Exceptionally facile CO addition to a saturated ruthenium complex

Article abstract of DOI:10.1021/ja051296h

Synthesis and characterization of Cp*Ru[η³-HC(PPh₂NPh)₂], 1, reveals it to have a piano stool structure with the ligand bound to Ru(II) via two N and the unique, sp³ hybridized carbon. While the analogous (cymene) Ru[η³-HC(PPh₂NPh)2]+ does not react with CO, under the same conditions, 1 adds one CO rapidly (25 °C, 1 atm CO). Characterization, including an X-ray structure determination, shows that CO has displaced one chelate ligand nitrogen, which then hangs off the molecule, free of Ru. DFT calculations reveal a possible mechanism via a remarkably low energy (+9.3 kcal/mol) intermediate, pendant N, but with one phenyl on phosphorus stabilizing Ru via donation from a C(ipso)=C(ortho) bond. DFT calculations show that the electronic energy change for binding CO is over 20 kcal/mol less favorable for cymene than for C₅Me5- as ligand; the reactivity difference is thus thermodynamic in origin. Copyright

Full text of DOI:10.1021/ja051296h

Published on Web 06/01/2005
Exceptionally Facile CO Addition to a Saturated Ruthenium Complex
Christine Bibal,† Yegor D. Smurnyy,†,‡ Maren Pink,† and Kenneth G. Caulton*,†
Department of Chemistry and Molecular Structure Center, Indiana UniVersity, Bloomington, Indiana 47405, and
Chemistry Department, Moscow State UniVersity, Moscow 199992, Russia
Received March 1, 2005; E-mail: caulton@indiana.edu
The literature on bisphosphinimino methanide ligands shows
remarkably Variable bonding (“d”)1-5 of the nucleophilic ligand
carbon to metal (I), as well as the possibility that the resonance
structure in IIa permits delocalization of any electron density away
from an otherwise sp3 carbanion, to yield a bidentate ligand mode
via only the two nitrogens. The η2 bonding of this ligand type has
another advantage. Stalke has recently advanced the idea6 that the
charge-separated resonance structure III contributes significantly
to the ground state of phosphinimines, thus enhancing the nucleo-
philicity at nitrogen. Structure IIb (together with relief of four-
membered ring strain) thus confers additional stability to the η2
binding mode. This opens the question of whether an η3 h η2
equilibrium might make an 18-electron ground-state structure I a
functional precursor to a 16-electron isomer IIa, and thus the
bisphosphinimino methanide a noninnocent ligand useful for
kinetically facile applications. The present work answers this in
the affirmative, but proves the above reasoning to be wrong.
Figure 1. ORTEP drawings (50% vibrational ellipsoids). (A) (C5-
Me5)Ru[HC(PPh2NPh)2]: Ru-N1 2.229(2); Ru-N2 2.226(1); Ru-C1
2.273(2) Å; N1-Ru-N2 82.68(5); C1-Ru-N1 70.22(5); C1-Ru-
N2 71.25(5); P1-C1-P2 120.87(9)°. (B) (C5Me5)Ru(CO)[η2-HC(PPh2-
NPh)2]: Ru-C48 1.863(2); Ru-C1 2.211(2); Ru-N1 2.156(2) Å; C48-
Ru-N1 93.79(10); C48-Ru-C1 97.17(10); C1-Ru-N1 72.49(7); C8-
P1-C14 106.80(11); C20-P2-C26 107.31(11); C1-P1-N1 99.37(10);
C1-P2-N2 109.58(11); P2-C1-P1 119.14(14)°.
the Cp* and the pendant PPh2NPh substituents in a trans relationship.
This suggests a single specific stereochemistry for Ru/N2 bond
scission. The N2/Ru separation, 3.68 Å, is clearly nonbonding, but
the atypically large bend of the carbonyl ( Ru1-C48-O1 )
167.7(2)°) may be to minimize repulsion between the inwardly
directed N2 and the CO. The N2/C48 distance, 2.84 Å, is about
0.15 Å shorter than the sum of van der Waals radii, and thus may
reflect incipient attraction between nucleophilic N2 and the carbonyl
carbon.9
Reaction of (Cp*RuCl)4 with LiCH(PPh2NPh)2 in benzene at
23 °C gives immediate conversion to a product characterized
spectroscopically (mirror symmetry) and crystallographically as Cp*-
Ru[HC(PPh2NPh)] (Figure 1a). The bisphosphinimino methanide
is η3 bound with a pyramidal (sp3) carbon and a Ru-C distance
consistent with a single bond. This compound is the analogue7 of
cationic (cymene)Ru[HC(PPh2NPh)2]+, but it shows remarkably
different reactivity. While this cymene-containing cation fails to
react in THF with 1 atm CO over 24 h at 23 °C, and starting
material is recovered, under these conditions, Cp*Ru[HC(PPh2-
NPh)2] reacts completely (eq 1) and immediately to give a
monocarbonyl adduct (υCO ) 1906 cm-1 in C6D6). The first
structural evidence against a structure derived from II for the
monocarbonyl is that the 31P{1H} NMR spectrum is an AX pattern,
with a very small JPP ) 1.5 Hz. Since the addition of CO to an
18-electron center cannot occur without some elimination of a
bound ligand (e.g., f η3-Cp*), this 31P NMR evidence suggests
that it is not part of Cp* and not the unique carbon (II), but rather
one imine nitrogen arm that has been dissociated. A single crystal
structure determination of this CO adduct (Figure 1b) indeed shows
one pendant nitrogen, N2, with retention of the Ru-C1 bond in a
four-membered ring η2 binding mode8 for the bisphosphinimino
methanide moiety. Note that only one diastereomer of this ring is
formed (NMR evidence of unseparated reaction solution), that with
How can the CO attack on an 18-electron center be so facile?
Conventional mechanistic thinking suggests a pre-equilibrium
“isomerization” to create unsaturation. DFT(PBE) searches of
minimum energy10 structures for (C5Me5)Ru[CH(PPh2NPh)2] found
one which was identical to the crystal structure (i.e., Figure 1a),
judging by bond lengths and angles within the Ru[η3-CH(PPh2-
NPh)2] substructure. Starting geometries designed to mimic the η2-
N/N ligand (type II), as a 16-electron complex analogous to the
well-known11 CpRu(Me2NC2H4NMe2)+ and its relatives, gave
Figure 2a, 15.1 kcal/mol above the η3-isomer. This species is
analogous to Cp*Ru(amidinate).12 Starting geometries designed to
mimic 16-electron species Cp*Ru[η2-HC(PPh2NPh)2] with the
ligand bound by the sp3 carbon and only one nitrogen led to a
† Indiana University.
‡ Moscow State University.
9
8944
J. AM. CHEM. SOC. 2005, 127, 8944-8945
10.1021/ja051296h CCC: $30.25 © 2005 American Chemical Society

C O M M U N I C A T I O N S
To understand the lack of CO addition to (cymene)Ru[η3-
HC(PPh2NPh)2]+ versus the facile reaction with the uncharged Cp*
analogue, we have optimized the structures of the cationic cymene
reactant and product. Comparison of the Cp* and cymene species,
(ring)Ru[η3-HC(PPh2NPh)2], shows no dramatic geometric differ-
ences (i.e., flaws) in the cymene species, although metal-ligand
bond lengths are about 0.08 Å shorter than those in the cymene
complex. However, while the reaction energy for the Cp* case is
-29.8 kcal/mol, it is only -8.8 kcal/mol for the cymene case. Since
the T*∆S term is about +8 kcal/mol at 298 K, together with a +4
kcal/mol term for the low [CO], CO addition is calculated to be
unfavorable for the cymene case, in agreement with observation.
The difference is thus thermodynamic in origin and must be
attributed to the weaker π basicity of the cationic species toward
the arriving CO ligand. Support for this hypothesis comes from
the calculated C/O stretching frequency for the Cp* adduct (1898
cm-1) versus the that for the (unobserved) cymene analogue (1967
cm-1).
Figure 2. DFT-optimized geometry of (a) doubly N-bound ligand isomer
and (b) the isomeric structure with one N and C(sp3), together with C(ortho)
and C(ipso) of one phenyl, bonded to Ru. In general, only the ipso phenyl
carbons are illustrated, except for one phenyl on P in b.
minimum only 9.3 kcal/mol higher than the ground state (Figure
2b). The implied weakness of one N-Ru bond in the ground state
was traced to structural features of the pendant PPh2NPh arm of
the ligand; it rotated during geometry optimization from being
nonbonding to Ru to a position where one phenyl ring from
phosphorus donates to the metal,13-16 thus compensating partly for
the N f Ru bond dissociation energy cost, and thereby stabilizing
this transient isomer. The C(ipso)-C(ortho) part of one phenyl
π-system donates to Ru (Ru-C(ipso) ) 2.48 Å, Ru-C(ortho) )
2.46 Å), compared to Ru-C(C5Me5) distances of 2.22 Å. Other
Ru-C distances to this phenyl are >3 Å. As a result, this C-C
distance is 0.03 Å longer than the other five C-C distances in that
phenyl ring. This phenyl donation occurs by deforming one HC-
(sp3)-P-C(ipso) angle 3° smaller than the other one on that P.
We suggest that this less stable isomer17 is thermally accessible
and thus of potential kinetic significance as the key to understanding
the facile CO addition to an 18-electron species (Figure 1).
Subsequent displacement of the aryl f Ru interaction in Figure
2b by CO is clearly favorable based on relative π acidity of phenyl
versus CO, disruption of aromaticity in the reactant, and relief of
steric congestion. Note particularly that the proposed intermediacy
of this isomer naturally accounts for the stereoselectivity of the
product, which has Cp* trans to the pendant PPh2NPh, provided
CO attacks on the side of the phenyl syn to Ru.
The broader implication of this result is that the many large
substituents in HC(PPh2NPh)2 not only create a crowded molecule,
but these phenyls can provide transient stabilization to “quasi-
unsaturated” intermediates,18 and thus play an active role in the
electronic structure of such intermediates: electronic factors
supplement steric factors. This ligand class is thus susceptible to
special substituent effects, as has already been demonstrated19 for
HC(PPh2NSiMe3)2-1. Moreover, in contrast to the usual concept
of the chelate effect keeping a ligand attached to a metal, here it
is the low energy (i.e., accessibility) of arm-off or alternative binding
modes that is the strength of this ligand class for promoting facile
ligand addition.
Acknowledgment. This work was supported by the Department
of Energy. We thank the Russian Academy of Sciences Joint
Supercomputer Center for computing time.
Supporting Information Available: CIF files and bond lengths
and angles for two X-ray structures, together with full details, and
Cartesian coordinates, and drawings of the DFT-optimized structures.
This material is available free of charge via the Internet at http://
pubs.acs.org.
References
(1) Gamer, M. T.; Dehnen, S.; Roesky, P. W. Organometallics 2001, 20, 4230.
(2) Gamer, M. T.; Roesky, P. W. J. Organomet. Chem. 2002, 647, 123.
(3) Wei, P.; Stephan, D. W. Organometallics 2002, 21, 1308.
(4) Hill, M. S.; Hitchcock, P. B. J. Chem. Soc., Dalton Trans. 2002, 4694.
(5) Evans, D. J.; Hill, M. S.; Hitchcock, P. B. Dalton Trans. 2003, 570.
(6) Kocher, N.; Leusser, D.; Murso, A.; Stalke, D. Chem.sEur. J. 2004, 10,
3622.
(7) Bibal, C.; Pink, M.; Smurnyy, Y. D.; Tomaszewski, J.; Caulton, K. G. J.
Am. Chem. Soc. 2004, 126, 2312.
(8) Jones, N. D.; Lin, G.; Gossage, R. A.; McDonald, R.; Cavell, R. G.
Organometallics 2003, 22, 2832.
(9) Benson, J. W.; Keiter, R. L.; Keiter, E. A.; Rheingold, A. L.; Yap, G. P.
A.; Mainz, V. V. Organometallics 1998, 17, 4275.
(10) All energies quoted here are electronic energies.
(11) Jimenez-Tenorio, M.; Puerta, M. C.; Valerga, P. Eur. J. Inorg. Chem.
2004, 17.
(12) Nagashima, H.; Kondo, H.; Hayashida, T.; Yamaguchi, Y.; Gondo, M.;
Masuda, S.; Miyazaki, K.; Matsubara, K.; Kirchner, K. Coord. Chem.
ReV. 2003, 245, 177.
(13) Dioumaev, V. K.; Szalda, D. J.; Hanson, J.; Franz, J. A.; Morris Bullock,
R. Chem. Commun. 2003, 1670.
(14) Marshall, W. J.; Grushin, V. V. Organometallics 2003, 22, 555.
(15) Alonso, E.; Fornies, J.; Fortuno, C.; Martin, A.; Orpen, A. G. Organo-
metallics 2000, 19, 2690.
(16) Cheng, T.-Y.; Szalda, D. J.; Morris Bullock, R. Chem. Commun. 1999,
1629.
(17) Vedernikov, A. N.; Caulton, K. G. New J. Chem. 2002, 26, 1267.
(18) Braunstein, P.; Naud, F. Angew. Chem., Int. Ed. 2001, 40, 680.
(19) Lin, G.; Jones, N. D.; Gossage, R. A.; McDonald, R.; Cavell, R. G. Angew.
Chem., Int. Ed. 2003, 42, 4054.
JA051296H
9
J. AM. CHEM. SOC. VOL. 127, NO. 25, 2005 8945