY P. S. WHARTON of methanesulfonyl chloride in pyridine, and (c) UNIVERSITY OF WISCONSIN warming the crude monomesylate I1 with a solu6, WIS. MADISON tion of potassium t-butoxide in t-butanol.5 The Received September 6, 1961 product was distilled a t 0.1 mm. from an oil bath at 100' and yielded a colorless liquid which solidi(7) V. Prelog and W. Kiing, Helv. Chim. Acta, 39, 1394 fied on cooling, m.p. 18-21'. Analysis by gas (1956). (8)The spectrum of a solution cu. 10% in carbon tetrachromatographys showed that a single sharp peak accounted for 9770 of the total area corresponding chloride containing tetramethylsilane was recorded on a Varian 60 mc. spectrometer. to effluent. The preparation of an oxime, m.p. (9) E. M. Kosower, W. D. Clowon, H. L. Goering, and 132-134' (found: C,72.92; H, 10.52; N, 7.80), con- J. C. Gross, J. Am. Chem. SOC.,83, 2013 (1961). (10) The ultraviolet spectra were recorded on a Cary firmed the presence of the ketonic function observed in the infrared spectrum, : : A 5.86 p . The spectrophotometer, Model 14M, using 1 mm. cells. (11) We are grateful to Dr. Kosower for supplying this formation of an unsaturated ten-membered ring spectroscopic-grade reagent. was demonstrated by hydrogenation with platinum oxide in ethanol-acetic acid: oximation of the product yielded an oxime, m.p. 70-75', in 25% yield, Organic Disulfides and Related Substances. which, after several crystallizations from hexane, V. The Eight-Membered Ring System Bis(o(2) For a review of fragmentation reactions see C. A. phenylene) TetrasulfidelS2 Grob, I.U.P. A.C. Kdkuld Symposium, Butterworths, London, 1959, p. 114. The fragmentation of 1,3-diol monotosylates is described by R. B. Clayton, H. B. Henbest, and M. Smith, J. Chem. SOC.,1982 (1957), and implies a stereospecific relation of product to reactant. The observed formation of I11 (trans double bond) implies the stereochemistry of I as drawn, in agreement with that reasonably assumed on the basis of reduction of the corresponding 3,Sdione with lithium aluminum hydride and sodium borohydride. See ref. 3 and C. B. C. Boyce and J. S. Whitehurst, J. Chem. SOC.,2680 (1960). It is interesting to coneider the possibility that fragmentation is of significance in the biogenesis of medium-sized ring sesquiterpenes. At least, this is an attractive speculation to add to two others which are widely held: that medium-sized ring sesquiterpenes (1) arise by direct cyclization of a farnesol-type precursor, and ( 2 ) are of significance in general sesquiterpene biogenesis. See L. Ruzicka, PTOC.Chem. SOC.,341 (1959) and Ezperientiu, 9, 357 (1953); J. B. Hendrickson, Tetrahedron, 7 , 82 (1959); D. H. R. Barton and P. de Mayo, Quart. RW.,' ii, 189 (1957). (3) F. Sondheimer and D. Elad, J. Am. Chem. SOC.,79, 5542 (1957). (4) This is a further examde of a new synthetic method deiclibed by P. S. Wharton and D. H. Bohlen, J. Org. Chem., 26, 3615 (1961). (5) The fragmentation of I1 to I11 could also be effected under solvolytic conditions in aqueous pyridine containing acetic acid. (6) A four-foot column of 0.5% polyethylene glycol (average m. wt. 4000) on 60-80 mesh glass beads was used in conjunction with an ionization detector.
sir: Bis(o-phenylene) tetrasullide (I) is of interest for several reasons: (a) Two forms are possible,
as-yJ s-s I
a rigid anti form (I*) with the benzene rings on opposite sides of the plane of the sulfur atoms and a syn form (I8,)with the rings on the same side; I., may exist in a conformation (Is2)with mutually perpendicular benzene rings. The stereoisomerism of I is interesting per se, but particularly in its probable connection with orbital dispositions about the sulfur atoms. (b) Both the isosterism of sulfur with carbon-carbon double bonds and the electron swarm about the sulfur atoms imply interesting properties. (c) Ring-ring interactions are possible in Is,. (1) Dibenzo[c,g] [1,2,5,6]tetrathiocin.
(2) Research supported by the Office of Ordnance Research, U. s. Army. Taken mainly from the Ph.D. dissertation of W.D.S., Vanderbilt University, 1960. Paper IV, L. Field, T. C. Owen, R. R. Crenshaw, and A. W. Bryan, J. Am. Chem. Soc., in press.
NOVEMBER
1961
COMMUNICATIOKS
4783
Iodine oxidation of o-benzenedithiol reportedly though 2-mercaptobenzothiazole gave a strong test yields a disulfide "C6H4S2,''3but the product "was at less than 1/300 the concentration. After 16 hr. insoluble in ordinary organic solvents" and evi- in refluxing benzene, the molecular weight was 700, dently was polymeric. At high dilution, we have showing extensive polymerization. obtained I and have established its identity All x-ray patterns are believed to represent only by molecular weights (cryoscopic, ebullioscopic, one substance. I showed no damage during x-ray x-ray), analysis, and absence of thiol. studies, which are being continued. o-Ben~enedithiol~ (12.50 g.) in 2.5 1. of benzene DEPARTMENT OF CHEMISTRY LAMAR FIELD was added (68 hr.) with stirring under nitrogen to VANDERBILT UNIVERSITY WILLIAM D. STEPHENS iodine (24.00 g.) in benzene (7.5 1.)-water (1 1.). TENN. ERNEST L. LIPPERT, JR. NASHVILLE, Evaporation5 of the mixture to 3 1. gave crop A Received July 28, 1961 (2.85 g., largely insoluble in molten camphor). Concentration of the washed (aqueous bisulfite, water) and dried fdtrate to600 ml. gave 2.10g. (17%) of crop B. Molecular weights of B were concentraThe Action of Phenyllithium on tion-dependent ; extrapolation to zero concentraBromomethyl- and tion gave 288 (Rast, camphor)6 and 340 (ebullioIodomethyltriphenylphosphonium Halides' scopic,' carbon disulfide) ; calcd., 280.4. When heated for 10 min. a t 180' or for 12 hr. near 146", I became insoluble in camphor; in contrast, it had Sir: a molecular weight (Rast) of 313-351 after 1-12 All known cases of triphenylphosphinealkylidene hr. a t 120' and of 311 after irradiation for 1 hr. formation using organometallic reagents involve I with ultraviolet light. The possibility of a rather specific temperature for polymerization seems proton abstraction from the [(CaH&P-C-H]X unlikely, since differential thermal analysiss showed I endotherms only a t ca. 216" (probably the m.p.) system.2 Such behavior was observed by us also in and a t 380". Crop B when recrystallized from ben- the case of chloromethyltriphenylphosphonium zene (73%), carbon disulfide (47%), and chloroform bromide, only triphenylphosphinechloromethylene (68%) gave I as small lemon yellow prisms of ill- being formed when .the phosphonium salt was defined m.p. (immersion a t 145" and heating a t treated with phenyllithium. We have investigated the reaction of bromoG'/min. gave m.p. ca. 215-230'). (Anal. Calcd. for C12H&: C, 51.39; H, 2.88; S, 45.74. Found: C, methyltriphenylphosphonium bromide4 and iodomethyltriphenylphosphonium iodide6 with phenyl51.58; H, 3.05; S, 45.90.) Evaporation of the filtrate from B to 75 ml. gave lithium in ether and subsequent Wittig reactions 1.21 g. of crop C, mol. wt. 513 (Rast). Molecular of the triphenylphosphinealkylidenes formed. I n weights of recrystallized A, B, and C by x-ray both cases a mixture of triphenylphosphinemethdiffraction were 280.1, assuming two molecules ylene and the respective triphenylphosphinehalomethylene was formed. Thus the action of one per unit cell of 586.4 A3. Another oxidation gave crop B (7% yield, m.p. molar equivalent of et,hereal phenyllithium on ca. 210-218'), mol. wt. 281.5 (x-ray); stability a t bromomethyltriphenylphosphonium bromide gave ca. 25" for 21 months is evidenced by absence an orange solution to which cyclohexanone was of apparent change in the x-ray pattern. Ebullio- added subsequently. The usual procedure2 for scopic molecular weights were 297 (chloroform, also carrying out such Wittig reactions was followed, carbon disulfide) , 342 (benzene) ; 207 ( b e n ~ e n e ) , ~ 192 (chloroform);9 the Rast mol. wt. was 323. No infrared absorption characteristic of a thiol was seen. Feigl's testlo was negative in dioxane, al(3) P. C. Guha and M. N. Chakladar, Quart. J . Ind. Chem. Soc., 2,318 (1925). (4) Prepared essentially according to W. R. H. Hurtley and S. Smiles, J . Chem. SOC.,1821 (1926). (5) All evaporations effected below 25". (6) In high concentration, I polymerized readily. As the last crystal melted, solid began appearing and did not melt at 40" higher. (7) Micro apparatus developed and kindly loaned by Prof. M. T. Bush. (8) Kindly performed by W. R. Adams, Redstone Division, Thiokol Chemical Corp. (9) Clark Microanalytical Laboratory, Urbana, Ill. (10) F. Feigl, Spot Tests. Vol. ZI. Organic Applicutiuns, Elsevier Publishing Co., New York, 1954, p. 164.
(1) Studies in Phosphinemethylene Chemistry, Part V. For Part IV see D. Seyferth and K. A. Brandle, J . Am. Chem. SOC.,83, 2055 (1961). ( 2 ) For a review see U. Schollkopf, Angm. Chem., 71,260 (19.59). (3) D. Seyferth, S. 0. Grim, and T. 0. Read, J. Am. Chem. SOC.,82, 1510 (1960): J . Am. Chem. SOC.,83, 1617 ( 1961). (4)Prepared by Dr. Erwin Hahn in these laboratories by the action of phosphorus tribromide in hot benzene on hydroxymethyltriphenylphosphonium bromide. An independent synthesis was reported recently by F. Ramirez et al., J. Am. Chem. SOC.,83,3539 (1961).
4784
COMMUNICA'L'IONS
VOL.
26
and the following producls were obtained in the yields given: benzene (59%) ; methylemryclohexane (18%); bromobenzene (40%); and bromomethylenecyclohexane (46%) .6 Yields were determined via gas chromatography, using unreacted cyclohexanone present in the reaction mixture (determined independently as the 2,Uinitrophenylhydrazone)' as a standard. The h t three products were isolated also using preparative scale gas chromatography, and their identity was conh e d by comparison with authentic samples. A similar procedure using iodomethyltriphenylphosphonium iodide gave benzene (28%); methylenecyclohexane (30%); iodobenzene (70%) ; and iodomethylenecyclohexane* (26%). Similar results were obtained using other carbonyl compounds. Thus, in contrast to the chloromethyltriphenylphmphonium salt, the bromomethyl- and iodomethyltriphenylphosphonium halides are attacked at the C-X bond as well as at the C-H bond by
phenyllithium. The data prcscnted show the C-X/C-H reaction ratio to be considerably greater for the iodomethylphosphonium halide. These results are in line with the known relative reactivity of carbon-halogen links toward organolithium reagents: C-Cl < C-Br < C-I.9 At present, this is the only Wittig-type procedure available as a general synthesis of vinylic bromides and iodides. Attempts are in progress to deflne conditions which will eliminate or a t least minimize attack at the C-Br and C-I links in the bromomethyl- and iodomethylphosphonium salts. Details of these and other related experiments will be given at a later date.
(5) Prepared by the reaction of methylene iodide and triphenylphosphinein warm benzene. (6) ny 1.5130. Anal. Calcd. for GHII13r:C, 48.02; H, 6.33. Found: C, 48.29;H, 6.24. (7) H. A. Iddles, A. W. Low, B. D. Itosen, and It. T. Hart, Anal. Chem., 11, 102 (1939).
Received September 15, 1961
Acknm2edgment. The authors are grateful to the
U. S. Army Research Office (Durham) for support of this work. DEPARTM~NT OF CI~EMISTTIY MASSACHIJSET~~ INSTITUTE OF
TECHNOLOQY CAMBRIDGE 30, MA EX^.
DIETMAR SEYFE~TH JAMES K. JIEEREN SAMUEL 0. GRIM
(8) n y 1.5623. A d . Calcd. for GHIII: C, 37.86; H, 4.99; I, 57.15. Found: C, 37.93; H, 5.20; I, 56.90. (9) It. G. Jones and FI. Gilman, Org. IZeactioras, 6, 342 (1951).
