[CONlRIHLTIOS
F R O M THE
DEPARTMEST O F CHEMISTRY, CORNELL UXIVERSITY]
Donor-Acceptor Bonding. VI. The Reactions of Trimethylamine, Dimethylamine, Monomethylamine, and Ammonia with Titanium Tetrachloride and of Trimethylamine with Titanium Trichloride' BY MORTONANTLER^
AND
A . W. LAUBENGAYER
RECEIVEDJUNE 11, 1955 T h e MesS-TiClr, MeaSH-TiCL, MeSHZ-TiCL and SH3-TiC14 systems have been studied. Addition reactions were found t o be complicated b y ammonolytic cleavage when G H links were present in the donor molecule and also by oxidationreduction reactions. The ability of the methylamines and S H 3t o reduce titanium(li-) decreases a s t h e number of methyl groups decrease, and in all cases increases as the temperature is raised. The addition compounds Me3L7.TiClaand (Meah')?, TiClj were isolated from t h e MeaN-Ticla adducts. T h e latter compound was prepared also in 100% yield by the direct reaction of M e 3 S and TiCI3. Both addition compounds are sublimable, crystalline solids of low thermal stability. The infrared absorption spectrum of (Me3S)2.TiC13is closely related t o t h a t of Meax. Its magnetic susceptibility is near the theoretical value calculated for one unpaired electron per titanium atom. The thermal decompositions of all gross adducts and the purified addition compounds were studied. The adduct of S H 3 and TiCI4, Irhich had the empirical composition (SH,),,TiCl, contained S H 4 C I . On heating t o 160' the presence of reduced titanium was detected, indicating t h a t the T i S C l reported as a product of the thermal decomposition of the system KHa-TiC14 may not b e a compound of tetravalent titanium. Possible reaction mechanisms and the structures of titanium halide molecular addition compounds are discussed,
Introduction Titanium tetrachloride has long been known to complex with electron pair donor molecules. However, only a few of its addition reactions have been critically studied, and this is particularly true of the systems involving nitrogen donors. I n contrast, the acceptor properties of TIC4 are almost unknown. The reactions of TiCl, with 3" have been investigated by numerous workers," and the formation of various amidochlorides, e.g., Ti(NH2)+21, has been shownJmto be characteristic of this system. Thermal decomposition in 'LYXUO of NH3Tic14 adducts yields a solid of the forniula TiNCl. The tendency of TiCl, to participate in oxidation-reduction reactions with certain organic systems has generally not been recognized, and it seemed desirable to consider this possibility with a series of related bases, the methylamines and "3. Aknmonolyticcleavage could be expected to be a complicating feature of the addition reactions with the donor molecules containing S-H links. Experimental Materials and Manipulation.-Eimer and Amend C . P . TiCl, was refluxed over mercury and distilled in an all glass system. lloisture was excluded by a P z O ~drying tube. Anh!rdrous TiC1, was prepared b!- reducing TiCL with hytlrogen3 (Calcd.: Ti, 31.05; C1, 68 Found: Ti, 32.1; CI, 68.0; reducing power, 99.27c, b! ation with standard Fe?(S04j3 t o the red end-point cf the iron thiocyanate complex). The hlatheson tank Me.iS used in the direct reac( I ) Presented hcforp the 1 ) i s i s i o n < > f Physical and Inurganic Chernist r y of t h e American Chemical ciety, Chicago, Ill., September, Iauliengayer and \Vm. C . Smith, 1%53 Paper \* o f this series, A \ T H r s JOCRNAI.. 76, 598.5 f I R 3 4 ) . 121 Abstracted from t h e thebis submitted t < > the Graduate School, Cornell U n i v e r s i t y , June I'J53, b y Morton .4ntler in partial fulfillment of the requirementi for t h e degree of Doctor of Philosophy. Di- a n d hexammines of TiCli have been reported by IV. C. Schumb a n d K . F. Sundstrom. THISJ O U R N A L , 5 6 , 596 (1933). (4) (a) H. Ruse, P o t r i l i t i z . . 16, 37 (1829). fbl H. Kose, ;bid., 2 4 , 1-11 (1832); i c ) J . P e t s t z , A u v Ciicrtr P h y ? . , [ 2 ] 44, 321 (1830); (d) A . Rosenheim a n d 0 . e m , 2 6 , 239 i1901); ( e ) \ I Biia a n d W 1Virhelai 36,4228 (1903); (1) A. Stahler. i b d . , 38, 2C,10 (190.5~; fs:) house, .I .Yoc. Cliurti l i i c l , 2 6 , 738 '1907); [ h i 0 Ruff a n d 9.50 (19081; ( i ! 1,. I > e n i i S ; i t h a n , Ph I > T h e s i s , A . 1'129. !i! A I3ragc.r. rk) C 11. Olson. U S Patent 2 . 4 13,778 (lC447) Patent 2,(iO6.81.5 (IK521: ( m i Cr !A- A . 1:wvIes anti I:. 11. Poll:ird .I. Clwti?. ,S,w , %>88 l l ! l + X ) . ~
tion with TiCla was dried and freed of impurities containing labile hydrogen bg treatment with freshly sublimed, ground Pros.; Matheson amines and S H 3used in all the reactions with Tic14 were dried by passage through porous B a 0 . 6 All manipulation of the solid adducts between the titanium halides and nitrogen bases was in a nitrogen atmosphere using a small, tight metal dry box. It was necessary to pay particular attention t o assure complete removal of moisture because of the extreme tendency of the adducts to absorb water, with subsequent decomposition. After loading, the box interior was flushed for a long time with oil pumped nitrogen, and then residual water was condensed on the cold surface of a glass bulb which could be filled with liquid nitrogen through connections t o the outside. I. The System Me3N-TiC1,$. Preparation of (MesK)a. TiC13.-Bis-trimethylamine titanium trichloride, (hlerK)2. TiCIii,was prepared by the direct reaction of TiCIJ with excess amine. Crystals of TiCI, were ground to a fine powder in the dry box and transferred t o the reaction vessel. The vessel was evacuated and M e 3 S condensed a t -78' onto the TiCIJ. The system was warmed t o ce. -5" and the suspension in liquid amine churned continuously with a glass-enclosed magnetic stirrer. After several hours, when reaction appeared complete, as evidenced by no further color change of the solids, excess amine was pumped completely from the system. The X-ray- powder diffraction pattern of the blue-green crystalline solid showed none of the lines characteristic of TiCI3. (MeaS)z,TiCl1can be recrystallized from liquid MesN. Analyses .-Samples were loaded into tared glass tubes in the dry box, sealed, and weighed. Analyses for titanium and chlorine were made gravimetrically on the same sample, the titanium by precipitation of hydrous titanic oxide from a dilute H S 0 3 solution followed by ignition to TiOn, and chlorine b y precipitation of AgCl from the filtrate. Carbon and hydrogen were determined by the conventional microcombustion technique. Analyses for nitrogen by the micro Dumas method were erratic. Similar difficulty was encountered with the adducts of the methylamines and TiCL.' .-lrzui. C d c d . for (Me,T),.TiCli: T i , 17.58; C1, 39.(14; C, 26.44; H , 6.66. Found: T i , 17.76; C1, 38.77; C, 26.51; H , 6.63. Attempts to titrate the Ti(II1) of (Me:iS)?.TiC1.I dissolved in water, dilute HC1, or dilute H2S04were made directly with standard Fez(S04)3,K2Cr2O7,and indirectly b y the titration with K2Cr20iof the Fe(I1) produced by t h e addition of excess Fe2(SOl)3. Low, highly erratic results were always obtained. Aqueous solutions of the dditicin compound were usually slightly cloudy. Volatility and Thermal Stability.-(lle3S )?.TiCI3can be slo~vlysublimed iiz m c u o a t moderate temperatures, c ' . ~ . , with a temperature differential from 20 t o 5'. At highrr .
L7. I n g r a m , 11,s. 'l'hesis. 'l'he U n i v e n i t y
