COMMUNICATIONS TO THE EDITOR
Oct. 20, 1959
The strange behavior of oxygen in these compounds has been emphasized recently by P. J. Wheatley4 and discussed in a review paper by K.J. Gillepsie and R. S. Nyholm.6 Surprisingly, in the course of an X-Ray study of [TiCI2(C5H6) ]?06we found definite evidence of colinearity of titanium and oxygen atoms. I n fact, the crystals of (I)> having* these unit cell constants: a = 7.47 if.; b = 9.86 A.; c = 12.58 A.; B = 127'5G', space group P21/c, have only two molecules per unit cell, so that the molecule must be centrosymmetrical. The colinearity of titanium and oxygen atoms is further confirmed by the Patterson and Fourier syntheses on the ac and bc planes. At the present stage of refinement of the structure (I? = 0.20), the fractional coordinates (x, y, z) of the heavy atoms have been assumed to be: Ti: 0.454, 0.132, 0.393; CII: 0.767, 0.132, 0.403; 0 1 : 0.178, 0.047, 0.187. It results that: 1, as above said, the Ti-0-Ti atoms are colinear. 2, All the carbon atoms of the cyclopentadiene ring are a t the same distance (2.35 f 0.05 A.) from the titanium atom. 3, Other relevant interatomic distances are: T i - 0 = 1.78 0.03 A.; Ti-C1 = 2.25 0.04 A.; Ti-CsH6' = 2.03 f 0.05 b. 4 Relevant bond angles are: 0-Ti-Cl = 104' i 2'; Cl-Ti-CI = 104" f 2'; C6H5-Ti-C17= 112' 2'; C5H6-Ti-07= 117' i 2". The T i - 0 distance is clearly shorter than the one expected for a single bond (1.78 b. instead of 1.92 k.),8 as was the case of the Ru-03 distance in the corresponding anion. Supposing an sp hybridization for the oxygen, and a d3shybridization for the titanium atoms, a partial double bond character of the T i - 0 bond may arise from donation of electrons from the py and pz filled oxygen orbitals to the d-y unfilled titanium orbitals. This fact may in turn stabilize the unusual sp hybridization a t the oxygen atom. We acknowledge the helpful suggestions of Prof. G. Natta, Prof. G. R. Levi and Dr. L. Porri, who also supplied us the sample.
*
*
*
(3) A. McL. Mathieson, D. P. Mellor and N . C. Stephenson, A d a Crysl., 5, 185 (1952). (4) P. J. Wheatley, in "Annual Review of Physical Chemistry," 8, 383 (1957). (5) R . J. Gillespie and R. S. Nyholm, Quart. Res., 11, 339 (1957). (6) Prepared by boiling in moist air a heptane solution of TiClr(CsHa). The yellow tabular crystals of [TiCl~(C1Hr)l~0, stable in air, may be recrystallized from heptane (private communication of L. Porri). (7) Referred to the center of the cyclopentadiene s-bonded ring. ( 8 ) Obtained by subtracting from the found Ti-C1 distance (2.25 A,) the difference between the covalent radii of chlorine and oxygen atoms (0.33 A,).
551 1
serious in the less polar solvents, e . g , AcOHS2 In this Communication we call attention to the large magnitude of such salt effects in the poorest ionizing media and their implications for reaction mechanism. The rate of ionization of p-methoxyneophyl p toluenesulfonate2b (I) may be followed in a variety of solvents by titration of generated toluenesulfonic acid, and this is accelerated by inclusion of, e.g., lithium perchlorate. With this salt, the pattern of salt effects in the 0-0.10 M range is the linear2 one of equation (1) or the more complex one of equation ( 2 ) , where k and K O are first order rate constants with and without added salt, respectively .
+
k = k0[1 b(LiC10,)j k = P[1 b(LiC10,) c(LiClO4)*/s]
+
+
(1) (2)
As summarized in Table I, such salt effects tend to become quite large in the less ionizing solvents, such as acetone, octanoic acid and ethyl acetate, and enormous in a solvent such as diethyl ether. In the latter medium, ionization rate is increased by a factor of 106 by 0;1 M lithium perchlorate. Because of the large salt effects on ionization illustrated in Table I, inclusion of a salt may drastically alter relative ionizing power of solvents. The comparison between acetic acid and diethyl ether is illustrated in Fig. 1. While rate of ioniza-
i 0 0
4 6 8 10 (LiCIOa),102 M. of k for I in Et90 and AcOH at 50.0'' ws. [LiClO,],
2
Fig. 1.-Plot DEPARTMENT OF INDUSTRIAL CHEMISTRY POLYTECHNIC INSTITUTE OF MILAN PAOLO CORRADINI MILAN(ITALY) GIUSEPPEALLEGRA tion of I in acetic acid a t 50" in the absence of RECEIVED JULY 22, 1959 lithium perchlorate exceeds t h a t in ether by a
LARGE SALT EFFECTS IN NON-POLAR SOLVENTS
Sir: Specific effects of added salts in accelerating rate of ionization of organic substrates such as alkyl toluenesulfonates tend to become quite (1) Research supported by the National Science Foundation.
factor of 2 X lo4, ether becomes a better ionizing medium than acetic acid at concentrations of lithium perchlorate above 0.036 M . For S N 2 displacement reactions with salts such as lithium chloride, the choice of a relatively poor ionizing solvent may not minimize b u t instead can promote (2) (a) S. Winstein, d at., THIS JOURNAL, 76, 2597 (1954); Chemistry ond Indusfry, 664 (1954): (b) S. Winstein and A. H. Fainberg, THIS JOURNAL,78, 2763 (1956).
5512
fin..
s..->.vr.
I ...7,,,XYC.
TABLEI EFFECTSOF LITHIUMPERCHLORATE O N IONIZATION OF I Solvent ACOH~~ 50% AcOH-AcnO hIe2SO HCOL-IIe? Ac.0 1 2 . 5 % AcOH-Dioxane Me2CO n-CiHiaCOOH EtOAc~ THF~ Et20
Temp., OC.
lO5ko sec-1
b
50.0 1 1 . 9 12.P 50.0 0.43 13.1b 75.0 18.2 0.oc 75.0 4.96 1.4c 75.0 3 . 4 1 47 I b 75.1 1.22 4G2$ 75.1 0.857 .t7.0c 7 5 . 0 0 434 4Glb,e 7 5 . 0 0 113 553b 75.0 0.0817 482' RO.0 0.000G~ 2.9s*** x 106
Av, fit
% of k 0.6 1.0 0.3
a , ii 2 ,3 0,5 I .6 3 .6 3.1 1.3
(LiC104) range: 0 0 4 . 0 6 M: 0-0.10 M ; 0-0.05 .If; 0-0.07 Af. e Equation 2; c = 1184. J Tetrahydro0.005 a t 75". Equation 2 ; c = 2.65 X lo6. furan.
, TO THE EDITOR
VOl. 81
results are given in Table I. These salts were also equilibrated a t higher temperatures with excess metal, quenched, arid analyzed. For PrC13, X,/hl values for (5 runs a t 978' averaged 2.34 f 0.03; for NdC13, G a t 9 X o , 2.00 i- 0.04; for N d h , 2 a t 970') 1.99 5 0.05; incomplete separation of metal may make the X111 ratios somewhat low. ,Although phase equilihrium studies with PrI3 are a t present incomplete, powder patterns of the product from reaction with excess metal a t >74O0 show a new phase and little PrI?to be present. The bronze product has an I Pr ratio less than 2 6 and does not appear to be the diiodide. TABLEI --Reduction
Q
competing ionization of the organic substrate and make it p r e d ~ r n i n a n t . ~ ( 3 ) S. Smith, J . Gall and D. Darwish, ~ ~ n p u b l i s h ework. d
System
7---Eutect~c-X,/M 7 , "C
limit
T, 'C
X/hI
2 50 & 0 . 0 4 W 4 i j 2 . 5 0 =k 0 . 0 ) f i 4 1 zt
N+YdCk
2.511 zk
f
3 2 01
+
Solid phase
.>
Pr, PrCId 0 0:i 811 =t2 SdClz 560 i 5 S d I ?
Pr-YrCl?
01 t i l 0
-
DEPARTMENT OF CHEMISTRY S. WISSTEIS Nd-SdIr 2 . 1 2 rz .o'? 492 i 2 ( 2 . 0 ) OF CALIFORXIA S. SMITH UXIVERSITY LOSANGELES 24, CALIF. D. DARWISH The dark green SdClz has been further identified RECEIVED SEPTEMBER 21, 1959 from powder pattern data as isomorphous with the RARE EARTH METAL-METAL HALIDE SYSTEMS. T H E PREPARATION OF NEODYMIUM(I1) HALIDES Sir :
I t has been suggested previously that the apparent solution of a number of metals in their molten halides is a result of the formation of a slightly stable, lower halide.' Although the suhhalide is frequently stable only in dilute solution. in some systems the amount of reduction is sufficient to exceed the normal salt-lower salt eutectic composition so that the latter can be obtained as a stable solid.* These metal-metal halide studies are presently being extended to the rare earth systems, where knowledge of the reduction characteristics under these conditions has been limited to the Ce-CeC13 system. Here the reduction limit recently has been reported to be about 9 mole % Ce (CeC12T3)in a solution in equilibrium with liquid Ce and solid CeC13a t 777°,3 in contrast to an earlier value of 327G.4 Although evidence for an oxidation state lower than three for neodymium in aqueous solution has been doubtful,jm6 and in liquid ammonia, inconcl~sive,~ reduction of the molten trichloride and triiodide by metal has been found to yield the corresponding neodymium (11) halide. \Xrith praseodymium, reduction only in solution is observed with the chloride, while a new phase is obtained with the iodide. The chlorides and iodides were prepared from the metalsS and their reactions with metal studied in tantalum containers both by cooling curves and by analysis of salt phases in equilibrium with excess metal. The essentials of the phase diagram (1) J D. Corbett, S. 17 Winhush and F. C . .4lbers, THISJ O U R N A L , 79, 3020 (1957). ( 2 ) J , D. Corhett and .\. Hershaft, ibid., 8 0 , 1530 (1058). (3) G. Mellors and S . Senderoff, J . P h y s . C h e m . , 63, 1110 (l9S9). (4) D. Cubicciotti, THISJ O C R N A L , 71, 4119 (1949). (5) C. Estee and G. Glocker, i b i d . , 7 0 , 1844 (1948).
( 0 ) H. Laitinen and E. Blodgett, ibid., 71, 2260 11949). ( 7 ) P. S. Gentile, Ph.D. Thesis, University of Texas, Austin. T e x n s , 1955. (8) a'e are indebted t o Drs. P. H . Spedding a n d A . H. Daane f o r the generous supply of pure metal and t h e benefit of their experience i n experimental techniques.
orthorhombic PbClz structure reported for SmClz and EuCla by Dijll and Klemm,g with lattice (SniC12: 4.49, constants of 4.31, 7.5s and 9.07 (.d, 9.97 -4.). Such a comparison has not yet been made for the dark purple NdIz. Phase diagram and X-ray data for the NdCla-NdClp system also show the existence of an intermediate phase near SdC12.:rin composition, melting prohably incongruently near 703'. Contribution S o . ;98. IYork was performed in the h i e s Laboratory of the 1:. S. .ltomic Energy Commission. r
- I
( 0 ) \T'. (1939)
Doll and \V, Klemm, Z . n l i o r e . ail,qe,ii. Cht't~z., 241, 2 1 6
INSTITUTE FOR ATOMICRESEARCH ~ S D LEOSARDF. DRUDISG DEPARTMEKT OF CHEMISTRY JOFIS D . CORBETT IOXYA S T A T E 1-TIYERSI.fY ANES, TOIT'A RECEIVED AUGVST7 , 1959 CARBENOID AND CATIONOID DECOMPOSITION OF DIAZO HYDROCARBONS DERIVED FROM TOSYLHYDRAZONES
S,ir: Tosylhydrazones (p-toluenesulfonylhydrazones) of aromatic aldehydes and ketones react with sodium in ethylene glycol to give aryldiazoalkanes ; tosy-lhydrazones of benzyl methyl ketone (equation l ) and cyclohexanone yield olefins and nitrogen. C8H,CH,C(Ci-I, I=.SSHO.SC;H~ + SaOCH2CH20H--+ C~II:CT-I-crrcF-rI+ s? i-ao.sc71f: -L ]
-
HOCH?CHI.OH i1 1
Carbon-skeleton rearrangements occur in decomposition of pinacolone and camphor tosylhydrazones' to give ".:~-dinieth?;l-'7-hutene and camphene.' .in investigation has now been made of reactions of arylsulfonylhydrazones with bases in protonic and aprotic solvents. The experimental condi(1) W. R . Bamford and T. S Stevens, J . Chew? SOL., 473,5 (1952). ( 2 ) See also R . Hirschmann, E . S . S n o d d y , J r , C. F. Iliskey and A'. L. Wender, Tms J O T , R K . A L , 7 6 , 1-013 f l 9 i 4 ) a n d G . H . Phillips, D. A. H. Tayloi and L. J . \\'yrnan, J . Chem. Soc., 1739 (1!454).
