diacetate (IIa), mp 202-206.5

---

May 20, 1952

269.3

COMMUNICATIONS TO THE EDITOR

oxygen bond. The strength of this interaction is a function of the aromatic substituent and of temperature, and ranges from a small amount of overlap of the orbitals to complete transference of the electron.

phenyltitanium triisopropylate were isolated, m.p. No decomposition occurred on storage 88-90'. under a nitrogen atmosphere a t 10'. However, the product decomposed rapidly when heated above its melting point. Analytical results are listed in Table I.

LABORATORY FOR INSULATION RESEARCH INSTITUTE OF TECHNOLOOY MASSACHUSETTS S. C. ABRAHAMS CAMBRIDGE 39, MASSACHUSETTS RECEIVED APRIL14, 1952

TABLE I ANALYSIS OF ORGANOTITANIUM DERIVATIVES Constituent

ISOLATION OF A COMPOUND CONTAINING T H E COVALENT TITANIUM-CARBON BOND

Sir:

e, 70

CsHsTi(OCsHr ) 8 Found Calcu(average) lated

Ti-Li complexa Found Calcu(average) lated

59.1 59.62 50.1 H, 70 9.2 8.63 8 . 9 Ti as TiOz, % 2 4 . 6 24.46 15.1 Halogen, % 0.23 (el) 15.2(Rr) 1-i, % Trace 2.5 CBHbTi( OC3H7)3.LiOCaH7.LiBr.(C*H&O.

49.95 8.13 15.12 15.11 ( R r ) 2.62

A study of the organic compounds of titanium has been carried out resulting in isolation of a compound containing the covalent titanium-carbon bond. Since titanium is a transition element with The reactions exhibited by phenyltitanium triisoits valence electrons divided between the third and fourth principal quantum groups, it cannot be propylate serve as a chemical proof of the presence expected to bear more than a superficial resem- of a titanium-carbon bond. An ether solution blance to other Group IV elements such as silicon, of the product gave a slowly developing organogermanium and tin which readily form stable co- metallic color test with Michler ketone; it oxivalent bonds with carbon. The consequent lack dized rapidly with oxygen to give a phenol derivaof stability and difficulty of formation of the tive; it reacted with water to give benzene, titantitanium-carbon bond is primarily responsible for ium hydrate and butanol; it reacted slowly with the repeated failures reported in the literature' diphenyl ketone to give triphenylcarbinol, and it failed to yield benzoic acid on treatment with since the first attempt by Cahours in 1861.2 Theoretical considerations which cannot be dis- "Dry Ice." These reactions are typical of an orcussed in this short space led us to the assumption ganometallic bond of low activity. The ether that relatively stable members of the class of com- solutions of the unisolated products obtained pounds of the type RnTi(CR')4-, might be pre- previously from phenylmagnesium bromide and the pared if n were 1 or, a t most, 2 and R were selected alkyl titanates exhibited similar reactions. The from the more electronegative organic groups. A titanium-carbon bond present in these compounds series of exploratory reactions between butyl was found to decompose spontaneously as follows CeH6:Tif +C6HI. + Tit3 titanate and various organomagnesium and organolithium reagents in 1: 1 to 4: 1 molar ratios sub- Free radical formation was shown by the ability stantiated this assumption and showed that the of the material to catalyze the polymerization of most stable carbon to metal bonds were formed with styrene. aromatic R groups. This work will be fully described in the JOURNAL The reaction between phenylmagnesium bromide a t a later date. and butyl titanate in a 1:1 molar ratio yielded a LEADCo., TITANIUM DIV. phenyltitanium derivative which was in too un- NATIONAL RESEARCH LABORATORIES DANIELF. HERMAN stable and reactive a state to be isolated from the SAYREVILLE, N. J. WALTERE;. NELSOS reaction mixture. When phenyllithium was used RECEIVED APRIL25, 1952 in place of the magnesium reagent, significant differences were observed leading toward the eventual isolation of phenyltitanium triisopropyl- CONFIGURATION AT C I ~OF 12-HYDROXYLATED ate in the following manner: The reaction of SAPOGENINS. REARRANGEMENT OF THE STEROID C/D RINGS pheriyllithium with isopropyl titanate in equimolar proportions led to the formation of a relatively Sir: insoluble crystalline, stable, lithium complex of Reduction of hecogenin with lithium aluminum phenyltitanium triisopropylate. Elementary an- hydride produced the Clz-epimeric diols: I, m.p. alysis and chemical properties indicated the follow- 216-220°, [ a ] " ~-32.4' (1.08, acetone). Found: ing formula : CCHETi (O C ~ H ~ ) ~ . L ~ O C ~(C2H~.L B ~ S H, 10.22; diacetate ( l a ) , m.p. 156-159', C,~75.40; [ a I z 3 -15' ~ (1.0, acetone). Found: C, 72.25; H&O. The lithium isopropylate portion of the molecule H, 9.34, and 11, m.p. 218.5-220.5', [ a I z 4-63.8' ~ reacted with titanium tetrachloride, as follows, to (1.05, acetone). Found: C, 74.79; H, 10.00; precipitate the lithium as the chloride, thus freeing diacetate (IIa), m.p. 202-206.5', [ a ] 2 3-65.1' ~ the organotitanium compound from the complex (1.0,acetone). Found: C, 71.88; H, 8.99. Molecular rotations of I, I1 and their respective diaceCaH&Ti( OCaH7)3.LiOCsH7.LiBr.(C*H&)zO+ */rTiCl, -----f tates compared with the Clz-epimeric cholanatesl CBHsTi(OCaH7)3 LiCl + LiBr + '/4Ti(OCaH,)4 + (GHE.)zO are in excellent agreement with the assigned con( 1 ) (a) B. Koechlin and T. Reichstein, Rels. Chim. A d a , 2S, 918 The lithium salts were filtered, and crystals of (1942): (b) E. Seebeck and T. Reichstein, i b i d . , 26, 536 (1943); (I

+

(1) H. Gilman and R. J. Jones, J . Org. Chcm., 10, 505 (1946). (2) M. A Cahours, A n n . chiwz, 131 63, 280 (1861).

T. F. Gallagher and W. P. Long, J . B i d Chcm., 162, 521 (1946); (d) D. H. R. Barton and W. Klyne, Chem b I n d . , 755 (1948).