Organometallics 1982, 1, 1723-1726
1723
and the complex Fe2(p-t-BupP)2Clz(PMe3)2 (1) appears to taining a metal-metal interaction.” The geometry of the our knowledge to be only the second d6 “tetrahedral dimer” tetrahedral, formally Co(II), end is very similar to one end of the iron triad to have been structurally characterized. of the Fe(I1) dimer (1). The PMe3 group on the other (ii) The stabilization of a mixed-valence complex: here the cobalt(1) atom lies in the C O ~ ( ~ - ~ - B U plane , P ) so ~ that it cobalt compound C O ~ ( ~ - ~ - B U ~ P ) (2) ~ Chas ~ (aPrelaM ~ ~ ) ~is nearly colinear (co(2)-co(l)-P(3) = 168.3 ( 1 ) O ) with the tively rare Co(I1)-Co(1) formal oxidation state due to the metal-metal bond. As expected (2) is paramagnetic (peff presence of one terminal chloride ligand on one cobalt. (iii) = 1.70 p B per dimer, by Evans’ method in solution), inCoordinative unsaturation: all three complexes are coordicating one unpaired electron per molecule. However, we dmatively unsaturated and none obeys the 18-electronrule. have not observed an EPR signal in benzene solution at The Ni(1) dimer [Ni(p-t-Bu2P)(PMe3)],(3) is a key memroom temperature or at -196 OC. ber of a class of unsaturated dinuclear phosphido-bridged As noted above [Ni(p-t-Bu2P)PMe312 (3) is a member complexes of Ni(1). Stable complexes of Ni(1) are rare, and of a relatively rare class of nickel(1) phosphido bridged so far only one other member of this class has been compounds for which structural data is available on only structurally characterized. (iv) Last, a key structural one.18 For 3 31P{1H)NMR datal8 are in accord with the feature in each case is the presence of the dinuclear M2X-ray crystal structure in which all the phosphorus and (p-t-BuzP), core in which the M2P2unit is planar and the nickel atoms are coplanar and the Me3P-Ni-Ni-PMe3 unit internuclear distances suggest the presence of metal-metal is linear, giving the molecule D2h symmetry. The Ni-Ni bonding. Since a number of bis(phosphid0)-bridgeddimers separation of 2.375 (3) A is in accord with a single methave a ”butterfly” arrangement of the p-PR2- units and a al-metal bond, giving each nickel atom an unsaturated ”bent” metal-metal bond,” we have investigated the 16-electron configuration. possibility that the planarity might be due mainly to Further studies on the chemistry of these complexes and electronic factors. However, preliminary results appear of bulky phosphido complexes in general are in progress. to discount this.’, Instead studies using space-filling CPK Acknowledgment. We thank the Dow Chemical models suggest that the planarity is due to steric reasons. Company, Midland, MI, and the Robert A. Welch FounThe two t-Bu groups on each phosphido phosphorus atom dation (R.A.J.) and the National Science Foundation occupy considerable space and permit relatively small (J.L.A.) for financial support. non-bonded H-H contacts between neighboring t-Bu Registry No. (Fe(fi-t-BuzP)C1(PMe3)Jz, 82808-28-4;Coz(~-tmethyl groups. Similar modeling studies of M2(p-Ph,P)2 82808-29-5; B U ~ P ) ~ C ~ ( P82918-17-0; M ~ ~ ) ~ ,(Ni(W-t-BuzP)(PMe3)),, and M2(p-Me2P),cores clearly show that steric effects are FeC1z(PMe3)z,55853-16-2; CoC1z(PMe3)z,53432-22-7; NiClZrelatively unimportant in determining their framework (PMe3)z,19232-05-4;Li-t-BuzP, 19966-86-0;Fe, 7439-89-6;Co, geometries. 7440-48-4;Ni, 7440-02-0. The use of excess Li-t-Bu,P in the synthesis of the iron complex F ~ , ( ~ - ~ - B u ~ P ) ~ C(1) ~ ~results ( P M in ~ ~consid), Supplementary Material Available: Tables of final fractional coordinates, bond distances and angles, anisotropic thermal erable decomposition, and we have been unable to isolate parameters, and structure factor amplitudes (32 pages). Ordering any pure products under these conditions. Also, 1 is information is given on any current masthead page. unstable in hydrocarbon solutions, rapidly producing paramagnetic decomposition products, and it can be isolated in only very low yield. This has so far precluded its (17) See, for example: Brown, D. B., “Mixed Valence Compounds”; D. Reidel Publishing Co: Boston, MA, 1980. full spectroscopic characterization. (18) Schafer, H. 2.Naturforsch., B: Anorg. Chem., Org. Chem. 1979, The solid-state structure, however, is of considerable 34B, 1358. See ref 4 for [Ni(PR3)(P(SiMe3)2]2. See also: Nobile, C. F.; interest (Figure 1). The terminal PMe3 and C1 ligands give Vasapollo, G.; Giannoccaro, P.; Sacco, A. Inorg. Chim. Acta 1981,48,261 each d6 iron(I1) atom a roughly tetrahedral geometry for [Ni(r-P-c-Hx2)(P-c-Hx,Ph)],. (excluding the Fe-Fe bond, ca. 2.8 A, in the geometry of each Fe atom). Very few mononuclear Fe(I1) complexes such as FeC1,2- and FeL42+(L = Ph3P0, (Me,N),PO) are known.1° Although the “tetrahedral” d6 dimer Ru2Cg2+l3 has been structurally characterized, this geometry is more Carbon Monoxide Activatlon by Organoactinides. common for d8 complexes such as [Ru(NO)(p-PPh,)Formyl Pathways in CO Homologation and (PPh3)I2l4and [Ir(CO)(p-PPh2)(PPh3)]2.15 The Fe-Fe Hydrogenation distance (ca. 2.8 A) represents a weak interaction since it is significantly longer than the single bond lengths required Dean A. Katahlra, Kenneth G. Moloy, and by the Fez(C0)6(PR2)2 dimers to satisfy the 18-electron Tobin J. Marks’ rule.16 By virtue of its asymmetric structure in which there is Department of Chemistry, Northwestern University one terminal chloride ligand, the cobalt(I1,I) complex Evanston, Illinois 6020 1 C O ~ ( ~ - ~ - B U ~ P ) ~is Ca particularly ~ ( P M ~ ~ interesting )~ exReceived August 12, 1982 ample of a dinuclear formally mixed-valence species con(11) See, for example: Ginsburg, R. E.; Berg, J. M.; Rothrock, R. K.; Collman, J. P.; Hodgson, K. 0.; Dahl,L. F. J. Am. Chem. SOC. 1979,101, 7218 and references therein. (12) Albright, T. A., private communication. Dalton Trans. 1975, (13) Raston, C. L.; White, A. H. J. Chem. SOC., 2410. (14) Reed, J.; Schultz, A. J.; Pierpont, C. G.; Eisenberg, R. Inorg. Chem. 1973,12, 2949. (15) Mason, R.; Sotofte, I.; Robinson, S. D.; Uttley, M. F. J. Organomet. Chem. 1972, 46, C61. (16) See: Burdett, J. K. J. Chem. SOC., Dalton Trans. 1977, 423 for references to other F e F e bonds. See also: S u m m e d e , R. H.; Hoffman, R. J. Am. Chem. SOC.1976, 98, 7240 for theoretical calculations.
0276-7333/82/2301-1723$01.25/0
Summary: The carbonylation of Th[(CH,),C,] ,(OR)H (R = CH[C(CH,),],) to yield an enediolate, {Th,OR),[ cis-OC(H)=C(H)O-] , or, in the presence [(CH,),C,] of H, the methoxide Th[(CH3),C,]2(OR)(OCH3) is argued on the basis of chemical and kinetic evidence to involve rate-limiting attack of a carbene-like v2-formyl (Th(v2OCH)) on a Th-H functionality to produce a ThOCH,Th +This contribution is dedicated to Professor Rowland Pettit, whose work continues to inspire us all.
0 1982 American Chemical Society
1724 Organometallics, Vol. 1 , No. 12, 1982
intermediate. The intermediate suffers subsequent carbonylation and rearrangement to yield the enediolate or hydrogenolysis to yield the methoxide.
Communications Scheme I
1. Hydride Insertion Th-H
We recently reported the rapid, reversible migratory insertion of CO into thorium-hydrogen bonds (A) to yield
lo\
k1
t; CO ,
Th-:C-H
k- I
ti
D
H
H
E 0
H
2. Ketene F o r m a t i o n
..Th-:C-H /O\
k1
C
mononuclear q2-formyls(B, eq l),followed by a slower (the rate a sensitive function of R) coupling reaction to yield cis-enediolates (C).l The unusual character of this efficient, regiospecific carbon-carbon double bond-forming process, apparent analogues of which also exist for other actinide2as well as early transiti~n-metal~ and main-group* compounds, has prompted a mechanistic investigation. We communicate here evidence that for R = CH[C(CH,),], the rate-limiting step in this transformation involves metal hydride attack on the q2-formyland that the course of the reaction can be easily diverted to effect clean CO hydrogenation to a methoxide. A priori, three mechanistic sequences seem most reasonable for enediolate formation (Scheme I) and can be tested against kineticlchemical information. A hydride insertion reaction (scenario 1) finds precedent in d- and f-element q2-acyl~ h e m i s t r yand, ~ ~ iwith ~ ~ ~use of the previously obtained kinetic and thermodynamic data for step a,lJ the sequence can be analyzed by standard, steady-state approximation method^.^>^^ Thus, as written, scenario 1 (1)(a) Fagan, P. J.; Moloy, K. G.; Marks, T. J. J. Am. Chem.SOC.1981, 103,6959-6962.(b) Katahira, D.A,; Moloy, K. G.; Marks, T. J. "Advances in Catalytic Chemistry 11", in press. (2)(a) Fagan, P. J.; Maatta, E. A.; Marks, T. J. ACS Symp. Ser. 1981, No. 152,53-78. (b) Manriquez, J. M.; Fagan, P. J.; Marks, T. J.; Day, C. S.; Day, V. W. J. Am. Chem. SOC.1978,101,7112-7114. (3)Wolczanski, P. J.; Bercaw, J. E. Acc. Chem. Res. 1980,13,121-127 and reference therein. (4)(a) Jiitzi, P.; Schroder, F.-W. J. Organomet. Chem. 1970,24,C43C44 and references therein. (b) Fischer. F. G.; Stoffers. 0. Justus Liebias Chem. 1933,500,253-270. (5)(a) Gell, K. I.; Posin, B.: Schwartz, J.; Williams, G. M. J. Am. Chem. SOC.1982, 104, 1846-1855. (b) Marsella, J. A.; Folting, K.; Huffman, J. C.; Caulton, K. G. Zbid. 1981,103,55965598.(c) Therlkel, R. S.; Bercaw, J. E. Ibid. 1981, 103, 2650-2659. (d) Marsella, J. A,; Caulton, K. G. Ibid. 1980,102,1747-1748. (e) Fachinetti, G.;Floriani, C.; Roselli, A.; Pucci, S. J. Chem. SOC.,Chem. Commun.1978,269-270. (0Floriani, C., private communication. (g) Erker, G.; Kropp, K.; Kriiger, C.; Chiang, A.-P. Chem. Ber. 1982,115, 2447-2460. (6)(a) Maatta, E. A.; Marks, T. J. J. Am. Chem. SOC.1981, 103, 35763578. (b)Maatta, E.A.; Moloy, K. G.; Marks, T. J., manuscript in preparation. (7)We estimate that k-' 2 lo's-' from NMR line shapes' and from thermodynamic measurements' that K = kl/k_, = 3.2 (5)M-l at 30 "C. It will be seen that k2 = 0.3 M-I s-'. (8)(a) Moore, J. W.; Pearson, R. G. 'Kinetics and Mechanism", 3rd ed.; Wiley: New York, 1981;pp 313-317 and references therein. (b) Frei, K.; G h t h a r d , H. H. Helu. Chim. Acta 1967,50,1294-1304.
Th-H
?\
Th-:C-H
+
co
t CO
' k - l
k2
Th-o\C/H
(f)
IIC II0
F Th-0
Th-0
/I
/I
fast
t Th-H
c
Th-0
0 3. F o r m y l Dimerization Th-H
?\
2Th-:C
-H
/\
t CO , L Th-:C-H
. k-I
kZ
Th-0
>c=c
/O-Th H'
H
(i)
predicts rate [Th]2[CO]. There is ample precedent in organoactinide chemistry for the postulated rapidity of step c relative to k2 (especially for bulky hydrocarbyl g r o ~ p s ) ~ and ~ ~ Jfor ~ Jthe ~ carbene-like hydrogen atom (9) The steady-state and equilibrium assumptions reduce t o the same expressions. d[ThH] -=dt
(a)
k,k2[ThHl2[COI k'kz == -[ThHI2[CO] k-1 + kZ[ThH] k-1
where kz[ThH]/k-' < 10" under the present conditions ([ThH] = 0.05 M, Pco 5 760 torr).7
(b)
d[ThH] dt
-=-
klkz[ThH][C0]2 klkz -- [ThH][C0l2 k-1 + k2[CO] k-l
where k2[CO]
