"906
C)OhlMUNICATIONS TO THE
EDITOR
Vol. 81
yield, b.p. 74-83' (0.3 xnm.); n2O~1.5002; @D -15'. The irone mixture was shown by infrared and gas-liquid partition chromatography to consist of about 57% a-neoisoirone8 (4-phenylsemic a r b a ~ o n em.p. , ~ 174-175' from ethanol. (Found : C, 74.02; H , S.43.)), 10.5% a-irone; 8.2% pirone, S.4% lights and 16% high boiling isomer (4-phenylsemicarbazone, m.p. 144' from ethanol (Found: C, 74.03; H , 8.18)). The structure of the latter isomer is being elucidated.
added to a synthetic mixture containing the metal ions and acid is recovered almost quantitatively under the same conditions for distillation. When pyrophosphate in excess is added to the mixture after reaction and the solution is subjected to continuous extraction with ether, in 6 hours 6G-70y0 of the fumaric acid is recovered, this recovery being identical to that observed when R F H + + rather than RFCH3++ is the oxidant. The extract however contains very little alcohol (it should be noted that the half-ester would (8) T h e structure of a-neoisoirone was established by Y.R . Naves and P. Ardizio, Bull. SOL.Chbz. Fvancc, 1419 (1955). respond to the test for CHJOH). If the reaction ( 9 ) Pseudoirone from Hoffmann-La Roche, Inc.,s Kutley, N. J., mixture then is refluxed to complete the cornplexing gave upon cyclization with 85% phosphoric acid 16% a-irone, 10 5 % by pyrophosphate, the alcohol can be recovered 8-irone, 7 % lights, and G G . 5 % a-neoisoirone (4-phenylsemicarbazone. essentially quantitatively on distillation. m p. 114' from ethanol). Mixed melting point with 4-phenylsemicarbazone of a-neoisoirone from cyclobutirone, 174'. Both derivatives From these and related observations, we conclude show identical infrared spectra in K B r pellets. that ester hydrolysis accompanies electron transfer THE GIVAUDAN CORPORATION (to account for the appearance of acid and the DELBWANNA, KEIV JERSEY H. E. ESCHINAZI extractability of fumaric acid) and that the CH30H RECEIVED APRIL 1, 1959 is left cobrdinated to Cr(II1) (to account for the holdback of alcohol by the solution). Blank SUBSTITUTION COUPLED TO ELECTRON experiments prove that CH,OH in solution does TRANSFER' not become associated with chromium when Cr++Sir: aq. reacts with RFH++. Thus we conclude further Rate comparisons2 made for a series of pentam- that transfer of alcohol to chromium must be direct, minecobaltic complexes reacting with Cr++ aq. and therefore that C r f + attacks the ester end of suggest that when appropriate bridging groups are the fumarato ligand. associated with Co(III), electron transfer takes GEORGEHERBERTJOSES LAD. R. T. &I. FRASER place by remote attack of Cr++, the electron flowing USIVERSITYOF CHICAGO D. I(.SEBERA ILL H. TAUBE through the bond system of the bridging group. CHICAGO, RECEIVED FEBRUARY 27, 1959 The results reported now seem to constitute definite proof for this conclusion.
i"
DIFLUOROAMINE'
Sir:
During the reaction of nitrogen trifluoride with arsenic a t elevated temperatures (250-300') to produce tetrafluorohydrazine2 small amounts of (to be subsequently represented as RFCH3++) difluoroamine were produced. reacts with Cr++aq. a t almost the same rate as The identity of difluoroamine has been estabdoes the corresponding complex with fumaric acid lished. Its molecular weight by vapor density in place of the half-ester. In both systems, good measurements is 52 (calculated 53). The mass stoichiometry for the oxidation-reduction process spectrum of difluoroamine given in Table I was is observed, one Co++aq. (and 5NH4+)being formed obtained on a Consolidated Electrodynamics for each Cr++aq. which reacts. Model 620 Mass Spectrometer and is consistent Now it should be noted that remote attack by with the formula HNF2. Cr++aq. on the half-ester will leave both Cr(II1) TABLEI and CH3 attached to the remote carboxyl. The resulting structure would not be stable, so t h a t ester FRAGMEXTATIOW PATTERN OF HI\-F2 hydrolysis may be expected to ensue on electron ni/e Ion Pattern, % transfer. We cite evidence that this reaction does 53 HSFz 100 indeed take place. When Cr++aq. in equivalent 52 SFL X i amount is added to 0.01 dl RFCH3++ in the 84 HNF+ 99 5 presence of 0.1 A f HC104, and the solution imme33 SF+ 4i 4 diately after reaction is titrated to pH 4.5 with 20 IIF 4 0 standard alkali, we observe that acid has been 5 2 19 F' produced equivalent in amount to the complex 15 SH X i which has reacted. When such mixtures after 11 S+ 33 0 reaction are distilled a t 4', variable amounts of The infrared absorption spectrum of difluoroCHIOH are detected in the distillate, but comprising only 2 to 7y0 of the total CHzOH. In a amine consists of strong doublets a t 7.0, 7.8, blank experiment, similar in every respect except 10.2 and a triplet a t 1 1 . 2 ~ . The n.m.r. proton specthat Cr+++aq. is used in place of Cr++aq., no trum consists of a triplet as would be expected CHlOH is found in the distillate. Free CH,OH from a proton spin-spin coupling with two equivalent fluorine nuclei. The center of the triplet is bCH3
+
+
+
(1) This work is supported by the Atomic Energy Commission under COntrdCt AT(l1-1)-378 I ? ) P K Sehera and €1 T a u h r t o I>P p i i h l i - h d (rlwrrihed in p i r t i n i iil I ( 1 , 37, ~ 124 ( I ' i i q i )
(1) Army Ordnance Contract DA-01-021-ORD-5135 (2) C R Colbura and A Kennedy. THIS J O U R N A L , 8 0 , 5004 (i%Rl
June 5, 1959
COMMUNICATIONS TO THE EDITOR
approximately 6 cycles on the high field side from benzene. The spin-spin splitting is about 24 cycles. The fluorine resonance spectrum consists of two broad bands arising from spin-spin coupling of the fluorine nuclei with the proton. The high field member of this doublet was measured a t 2100 cycles on the low field side of trifluoroacetonitrile. The vapor pressure of difluoroamine was measured and can be expressed by a Clausius-Clapeyron equation log P(mm.) = -1298/T
+ 8.072
2907
find a more rewarding synthesis we turned to the acylation of the magnesium salt of ethyl hydrogen tnethylmalonate and have been able to develop this method into a very convenient one-step synthesis of a-substituted p-ketoesters. By treatment of ethyl hydrogen methylmalonate2 (I, R = CH3) with two equivalents of magnesium ethoxide, then removal of the ethanol formed,3 or, better, with two equivalents of isopropylmagnesium bromide in tetrahydrofuran, the half-ester was converted to the tetrahydrofuran-soluble magnesium chelate (11, R = CH3).4 Addition of
The boiling point of difluoroamine is -23' and the heat of vaporization is estimated to be 5940 cal./mole. The melting point is approximately - 131' and its critical temperature measured by the Cagniard de la Tour tube method is 130'. From an extrapolation of the vapor pressure data a k, critical pressure of 93 atmospheres was calculated. A material thought to be difluoroamine has been !I1 ,COaC&L YOzC2H6 reported previously by Ruff and Staub3 and S t a ~ b . ~ They found it to be a product of the electrolysis R-CH R-CH-C-R' II of ammonium bifluoride and reported a boiling 'COZH 0 point "ostensibly about an approximate I IV b.p. --65O, and m.p. -125'.$ These figures are ' led to the somewhat different from those reported herein, acetyl chloride to this solution a t 0 b.p. -23' and m.p. -131'. We have failed in rapid evolution of carbon dioxide and the formamany attempts to obtain difluoroamine by the tion of the new magnesium chelate (111, R = CHI, electrolysis of ammonium bifluoride. Burg5 has R' = CHI). Removal of most of the solvent, pointed out the uncertainty of the identification and hydrolysis of the residue with dilute aqueous ammonium chloride afforded a 62% yield of ethyl of difluoroamine by Ruff. Caution should be used in working with difluoro- a-methylacetoacetate (IV, R = CH3, R' = CH3), amine. We have had occasional explosions with b.p. 70' (13 mm.).5 Similarly we have prepared both solid and liquid difluoroamine. Difluoro- ethyl 3-keto-2-methylcaproate (IV, R = CH3, amine has been stored a t room temperature in R' = C3H7), b.p. 86-89'(10 mm.),6 in 69% yield glass vessels for long periods of time with no ap- from butyryl chloride; ethyl a-methylbenzoylacetate (IV, R = CH3, R' = CeHb), b.p. 90-94' parent decomposition, The authors wish to express their appreciation to (0.2 mm.),' in 52% yield from benzoyl chloride; Dr. Keith S. McCallum for the n.m.r. and infrared and diethyl 3-keto-2-methyladipate (IV, R = CHI, R' = CHzCH2COzCzHb), b.p. 80-82°(0.08 mm.) spectra of difli-oroamine. C, 57.45; H, 7.66), in 71% yield from p-carbeth(3) 0 . Ruff and L. Staub, 2. anorg. allgem. Chem., 198, 32-38 oxypropionyl chloride. Application of this method (1831). to the formation of the desired p-ketoester (IV, (4) Lisbeth Staub, Thesis, Breslau, 1932. (5) A. B. Burg in Simons, "Fluorine Chemistry," Vol. I, Academic R = CHa, R' = o - C H ~ O C ~ H ~ C H ~ Cb.p. H~). Press, N e w York, N. Y.,1950, p. 88. 117'(0.1 mm.)(C, 68.04; H, 7.53) led to a 63% KOHMA N D HAASCOMPANY yield from o-methoxycinnamic acid without isoREDSTONE ARSENAL AL KENNEDY lation of the acid chloride. RESEARCHDIVISION CHARLES B. COLBURN When ethyl hydrogen phenylmalonate* was emALABAMA HUNTSVILLE, ployed, acetylation of the resulting magnesium RECEIVED MARCH24, 1959 chelate (11, R = CeH5) afforded a 60% yield of ethyl a-phenylacetoacetate (IV, R = C6H5, R' = CsHs), b.p. 128-134'(8 m n ~ . ) Diethyl ~ 3A NEW SYNTHETIC METHOD FOR THE PREPARATION OF a-SUBSTITUTED 8-KETOESTERS keto-2-phenyladipate (TV, R = CaH5, R' = Sir : CH2CHZC0zC2H5), b.p. 137-138'(0.25 mm.) (C, (2) J. R. Roland and S. M. McElvain, TKISJOURNAL,69, 132 During the course of a synthetic program directed toward polycyclic terpenes, we had need of (1937). (3) The procedure reported for the preparation of diethyl ethoxysubstantial quantities of the P-ketoester, ethyl magnesiomalonate was used, J. A. Price and D. S . Tarbell, Org. S y n . , 5-(o-methoxyphenyl)-3-keto-2-methylvalerate(IV, 37, 20 (1957). (4) Cf. a similar magnesium chelate prepared by carboxylation R = CH3, R' = o - C H ~ O C ~ H ~ C H Z C HOne Z). approach (methylation of the pyrrolidine enamine of nitroalkanes with methylmagnesium carbonate, R. M. Stiles and H. Finkbeiner. THISJOURNAL, 81, 505 (1959). of the unsubstituted 8-ketoester, obtained via ( 5 ) K. Folkers and H. Adkins, ibid., 53, 1416 (1931). partial hydrolysis' of the corresponding malonate) (8) C. Glacet. A n n . chim., 1121 2, 293 (1947). (7) J. B. Dorsch and S. M. McElvain, THIS JOURNAL, 54, 2960 led to the desired product in only 31% yield from o-methoxyhydrocinnamic acid. In an effort to (1930).
1 :t
(1) B. R. Baker, R . E. Schaub and J. H. Williams, J . Org. Chcm., 17, 116 (1952).
1
( 8 ) E. J. Corey, ibid., 74, 5903 (1952). (9) R. H. Kimball, J. D . Jefferson and A. B. Pike, Org. S y n , Coll.
Vol. 11, 284 (1955).
