1138
/
Journal of the American Chemical Society
non-nearest-neighbor Interactions in ethane gives rise to a very small barrier. However, when these interactions are neglected for the M2LBseries, there is essentially no change in the magnitude of the barrier. This supports our contention that the rotational barrier in these dimers is due to the tilting or hybridization of the orbitals at the metal itself. (9) We are grateful to the National Science Foundation for the support of this work through Research Grant CHE-7606099.
Thomas A. Albright Department of Chemistry, University of Houston Houston, Texas 77004 Roald Hoffmann* Department of Chemistry, Cornell University Ithaca, New York 14853 Received May 25, 1978
Thermochemistry of Some Metal-to-Metal Triple Bonds
100.24
/
November 22, 1978
Table I. Standard Enthalpy of Formationa of Dimethylamido Compounds of Molybdenum and Tungsten
AH298Su~ 72.4
A.Hro(c)
M W Me2)4 Mo2(NMe2)6 W ( N Me2)6 WANMe216 (I
59.0 17.2 178.9 19.2
111
89.1 113.3
AHro(g) 131.4 128.2 268.0
132.5
I n kJ mol-'.
Table 11. E(M-NMe2) and Corresponding D ( M 2 M ) " for Various Formal Oxidation Numbers of the Metal Atom formal oxidation number 7
4
5
6
E( Mo-N Me*)
288
D( M
D( W-N M q )
200 33 I
D(W2W)
340
255b 396 295 558
223 592 258 775
I90 788 222h 995
O ~ M O )
a I n kJ mol-'. Experimental value. Sir: Although multiple bonds between transition metal atoms D(M-M) 6B(M-NMe2) = A H D have now become very well and compounds con(1) taining them are quite well characterized structurally and to evaluate D(M-M), where AHDrepresents the sum of all D spectroscopically, as well as chemically, there have been no values, a quantity obtainable from the AHfO values. The amthermochemical data bearing on the strengths of these prebiguity arises because there i s 2 0 unequivocal way to decide sumably very strong bonds. There have, indeed, been coarse what values are to be used for D(M-NMe2). Those obtained bond energy estimates ranging from too high,5 to plausibly experimentally in the mononuclear molecules, where formal ir~termediate,~,' to too low,8 but no heats of formation of peroxidation numbers are different, are not necessarily approtinent compounds have heretofore been reported. Some heats priate. It is well established for other_sorts of M-X bonds (X of formation have now been measured and are reported here. = C, F, CI, Br, I, for example) that D(M-X) varies with the From these it is, in principle, possible to evaluate the energies oxidation number of M. From the known correlakons for these of the multiple metal-to-metal ( M L M ) bonds,9 but, in other sorts of bonds, and using the two measured D(M-NMe2) practice, there are ambiguities. These too will be considered values, we can estimate the dependence of D(M-NMe2) values and the question of what may be credible M L M bond eneron oxidation number." gies, and with what ranges of uncertainty, will be discussed. W e thus finally arrive at the figures in Table 11, where we The measurements themselves and other technical points will present b(M-NMe2) values for oxidation numbers 3-6 and be the subject of specialized reports to appear elsewhere. the b ( M L M ) values which result when each of these is emAll species containing M L M bonds with n 1 3 are of at ployed in eq 1. Even though b(M-NMe2) values vary only least the complexity represented by the general formula moderately with oxidation number, the factor of 6 in eq 1 [X,M"MX,]*J. In such a species there are only two kinds causes the D(M-2-M) values to span a considerable range. of bond and, if we consider only neutral molecules (y = 0), the W e are not prepared positively to exclude any of the problem of determining D(M"M), the bond enthalpy of the D ( M 2 M ) values in Table 11, but we believe that some are M A M bond, is reduced to the following two steps: (1) meamore plausible than others. The metal atoms in M2(NMe2)6 suring the enthalpy of formation of X,MMX,(g); (2) estihave formal oxidation numbers of 3, but each metal atom acmating the value of b(M-X). The first of these steps is the less tually has a valence of 6. If it is assumed that a valence of 6 troublesome, though by no means trivial. The second poses implies the same b ( M - N M e 2 ) value in all cases, then the insidious difficulties. highest D ( M A M ) values are the best estimates, and this would A review of all available compounds showed that the most make these triple bonds among the strongest chemical bonds attractive candidates for study are the triply-bonded molecules known. The Mo-Mo and W-W quadruple bonds would be X3MMX3 where M = M o or W. Species of this type with X even stronger-perhaps the strongest bonds known. If, on the = alkyl-, alkoxy-, or dialkylamide are known. In order to have other hand, this equating of valence number with formal oxiany chance of success in step 2, it is required that there exist dation number overestimates the value of the latter to be used for each X3MMX3 a t least one MX, molecule, the heat of in Table 11, the true D ( M L M ) values are lower. Perhaps formation of which can also be measured. On this basis, our selection was reduced to Mo2(NMe& and M O ( N M ~for ~ ) ~ formal oxidation numbers as low as 4 are appropriate. Our tentative suggestion is to assign D ( M 2 M ) values in the range molybdenum and to W*(NMe& and W(NMe2)h for tungsten. 592 f 196 and 775 f 218 kJ mol-l for Mo and W, respecFor each of these four compounds the structure is known,I0-l4 tively. These values are reasonably concordant with the establishing it to be molecular in character, with equivalent plausibly intermediate estimates made earlier for some quaM-N bonds. I n addition, each one can be volatilized and all druple bonds, viz., 640 f 120 kJ mol-' for D ( M o - ~ M o )and are available in appropriate quantity and purity to allow ac560 f 120 kJ mol-' for D ( R e L R e ) . In short, M L M and curate thermochemical measurements. M L M bonds are very strong ones, though probably not the The thermochemical datal5 are presented in Table I. From strongest homonuclear bonds known (cf. D(N=N) = 946 kJ the AHfO data for the two mononuclear compounds and using mol-'). The experimental work is being extended to other standard16 values of AHfO [Mo, g] = 658.1, AHfO [W, g] = compounds with M L M bonds.I8 859.9, and AHfo[NMe2, g] = 123.4 kJ mol-I, one may straightforwardly deduce the following b(M-N) values ( f 5 References and Notes kJ mol-'): D(Mo-NMe2) = 255 kJ mol-' in Mo(NMe2)4 and (1) F. A. Cotton, Chem. SOC. Rev., 4, 27 (1975). D(W-NMe2) = 222 kJ mol-l in W(NMe2)6. (2) F. A. Cotton, Acc. Chem. Res., 11, 225 (1978). (3) M. H. Chishoim and F. A. Cotton, Acc. Chern. Res., 11, 356 (1978). We now employ the equation
+
0002-7863/78/1500-7738$01 .OO/O
0 1978 American Chemical Society
7739
Communications t o the Editor (4) W . C. Trogler and H. B. Gray, Acc. Chem. Res., 11, 232 (1978) (5) F. A. Cotton and C. 6. Harris, Inorg. Chem., 6, 924 (1967). (8) W. C. Troaler, C. D. Cowman, H. B. Gray, and F. A. Cotton, J. Am. Chem. Sm., 99,2993 (1977). (7) R. J. H. Clark and N. R. D'Urso, J. Am. Chem. SOC.,100, 3088 (1978). (8) G. L. Geoffroy, H. B. Gray, and G. S. Hammond, J. Am. Chem. SOC.,96,
a r
'
i
0.98B
5565 . _ _ _ I 119741 . _ .
(9) The notatio'nMLM is used for bonds of order 3-4 where purely pictorial symbols such as M Z M and M=M are cumbersome and/or easily misread and the intermediate bond order of 3.5, which occurs very frequently, cannot be given a purely pictorial representation. (10) M. H. Chisholm, F. A. Cotton, B. A. Frenz, W. W. Reichert, L. W. Shive, and 6. R . Stults, J. Am. Chem. SOC.,98, 4469 (1976). (11) M. H. Chisholm, F. A. Cotton, M. W. Extine, and B. R. Stults, J. Am. Chem. SOC.,96, 4477 (1976). (12) D. C. Bradley and M. H. Chisholm, J. Chem. SOC.A, 274 (1971). (13) D. C. Bradley, M. H. Chisholm, C. E. Heath, and M. B. Hursthouse, Chem. Commun., 1261 (1969); D. C. Bradley, M. H. Chisholm, and M. W. Extine, horg. Chem., 16, 1791 (1977). (14) M. H. Chisholm, F. A. Cotton, and M. W. Extine, Inorg. Chem., 17, 1329 (1978). (15) A l l of the thermochemical measurements were made at the University of Manchester and they will be reported fully in forthcoming papers. Briefly, the following methods were used to measure enthalpies from which enthalpies of formation could then be calculated. For W(NMe2)6the heat of hydrolysis in 0.1 M aqueous HCl was measured; the products were [NMe2H2]CIand H2WO4.For WdNMe2)Bboth combustion bomb calorimetry and an oxidative hydrolysis reaction employing acidic dichromate were used. The same type of oxidative hydrolysis reaction was also used for MO(NMe2)., and Mo2(NMe&. The enthalpies of sublimation of all compounds were measured using the vacuum-sublimation, drop-microcalorimeter technique previously described by F.A. Adedeji et al., J. Organomet. Chem., 97, 221 (1975). (16) NatI. Bur. Stand.(U.S.) Circ., 270-6(1973): D. D. Wagman, W. H. Evans, V. B. Parker, I. Halow, S. M. Bailey, and R. H. Schumm, NatI. Bur. Stand. (U.S.) Tech. Note, 270-6 (1973); E. R. Piante and A. 8.Sessoms, J. Res. Natl. Bur. Stand., Sect. A, 77, 237 (1973). (17) The arguments on which these estimates are based are not particularly controversial and will be explained in detail elsewhere. (18) Support for this work was provided by the Science Research Council (Manchester), the National Science Foundation (TAMU. Princeton), and The Petroleum Research Fund (Princeton), administered by the American Chemical Society.
130
r
50
Or
T
0965
.
LO
30
2o
F. A. Cotton* Department of Chemistry, Texas A& M Unicersity College Station, Texas 77843 Receiced July 21, 1978
120
b
J. A. Connor, G . Pilcher, H.A. Skinner Chemistry Department, VniGersity of Manchester Manchester M I 3 9PL, United Kingdom
M. H. Chisholm Department of Chemistry, Princeton University Princeton, New Jersey 08540
10 -0
lR1,
110
1
1
1 110
120
'orW1 10
1.30
'
-P
Figure 1. Activation energy (E,) for hydrogen abstraction from various alkanes at the marked ositions (a) by CH3.Io and (b) by CFy" radicals vs. delocalizabilities DY' (in units of p).
P
F H&H H
Perturbation Molecular Orbital Treatment of Free-Radical Hydrogen-AbstractionReactions Sir: The successful PMO treatment' of the reactivity and regioselectivity in dicyanomethyl radical additions to alkenes and of the reactivity in CF3. radical additions to alkenes using M I N D 0 / 3 data2-4 prompted us to extend this approach to hydrogen-abstraction reactions for which kinetic data are a ~ a i l a b l ee.g., , ~ the reactions of CF3. and CH3. radicals with alkanes.6 For the reaction of each radical we calculated Fukui's delocalizabilities D$R),7*8 according to occ
c .2
unocc
,
(-
P)
q-a This treatment implies that only the mutual interactions of the radical SOMO and the A 0 of the hydrogen to be abstracted in all occupied and unoccupied MO's of the alkane are considered. The resonance integral @ for each series of abstractions 0002-7863/78/1500-7739$01 .OO/O
0
T
H
1
O 0 13
O I
Dr
-0
Figure 2. Logarithm of relative rates of hydrogen abstractionI3 (In krei) from 1-fluorobutane b y chlorine atoms vs. delocalizability (DIR') (in units o f 8) as calculated for the shown conformations.
was taken as constant. This approximation may be justified on employing the same reagent (radical) and similar subs t r a t e ~The . ~ results of the calculations are listed in Table I. As can be seen from Figures 1a and 1b there is a fairly good
0 1978 American Chemical Society
