Cationic and formally zwitterionic rhodium(I) and iridium(I) derivatives of a P,N-substituted indene: A comparative synthetic, structural, and catalytic investigation

Article abstract of DOI:10.1021/om060758c

A family of neutral, cationic, and formally zwitterionic Rh(I) and Ir(I) complexes featuring 1-PⁱPr₂-2-NMe₂-indene (3a), 3-PⁱPr₂-2-NMe₂-indene (3b), or 3-P _iPr₂-2-NMe2-indenide (3c) ligands have been prepared and structurally characterized. Whereas treatment of 3a with 0.5 equiv of [(COD)RhCl]₂ afforded (COD)RhCl(κ¹-P,N-3a) (4a) in 96% isolated yield, [(COD)M(κ²-P,N-3a)]⁺X ^- ([5a]⁺PF₆^-, M = Rh; [6a] ⁺SO₃CF₃^-, M = Ir) complexes were obtained in 14% and 41% isolated yield from 3a and the appropriate in-situ-prepared [(COD)M(THF)₂]⁺X^- precursor. The isomeric complexes [(COD)M(κ²-P,N-3b)]⁺X ^- ([5b]⁺X^-, M = Rh; [6b]⁺X ^-, M = Ir) were prepared using a similar protocol employing NEt ₃, with isolated yields ranging from 37% to 94% depending on the identity of M and X. The zwitterionic complexes (COD)M(κ²-P,N- 3c) (5c, M = Rh; 6c, M = Ir) were prepared in 89% and 93% isolated yield, either upon treatment of [5b]⁺X^- or [6b]⁺X ^- with NaN(SiMe₃)₂ or via addition of 0.5 equiv of [(COD)-MCl]₂ to [3c]Li. The Rh(I) complexes (CO)(Cl) Rh(κ²-P,N-3b) (7) and [(CO)₂Rh(κ²- P,N-3b)]⁺PF₆^- (8) were prepared in 43% and 66% isolated yield. Single-crystal X-ray diffraction data were obtained for 4a·CH₂Cl₂, [5b]⁺BF^{4 -}, 5c, [6b]⁺SO₃CF₃^-, [6b]⁺PF₆^-, 6c, and 7, the analysis of which revealed an effectively isostructural relationship among the metal coordination environments in [5b]⁺X^-, [6b]⁺X^-, 5c, and 6c. Despite the structural similarities that are apparent within the cation/zwitterion pairs [5b]⁺X^-/5c and [6b] ⁺X^-/6c, reactivity studies involving these complexes highlighted important distinctions within and between these pairs of complexes. While both the κ²-P,N-3b and the COD ligands in [5b] ⁺BF₄^- were displaced upon exposure to excess PMePh₂, under similar conditions only the COD ligand in 5c underwent substitution, affording (PMePh₂)₂Rh(κ²-P, N-3c) (9) in 66% isolated yield. Temperature-dependent ¹H NMR line shape changes observed for 9 were rationalized in terms of the operation of at least one dynamic process in this complex (ΔG?₂₉₈ ≈14 kcal/mol). Further reactivity differences were observed when [5b] ⁺PF₆^-, [6b]⁺PF₆ ^-, 5c, and 6c were employed as catalysts for the hydrogenation and hydrosilylation of alkenes. While [6b]⁺PF₆^- proved to be the most effective styrene hydrogenation catalyst in CH ₂Cl₂ or THF, similarly high conversion to ethylbenzene was achieved by use of 5c as a catalyst in benzene, a solvent in which [5b] ⁺PF₆^- and [6b]⁺PF₆ ^- are not soluble. The catalytic utility of 5c and 6c was demonstrated further in hydrosilylation reactions employing triethylsilane and styrene, wherein these zwitterionic complexes were observed to outperform their cationic relatives. While nearly quantitative yields were obtained by use of either 5c (toluene or THF) or 6c (1,2-dichloroethane or THF) as a catalyst, these complexes were found to exhibit remarkably high, but divergent, selectivity for E-1-triethylsilyl-2-phenylethene (10b) and 1-triethylsilyl-2- phenylethane (10a), respectively. By comparison, 10a and 10b were produced in a 1.3:1 ratio in reactions employing Crabtree's catalyst ([(COD)Ir(PCy ₃)(Py)]⁺PF₆^-, COD = η⁴-1,5-cyclooctadiene, Cy = cyclohexyl, Py = pyridine; 1) under similar conditions. When catalytic activity/selectivity and solvent compatibility are considered, the results of these catalytic studies suggest that zwitterionic species such as 5c and 6c represent an effective class of catalyst complexes for the addition of E-H bonds to unsaturated substrates, whose properties are in some cases complementary to those of more traditional Rh(I) and Ir(I) [(COD)M(κ²-P,N)]⁺X^- salts.

Full text of DOI:10.1021/om060758c

594
Organometallics 2007, 26, 594-608
Cationic and Formally Zwitterionic Rhodium(I) and Iridium(I)
Derivatives of a P,N-Substituted Indene: A Comparative Synthetic,
Structural, and Catalytic Investigation
Judy Cipot,† Robert McDonald,‡ Michael J. Ferguson,‡ Gabriele Schatte,§ and
Mark Stradiotto*,†
Department of Chemistry, Dalhousie UniVersity, Halifax, NoVa Scotia, Canada B3H 4J3,
X-Ray Crystallography Laboratory, Department of Chemistry, UniVersity of Alberta, Edmonton, Alberta,
Canada T6G 2G2, and Saskatchewan Structural Sciences Centre, UniVersity of Saskatchewan, Saskatoon,
Saskatchewan, Canada S7N 5C9
ReceiVed August 21, 2006
A family of neutral, cationic, and formally zwitterionic Rh(I) and Ir(I) complexes featuring 1-PiPr2-
2-NMe2-indene (3a), 3-PiPr2-2-NMe2-indene (3b), or 3-PiPr2-2-NMe2-indenide (3c) ligands have been
prepared and structurally characterized. Whereas treatment of 3a with 0.5 equiv of [(COD)RhCl]2 afforded
-
(COD)RhCl(κ1-P,N-3a) (4a) in 96% isolated yield, [(COD)M(κ2-P,N-3a)]+X- ([5a]+PF6 , M ) Rh;
-
[6a]+SO3CF3 , M ) Ir) complexes were obtained in 14% and 41% isolated yield from 3a and the
appropriate in-situ-prepared [(COD)M(THF)2]+X- precursor. The isomeric complexes [(COD)M(κ2-P,N-
3b)]+X- ([5b]+X-, M ) Rh; [6b]+X-, M ) Ir) were prepared using a similar protocol employing NEt3,
with isolated yields ranging from 37% to 94% depending on the identity of M and X. The zwitterionic
complexes (COD)M(κ2-P,N-3c) (5c, M ) Rh; 6c, M ) Ir) were prepared in 89% and 93% isolated yield,
either upon treatment of [5b]+X- or [6b]+X- with NaN(SiMe3)2 or via addition of 0.5 equiv of [(COD)-
-
MCl]2 to [3c]Li. The Rh(I) complexes (CO)(Cl)Rh(κ2-P,N-3b) (7) and [(CO)2Rh(κ2-P,N-3b)]+PF6 (8)
were prepared in 43% and 66% isolated yield. Single-crystal X-ray diffraction data were obtained for
-
-
-
4a‚CH2Cl2, [5b]+BF4 , 5c, [6b]+SO3CF3 , [6b]+PF6 , 6c, and 7, the analysis of which revealed an
effectively isostructural relationship among the metal coordination environments in [5b]+X-, [6b]+X-,
5c, and 6c. Despite the structural similarities that are apparent within the cation/zwitterion pairs [5b]+X-/
5c and [6b]+X-/6c, reactivity studies involving these complexes highlighted important distinctions within
-
and between these pairs of complexes. While both the κ2-P,N-3b and the COD ligands in [5b]+BF4
were displaced upon exposure to excess PMePh2, under similar conditions only the COD ligand in 5c
underwent substitution, affording (PMePh2)2Rh(κ2-P,N-3c) (9) in 66% isolated yield. Temperature-
1
dependent H NMR line shape changes observed for 9 were rationalized in terms of the operation of at
least one dynamic process in this complex (∆Gq ≈ 14 kcal/mol). Further reactivity differences were
298
-
-
observed when [5b]+PF6 , [6b]+PF6 , 5c, and 6c were employed as catalysts for the hydrogenation and
hydrosilylation of alkenes. While [6b]+PF6- proved to be the most effective styrene hydrogenation catalyst
in CH2Cl2 or THF, similarly high conversion to ethylbenzene was achieved by use of 5c as a catalyst in
benzene, a solvent in which [5b]+PF6- and [6b]+PF6- are not soluble. The catalytic utility of 5c and 6c
was demonstrated further in hydrosilylation reactions employing triethylsilane and styrene, wherein these
zwitterionic complexes were observed to outperform their cationic relatives. While nearly quantitative
yields were obtained by use of either 5c (toluene or THF) or 6c (1,2-dichloroethane or THF) as a catalyst,
these complexes were found to exhibit remarkably high, but divergent, selectivity for E-1-triethylsilyl-
2-phenylethene (10b) and 1-triethylsilyl-2-phenylethane (10a), respectively. By comparison, 10a and
10b were produced in a 1.3:1 ratio in reactions employing Crabtree’s catalyst ([(COD)Ir(PCy3)(Py)]+PF6 ,
-
COD ) η4-1,5-cyclooctadiene, Cy ) cyclohexyl, Py ) pyridine; 1) under similar conditions. When
catalytic activity/selectivity and solvent compatibility are considered, the results of these catalytic studies
suggest that zwitterionic species such as 5c and 6c represent an effective class of catalyst complexes for
the addition of E-H bonds to unsaturated substrates, whose properties are in some cases complementary
to those of more traditional Rh(I) and Ir(I) [(COD)M(κ2-P,N)]+X- salts.
square-planar complexes of the heavier group 9 metals has
emerged as an important theme in modern organometallic
research, since such species are among the most active and
widely used classes of homogeneous catalysts for the addition
of E-H and E-E bonds (E ) main group element) to
Introduction
Building on the seminal investigations by Schrock and
Osborn,1 and later by Crabtree,2 the development of cationic
* Author to whom correspondence should be addressed. Fax: 1-902-
494-1310. Tel: 1-902-494-7190. E-mail: mark.stradiotto@dal.ca.
† Dalhousie University.
(1) (a) Schrock, R. R.; Osborn, J. A. J. Am. Chem. Soc. 1976, 98, 2134.
(b) Schrock, R. R.; Osborn, J. A. J. Am. Chem. Soc. 1976, 98, 2143.
(2) Crabtree, R. Acc. Chem. Res. 1979, 12, 331.
‡ University of Alberta.
§ University of Saskatchewan.
10.1021/om060758c CCC: $37.00 © 2007 American Chemical Society
Publication on Web 01/03/2007

Rh and Ir κ2-P,N Complexes
Organometallics, Vol. 26, No. 3, 2007 595
unsaturated substrates.3 The groundbreaking work of Knowles4
and more recent discoveries by Pfaltz5 demonstrate that the
rational development of such catalysts can enable significant
advancements in metal-mediated synthesis. However, the range
of experimental conditions under which such cationic group 9
species can be employed is limited, since these catalysts
commonly exhibit poor solubility in low-polarity media and can
be rendered inactive in strongly coordinating solvents.6 Fur-
thermore, the systematic design of cationic catalysts is com-
plicated by the fact that the choice of accompanying outer-sphere
counteranion in such discrete salts can influence catalytic
performance (i.e., activity, selectivity, lifetime, etc.), with some
of the most effective counteranions (e.g., tetrakis(3,5-(trifluo-
romethyl)phenyl)borate, BArF-) being relatively costly to obtain
from commercial sources and/or hazardous to prepare.7 In
principle, the use of less polar neutral catalyst complexes
represents a way in which to circumvent the aforementioned
drawbacks associated with cationic species; however, neutral
group 9 catalysts can be less active than their cationic relatives.
This trend is particularly evident in the case of Ir(I) alkene hydro-
genation catalysts. While Crabtree’s catalyst ([(COD)Ir(PCy3)-
(Py)]+PF6-, COD ) η4-1,5-cyclooctadiene, Cy ) cyclohexyl,
Py ) pyridine; 1)2 is capable of hydrogenating a much wider
range of alkene substrates than can be reduced by use of
Wilkinson’s catalyst ((PPh3)3RhCl, 2),8 the related neutral Ir(I)
complex (PPh3)3IrCl is inactive for alkene hydrogenation.9
ligand by more strongly coordinating Lewis bases.12 Other
classes of Rh(I) zwitterions featuring homobidentate or homo-
tridentate anionic ancillary ligands based on either P or N donors
have also emerged, including those in which the anionic charge
carrier is a substituted carborane,13 sulfate,14 or borate15 frag-
ment. Notably, Betley and Peters15b,c have demonstrated that
zwitterionic square-planar Rh(I) complexes featuring either κ2-
[Ph2B(CH2PR2)2]- or κ2-[Ph2B(CH2NR2)2]- ancillary ligands
are active catalysts for various E-H bond additions to alkenes
in the presence of both high- and low-polarity solvents.
However, a more recent report by Tilley and co-workers15a
highlights the fact that this general class of zwitterions are also
susceptible to decomposition by way of B-C bond cleavage to
yield nonzwitterionic bis(phosphino)borane Rh(I) derivatives.16
By comparison, zwitterionic complexes featuring formally
cationic Ir centers are much less common than their Rh relatives,
and while the synthetic utility of cationic Ir(I) catalysts such as
1 is well-established,2,5 the reactivity properties of related Ir(I)
zwitterions have not been explored systematically.17
In this context, we have established recently synthetic routes
to an entirely new class of zwitterionic metal complexes derived
from donor-substituted indenes.18 In contrast to the facial (η5)
binding of metal fragments that is commonly observed in
indenylmetal chemistry,19,20 these unusual zwitterions can be
(11) Selected synthetic and catalytic studies: (a) Vogels, C. M.; Decken,
A.; Westcott, S. A. Can. J. Chem. 2006, 84, 146. (b) Van den Hoven, B.
G.; Alper, H. J. Am. Chem. Soc. 2001, 123, 10214, and references therein.
(c) Dai, C.; Robins, E. G.; Scott, A. J.; Clegg, W.; Yufit, D. S.; Howard,
J. A. K.; Marder, T. B. Chem. Commun. 1998, 1983. (d) Kishimoto, Y.;
Itou, M.; Miyatake, T.; Ikariya, T.; Noyori, R. Macromolecules 1995, 28,
6662. (e) Aresta, M.; Quaranta, E.; Albinati, A. Organometallics 1993, 12,
2032. (f) Westcott, S. A.; Blom, H. P.; Marder, T. B.; Baker, R. T. J. Am.
Chem. Soc. 1992, 114, 8863. (g) Longato, B.; Pilloni, G.; Graziani, R.;
Casellato, U. J. Organomet. Chem. 1991, 407, 369. (h) Schrock, R. R.;
Osborn, J. A. Inorg. Chem. 1970, 9, 2339.
Zwitterionic group 9 complexes that feature formal charge
separation between a cationic metal fragment and an associated
anionic ancillary ligand represent appealing candidates for
catalytic studies, since in principle such species espouse the
desirable traits of discrete salts and traditional neutral complexes
in offering a formally cationic metal fragment within an overall
neutral molecular framework.10 However, formally zwitterionic
group 9 complexes are much less common than their cationic
relatives, as a result of the challenges associated with construct-
ing such charge-separated species. The first zwitterionic Rh(I)
complexes to be explored in catalytic applications were those
supported by η6-coordinated tetraphenylborate and related
ligands.11 While such zwitterions have proven capable of
mediating a wide range of substrate transformations involving
E-H and/or E-E bond activation under mild conditions and
with a high degree of selectivity,11 it has been shown that these
complexes can be transformed into conventional cationic
[L4Rh]+BPh4- species upon substitution of the η6-boratoarene
(12) Zhou, Z.; Facey, G.; James, B. R.; Alper, H. Organometallics 1996,
15, 2496.
(13) Teixidor, F.; Vin˜as, C.; Abad, M. M.; Whitaker, C.; Rius, J.
Organometallics 1996, 15, 3154.
(14) (a) Barbaro, P.; Bianchini, C.; Dal Santo, V.; Meli, A.; Moneti, S.;
Psaro, R.; Scaffidi, A.; Sordelli, L.; Vizza, F. J. Am. Chem. Soc. 2006,
128, 7065, and references therein. (b) Kla¨ui, W.; Schramm, D.; Peters, W.
Eur. J. Inorg. Chem. 2001, 3113. (c) Dorta, R.; Shimon, L.; Milstein, D. J.
Organomet. Chem. 2004, 689, 751.
(15) (a) Turculet, L.; Feldman, J. D.; Tilley, T. D. Organometallics 2004,
23, 2488. (b) Betley, T. A.; Peters, J. C. Angew. Chem., Int. Ed. 2003, 42,
2385. (c) Betley, T. A.; Peters, J. C. Inorg. Chem. 2002, 41, 6541.
(16) The related degradation of a (phosphino)borate ligand has also been
observed in the context of Cu(I) chemistry: Thomas, J. C.; Peters, J. C.
Polyhedron 2004, 23, 2901.
(3) For selected reviews, see: (a) Beletskaya, I. P.; Moberg, C. Chem.
ReV. 2006, 106, 2320. (b) Alonso, F.; Beletskaya, I. P.; Yus, M. Chem.
ReV. 2004, 104, 3079. (c) Tanaka, M. Top. Curr. Chem. 2004, 232, 25. (d)
Reichl, J. A.; Berry, D. A. AdV. Organomet. Chem. 1999, 43, 197. (e) Han,
L.-B.; Tanaka, M. Chem. Commun. 1999, 395. (f) Beletskaya, I.; Pelter, A.
Tetrahedron 1997, 53, 4957. (g) Dickson, R. S. Homogeneous Catalysis
with Compounds of Rhodium and Iridium; Reidel: Dordrecht, 1985.
(4) Knowles, W. S. Angew. Chem., Int. Ed. 2002, 41, 1998.
(5) Bell, S.; Wu¨stenberg, B.; Kaiser, S.; Menges, F.; Netscher, T.; Pfaltz,
A. Science 2006, 311, 642.
(6) (a) Crabtree, R. H.; Demou, P. C.; Eden, D.; Mihelcic, J. M.; Parnell,
C. A.; Quirk, J. M.; Morris, G. E. J. Am. Chem. Soc. 1982, 104, 6994. (b)
Shapley, J. R.; Schrock, R. R.; Osborn, J. A. J. Am. Chem. Soc. 1969, 91,
2816.
(17) For selected preparations of formally zwitterionic Ir complexes,
see: (a) Tejel, C.; Ciriano, M. A.; Jime´nez, S.; Oro, L. A.; Graiff, C.;
Tiripicchio, A. Organometallics 2005, 24, 1105. (b) Feldman, J. D.; Peters,
J. C.; Tilley, T. D. Organometallics 2002, 21, 4050. (c) Feldman, J. D.;
Peters, J. C.; Tilley, T. D. Organometallics 2002, 21, 4065. (d) Torres, F.;
Sola, E.; Mart´ın, M.; Ochs, C.; Picazo, G.; Lo´pez, J. A.; Lahoz, F. J.; Oro,
L. A. Organometallics 2001, 20, 2716. (e) Peters, J. C.; Feldman, J. D.;
Tilley, T. D. J. Am. Chem. Soc. 1999, 121, 9871. (f) Reference 15a herein.
(18) (a) Wile, B. M.; Burford, R. J.; McDonald, R.; Ferguson, M. J.;
Stradiotto, M. Organometallics 2006, 25, 1028. (b) Rankin, M. A.;
McDonald, R.; Ferguson, M. J.; Stradiotto, M. Angew. Chem., Int. Ed. 2005,
44, 3603. (c) Cipot, J.; McDonald, R.; Stradiotto, M. Chem. Commun. 2005,
4932. (d) Stradiotto, M.; Cipot, J.; McDonald, R. J. Am. Chem. Soc. 2003,
125, 5618.
(7) (a) Smidt, S. P.; Zimmermann, N.; Studer, M.; Pfaltz, A. Chem.-
Eur. J. 2004, 10, 4685. (b) Reger, D. L.; Wright, T. D.; Little, C. A.; Lamba,
J. J. S.; Smith, M. D. Inorg. Chem. 2001, 40, 3810. (c) Moore, E. J.;
Waymouth, R. Chem. Eng. News 1997, 75 (March 17), 6.
(8) Osborn, J. A.; Jardine, F. H.; Young, J. F.; Wilkinson, G. J. Chem.
Soc. A 1966, 1711.
(9) Bennett, M. A.; Milner, D. L. J. Am. Chem. Soc. 1969, 91, 6983.
(10) For the purposes of the discussions herein, “zwitterionic” complexes
are defined as those species that formally lack conventional resonance
structures that place an anionic charge onto one of the donor atoms of the
ancillary ligand.
(19) (a) Lobanova, I. A.; Zdanovich, V. I. Russ. Chem. ReV. 1988, 57,
967. (b) Birmingham, J. M. AdV. Organomet. Chem. 1964, 2, 365.
(20) For some recent examples of η5-indenyl-rhodium and -iridium
complexes, see: (a) Santi, S.; Ceccon, A.; Carli, F.; Crociani, L.; Bisello,
A.; Tiso, M.; Venzo, A. Organometallics 2002, 21, 2679. (b) Santi, S.;
Ceccon, A.; Crociani, L.; Gambaro, A.; Ganis, P.; Tiso, M.; Venzo, A.;
Bacchi, A. Organometallics 2002, 21, 565. (c) Schumann, H.; Stenzel, O.;
Dechert, S.; Girgsdies, F.; Blum, J.; Gelman, D.; Halterman, R. L. Eur. J.
Inorg. Chem. 2002, 211. (d) Rupert, K. C.; Liu, C. C.; Nguyen, T. T.;
Whitener, M. A.; Sowa, J. R., Jr. Organometallics 2002, 21, 144. (e)
Schumann, H.; Stenzel, O.; Dechert, S.; Girgsdies, F.; Halterman, R. L.

596 Organometallics, Vol. 26, No. 3, 2007
Cipot et al.
Scheme 1. Modular Synthesis of the P,N Ligands 3a, 3b,
and 3c
Scheme 2. Synthesis of Rh(I) and Ir(I) Complexes Derived
from 3aa
viewed as comprising a formally cationic metal fragment whose
charge is counterbalanced by a sequestered, uncoordinated 10π-
electron indenide unit that is built into the backbone of a
bidentate ancillary ligand.21 In light of the superior catalytic
properties commonly exhibited by platinum-group metal com-
plexes featuring κ2-P,N ancillary ligands compared to analogous
P,P or N,N species,2,22 and given the lack of published synthetic
routes to heterobidentate group 9 zwitterions, we identified P,N-
substituted indenides as particularly important ligand targets.
As outlined in Scheme 1 for 1-PiPr2-2-NMe2-indene (3a), the
indene precursors to such indenide ligands are prepared easily
in a high-yielding and modular fashion via condensation of the
commercially available reagents 2-indanone and HNR2, followed
by deprotonation of the ensuing 2-aminoindene and quenching
with a chlorophosphine. While in solution the first-formed allylic
indene 3a evolves slowly to an equilibrium mixture of 3a and
the vinylic isomer 3b,23 both are readily deprotonated to afford
cleanly the desired P,N-indenide 3c, which has proven to be an
effective ancillary ligand for supporting κ2-P,N zwitterionic
complexes.18
aReagents: (i) 0.5 [(COD)RhCl]2; (ii) 0.5 [(COD)MCl]2, AgX (M
) Rh, X ) PF6; M ) Ir, X ) SO3CF3); (iii) NEt3; (iv) 0.5
[(COD)MCl]2, AgX (X ) SO3CF3, BF4, or PF6) then NEt3 (except for
M ) Rh and X ) BF4); (v) NaN(SiMe3)2; (vi) n-BuLi, then 0.5
[(COD)MCl]2 (COD ) η4-1,5-cyclooctadiene).
exploited as representative E-H bond addition reactions. The
synthetic utility of these new group 9 complexes was confirmed
during the course of this catalytic survey, and their performance
was benchmarked against those of the prototypical group 9
catalysts, 1 and 2. When catalytic activity/selectivity and solvent
compatibility are considered, the results of these catalytic studies
suggest that zwitterionic Rh(I) and Ir(I) species supported by
κ2-P,N-3c represent an effective class of catalyst complexes for
the addition of E-H bonds to unsaturated substrates, whose
properties are in some cases complementary to those of more
traditional Rh(I) and Ir(I) [(COD)M(κ2-P,N)]+X- salts. Herein
we report the findings of this multifaceted study involving κ2-
P,N Rh(I) and Ir(I) derivatives of 3c, the first formally
zwitterionic analogues of ubiquitous heterobidentate κ2-P,N
square-planar group 9 cations. A portion of these results has
been communicated previously.18c,d
In an effort to assess the extent to which group 9 zwitterions
supported by κ2-P,N-3c compare with more conventional square-
planar Rh(I) and Ir(I) cationic complexes, we have carried out
a series of synthetic and structural studies in which cations
supported by κ2-P,N-3b were compared with zwitterionic
relatives that feature the anionic ligand κ2-P,N-3c.24,25 So as to
correlate these structural results with experimentally observed
reactivity, the hydrogenation and hydrosilylation of alkenes were
Results and Discussion
Organometallics 2001, 20, 5360. (f) Schumann, H.; Stenzel, O.; Dechert,
S.; Girgsdies, F.; Halterman, R. L. Organometallics 2001, 20, 2215. (g)
Westcott, S. A.; Kakkar, A. K.; Taylor, N. J.; Roe, D. C.; Marder, T. B.
Can. J. Chem. 1999, 77, 205.
(21) For discussions regarding the aromatic character of indenide anions,
see: Jiao, H.; Schleyer, P. v. R.; Mo, Y.; McAllister, M. A.; Tidwell, T. T.
J. Am. Chem. Soc. 1997, 119, 7075.
Synthesis and Characterization of Neutral and Cationic
Complexes of 3a. In a preliminary synthetic survey, the ability
of 3a to support neutral and cationic Rh(I) and Ir(I) species
was examined. Addition of 3a to 0.5 equiv of [(COD)RhCl]2
afforded (COD)RhCl(κ1-P,N-3a) (4a) in 96% isolated yield
(Scheme 2). Data obtained from NMR spectroscopic studies
indicated that the connectivity within the monodentate ligand
3a is retained in the square-planar complex 4a. This was
confirmed through the crystallographic characterization of 4a‚
CH2Cl2; an ORTEP26 diagram is provided in Figure 1, while
salient X-ray experimental data and selected metrical parameters
are collected in Tables 1 and 2, respectively. In 4a the Rh-
alkene distances trans to P are significantly longer than those
trans to Cl, in keeping with established structural trends in
(22) For selected reviews, see: (a) Ka¨llstro¨m, K.; Munslow, I.; Ander-
sson, P. G. Chem.-Eur. J. 2006, 12, 3194. (b) Guiry, P. J.; Saunders, C. P.
AdV. Synth. Catal. 2004, 346, 497. (c) Chelucci, G.; Orru`, G.; Pinna, G. A.
Tetrahedron 2003, 59, 9471. (d) Helmchen, G.; Pfaltz, A. Acc. Chem. Res.
2000, 33, 336. (e) Gavrilov, K. N.; Polosukhin, A. I. Russ. Chem. ReV.
2000, 69, 661. (f) Cui, X.; Burgess, K. Chem. ReV. 2005, 105, 3272.
(23) Cipot, J.; Wechsler, D.; Stradiotto, M.; McDonald, R.; Ferguson,
M. J. Organometallics 2003, 22, 5185.
(24) For some conceptually related studies examining the effect of formal
charge separation on metal-centered reactivity, see: (a) Flanagan, S. P.;
Guiry, P. J. J. Organomet. Chem. 2006, 691, 2125. (b) Fujisawa, K.; Ono,
T.; Ishikawa, Y.; Amir, N.; Miyashita, Y.; Okamoto, K.; Lehnert, N. Inorg.
Chem. 2006, 45, 1698. (c) Hahn, C. Chem.-Eur. J. 2004, 10, 5888. (d)
Thomas, C. M.; Peters, J. C. Inorg. Chem. 2004, 43, 8. (e) Thomas, J. C.;
Peters, J. C. J. Am. Chem. Soc. 2003, 125, 8870. (f) Padilla-Mart´ınez, I. I.;
Poveda, M. L.; Carmona, E.; Monge, M. A.; Ruiz-Valero, C. Organome-
tallics 2002, 21, 93. (g) Seymore, S. B.; Brown, S. N. Inorg. Chem. 2000,
39, 325. (h) Hoic, D. A.; Davis, W. M.; Fu, G. C. J. Am. Chem. Soc. 1996,
118, 8176. (i) Lorkovic, I. M.; Duff, R. R., Jr.; Wrighton, M. S. J. Am.
Chem. Soc. 1995, 117, 3617. (j) Quan, R. W.; Bazan, G. C.; Kiely, A. F.;
Schaefer, W. P.; Bercaw, J. E. J. Am. Chem. Soc. 1994, 116, 4489. (k)
References 15b,c herein.
(25) A computational interrogation of Rh(I) complexes of this type
employing DFT methods, as well as the quantum theory of atoms in
molecules (QTAIM), has afforded an opportunity to evaluate changes in
the electronic characteristics of a formally cationic group 9 metal fragment
on going from cationic complexes featuring κ2-P,N-3b ligation to isosteric
zwitterions incorporating κ2-P,N-3c. We will report on the results of such
DFT/QTAIM studies elsewhere.
(26) ORTEP-3 for Windows, version 1.074: Farrugia, L. J. J. Appl.
Crystallogr. 1997, 30, 565.

Rh and Ir κ2-P,N Complexes
Organometallics, Vol. 26, No. 3, 2007 597
Figure 1. ORTEP diagrams for 4a‚CH2Cl2, [5b]+BF4-, 5c, [6b]+SO3CF3-, [6b]+PF6-, and 6c, shown with 50% displacement ellipsoids
and with the atomic numbering scheme depicted. The dichloromethane solvate and selected hydrogen atoms have been omitted for clarity.
-
Only one of the two crystallographically independent molecules of [6b]+PF6 is shown.
coordination chemistry. As well, the somewhat short N-C2
distance (1.398(3) Å) suggests that the uncoordinated NMe2
fragment in 4a is in partial conjugation with the adjacent indene
framework. In contrast to the rearrangement of 3a to 3b that is
observed both for uncoordinated 3a in solution23 and for 3a
within the coordination sphere of cationic Rh(I) and Ir(I)
complexes (vide infra), the corresponding isomerization of 4a
to (COD)RhCl(κ1-P,N-3b) was not detected. Analogous reac-
tions employing [(COD)IrCl]2 resulted in the consumption of
3a, along with the clean formation of two phosphorus-containing
products (δ 31P ) 6.8 and 36.1). We have thus far not been
able to isolate and unequivocally identify these products, which
may correspond to a (COD)IrCl(κ1-P,N-3a,b) mixture. However,
the high-frequency 31P NMR resonance is also consistent with
the formation of [(COD)Ir(κ2-P,N-3b)]+Cl- ([6b]+Cl-, vide
infra).
(41%) were we successful in exploiting the differing solubilities
of the κ2-P,N-3a and κ2-P,N-3b isomers as a means of isolating
pure samples of these complexes (Scheme 2). Data obtained
1
from H and 13C NMR experiments are in keeping with the
C1-symmetric nature of [5a]+PF6 and [6a]+SO3CF3 and
confirm the allylic structure of the bound 3a ligand in these
complexes. It is evident that [5a]+X- and [6a]+X- represent
kinetic products in these reactions and that the net 1,3-proton
shift within the indene backbone of 3a, leading to thermody-
namically favored [5b]+X- or [6b]+X- products featuring κ2-
P,N-3b ligands, is facilitated upon coordination to [(COD)M]+
fragments in this system. We have observed that while in
solution the isomerization of isolated [5a]+PF6- or [6a]+SO3CF3-
to the corresponding κ2-P,N-3b complex is slow and does not
proceed to completion after several days at 24 °C. under similar
conditions such transformations are rendered rapid (hours) and
quantitative in the presence of NEt3. In contrast, we have
demonstrated previously that uncoordinated 3a is not trans-
formed cleanly into 3b upon treatment with amine.23
-
-
In the pursuit of [(COD)M(κ2-P,N-3a)]+X- ([5a]+X-, M )
Rh; [6a]+X-, M ) Ir; X ) SO3CF3, BF4, or PF6) complexes,
appropriate [(COD)M(THF)2]+X- precursors (prepared in-situ)
were treated with 3a at 24 °C, and the progress of each reaction
was monitored by use of 31P NMR methods. Notably, the nature
of the metal/counteranion combination was found to figure
importantly in determining the course of these transformations.
For reactions featuring SO3CF3- or PF6- counterions, [5a]+X-
or [6a]+X- and the corresponding [(COD)M(κ2-P,N-3b)]+X-
([5b]+X-, M ) Rh; [6b]+X-, M ) Ir) isomer were observed
in various proportions as the only products in the crude reaction
mixture, even after short reaction times (minutes). While a
similar isomeric [6a,b]+X- mixture was detected spectroscopi-
Synthesis and Characterization of the Cationic Complexes
[5b]+X- and [6b]+X-. The incomplete rearrangement of 3a to
3b in solution prevents the direct use of the latter in the synthesis
of κ2-P,N-3b metal complexes. However, we have discovered
that the target complexes [5b]+X- and [6b]+X- are formed quan-
titatively (31P NMR) upon addition of 3a to [(COD)M(THF)2]+X-
(prepared in-situ), followed by treatment with NEt3 (Scheme
2); [5b]+BF4- is generated cleanly from [(COD)Rh(THF)2]+BF4-
and 3a, even in the absence of added amine (vide supra). After
workup, [5b]+X- (X ) SO3CF3, 87%; BF4, 94%; PF6, 73%)
or [6b]+X- (X ) SO3CF3, 89%; BF4, 37%; PF6, 78%) were
-
cally in reactions employing the BF4 counteranion, under
analogous conditions the Rh(I) cation [5b]+X- was observed
exclusively. Only for [5a]+PF6 (14%) and [6a]+SO3CF3
-
-

Table 1. Crystallographic Data for 4a, [5b]+BF4-, [6b]+SO3CF3-, [6b]+PF6-, 5c, 6c, and 7
[5b]+BF4
[6b]+SO3CF3
[6b]+PF6
5c
6c
7
-
-
-
4a·CH2Cl2
empirical formula
fw
C26H40Cl3NPRh
606.82
C25H38BF4NPRh
573.25
C26H38F3O3NPSIr
724.80
C25H38F6NP2Ir
720.70
C25H37NPRh
485.44
C25H37NPIr
574.73
C18H26ClNOPRh
652.55
cryst dimens
cryst syst
space group
a (Å)
b (Å)
c (Å)
0.61 × 0.31 × 0.17
monoclinic
P21/n
15.773(1)
8.2243(6)
20.994(2)
91.688(1)
2722.2(3)
4
0.26 × 0.24 × 0.23
monoclinic
P21/n
9.893(1)
18.375(2)
14.129(1)
90.911(2)
2568.1(5)
4
0.22 × 0.14 × 0.08
monoclinic
P21/n
9.9608(6)
20.270(1)
14.0988(9)
103.575(1)
2767.2(3)
4
0.53 × 0.13 × 0.13
monoclinic
P21/m
14.535(1)
9.9545(9)
18.506(2)
91.182(1)
2677.0(4)
4
0.28 × 0.25 × 0.10
monoclinic
P21/n
9.7054(6)
16.985(1)
14.5743(9)
109.392(1)
2266.2(2)
4
0.36 × 0.18 × 0.13
monoclinic
P21/n
9.6714(5)
16.9783(9)
14.5080(7)
109.4177(7)
2246.8(2)
4
0.08 × 0.08 × 0.05
monoclinic
P21
9.4380(4)
10.7550(7)
9.7100(6)
102.118(3)
963.7(1)
2
â (deg)
V (Å3)
Z
F
calcd (g cm-3
)
1.481
0.996
52.74
1.483
0.770
52.86
1.740
5.007
52.82
1.788
5.164
52.84
1.423
0.835
52.76
1.699
6.025
52.72
1.522
1.111
54.94
µ (mm-1
)
2θ limit (deg)
-19 e h e 16
-10 e k e 10
-26 e l e 26
15 470
5548
0.0205
-10 e h e 12
-22 e k e 20
-17 e l e 17
14 027
5224
0.0310
-12 e h e 12
-25 e k e 25
-17 e l e 17
19 876
5676
0.0621
-18 e h e 18
-12 e k e 12
-23 e l e 23
20 721
5809
0.0370
-12 e h e 12
-21 e k e 17
-17 e l e 18
12 397
4611
0.0321
-12 e h e 12
-21 e k e 21
-18 e l e 18
17 112
4579
0.0254
-12 e h e 12
-13 e k e 13
-12 e l e 12
4198
4198
NA
total no. of data collected
no. of indep reflns
Rint
no. of obsd reflns
range of transmn
no. of data/restraints/params
5091
4661
4270
5030
4033
4273
3818
0.8490-0.5818
5548/0/289
0.0328
0.0874
1.019
0.8428-0.8249
5224/0/298
0.0510
0.1415
1.147
0.6902-0.4055
5676/0/325
0.0315
0.0654
1.002
0.5533-0.1706
5809/0/571
0.0290
0.0740
1.088
0.9212-0.7998
4611/0/253
0.0329
0.0861
1.083
0.5081-0.2203
4579/0/253
0.0174
0.0426
1.086
0.9465-0.9164
4198/1/214
0.0398
0.0849
1.078
2
R1 [Fo2 g 2σ(Fo )]
2
wR2 [Fo2 g -3σ( Fo )]
goodness-of-fit

Rh and Ir κ2-P,N Complexes
Organometallics, Vol. 26, No. 3, 2007 599
Table 2. Selected Interatomic Distances (Å) and Angles (deg) for 4a, [5b]+BF4-, [6b]+SO3CF3-, [6b]+PF6-, 5c, and 6c
-
-
-
-
4a‚CH2Cl2
[5b]+BF4
[6b]+SO3CF3
[6ba]+PF6
[6bb]+PF6
5c
6c
M-P
2.3474(6)
2.305(1)
2.211(4)
2.239(5)
2.232(5)
2.140(5)
2.136(5)
1.809(4)
1.459(6)
1.508(6)
1.340(7)
1.486(6)
1.384(7)
1.396(7)
1.384(8)
1.390(7)
1.382(7)
1.502(7)
1.411(6)
3.074
2.299(1)
2.192(4)
2.212(4)
2.184(4)
2.152(4)
2.108(5)
1.807(5)
1.475(6)
1.500(6)
1.339(6)
1.491(6)
1.380(7)
1.396(7)
1.381(7)
1.383(7)
1.390(6)
1.498(6)
1.404(7)
3.072
2.301(2)
2.203(6)
2.17(1)
2.22(1)
2.14(3)
2.32(1)
2.231(8)
2.22(1)
2.170(9)
2.122(8)
2.13(1)
1.810(7)
1.460(9)
1.491(7)
1.346(7)
1.466(8)
1.380(8)
1.387(8)
1.364(9)
1.403(9)
1.382(9)
1.508(8)
1.394(7)
3.060
2.3173(6)
2.242(2)
2.275(2)
2.238(2)
2.123(3)
2.123(3)
1.758(2)
1.479(3)
1.386(3)
1.419(3)
1.446(3)
1.412(3)
1.387(4)
1.398(4)
1.383(4)
1.414(4)
1.430(4)
1.442(4)
3.088
2.3171(6)
2.216(2)
2.201(3)
2.231(3)
2.127(2)
2.116(2)
1.755(2)
1.478(3)
1.381(3)
1.413(3)
1.435(3)
1.406(3)
1.374(4)
1.411(4)
1.370(4)
1.412(4)
1.424(4)
1.440(3)
3.070
M-N
c
c
d
d
M-Calkene1
M-Calkene2
M-Calkene3
M-Calkene4
P-Cind
N-C2
C1-C2
C2-C3
C3-C3a
C3a-C4
2.202(3)
2.216(3)
2.113(3)
2.138(3)
1.877(2)
1.398(3)
1.533(3)
1.352(4)
1.451(4)
1.393(4)
1.396(5)
1.385(5)
1.400(4)
1.386(4)
1.528(3)
1.406(4)
2.974
2.13(3)
1.812(5)
1.462(8)
1.519(7)
1.321(7)
1.486(7)
1.391(8)
1.388(8)
1.381(8)
1.394(9)
1.379(7)
1.510(8)
1.406(7)
3.060
3.192
84.2(2)
111.6(5)
124.5(5)
116.9(4)
101.1(2)
C4-C5
C5-C6
C6-C7
C7-C7a
C1-C7a
C3a-C7a
M‚‚‚C2e
M‚‚‚C3e
P-M-N
M-N-C2
3.889
3.198
84.2(1)
112.1(3)
124.2(4)
117.4(3)
101.3(2)
3.199
84.3(1)
112.4(3)
123.8(4)
117.1(3)
101.7(2)
3.192
83.1(4)
111.3(6)
124.2(5)
117.1(4)
101.1(5)
3.190
3.189
83.92(5)
110.5(2)
120.6(2)
118.2(2)
102.15(8)
83.90(5)
110.8(1)
121.0(2)
117.5(2)
102.23(8)
N-C2-C3
C2-C3-P
C3-P-M
a,b Within the first and second independent molecules of [6b]+PF6-, respectively. The metal-cyclooctadiene distances trans to phosphorus. The metal-
c
d
e
cyclooctadiene distances trans to nitrogen or chloride. Transannular distance measured in the final refined structure.
isolated as analytically pure orange solids.27 The H, 13C, and
31P NMR data obtained for [5b]+X- and [6b]+X- are in each
case consistent with a square-planar [(COD)M(κ2-P,N-3b)]+X-
complex possessing Cs symmetry. As well, NMR signals
associated with the respective Rh(I) and Ir(I) cations are not
influenced by the nature of the accompanying counteranion, in
keeping with the proposed outer-sphere nature of X in these
salts. For [5b]+BF4-, [6b]+SO3CF3-, and [6b]+PF6-, the
solution structural characterization of these complexes was
confirmed on the basis of data obtained from single-crystal X-ray
diffraction experiments (Figure 1; Tables 1 and 2). The metrical
parameters associated with the cationic fragments in these
complexes differ only modestly, and no short contacts are
observed between M and X,28 or between M and the adjacent
C2dC3 unit. As well, the interatomic distances within the metal
coordination sphere of these complexes fall within the range of
those observed in related group 9 complexes,29 with the
M-alkene distances trans to P being significantly longer than
those trans to N.
served, and entirely new coordination motifs for this versatile
ligand are continuing to emerge.30 Alternatively, we envisioned
the use of donor-substituted variants of this anionic hydrocarbon
in the construction of zwitterionic κ2-P,N group 9 metal
complexes, in which the uncoordinated indenide unit serves as
a sequestered anionic charge reservoir, rather than as a site for
metal binding. Having successfully prepared cationic square-
planar group 9 complexes supported by κ2-P,N-3b, we turned
our attention to the synthesis of structurally analogous zwitte-
rions featuring κ2-P,N-3c ligation. Upon exposure to NaN-
(SiMe3)2, both [5b]+X- and [6b]+X- are transformed quanti-
tatively (31P NMR) into the desired κ2-P,N-3c complexes 5c
and 6c, respectively (Scheme 2). These zwitterionic species are
also generated cleanly from [3c]Li and 0.5 equiv of [(COD)-
MCl]2 and can be isolated as analytically pure orange-red solids
in high yield (M ) Rh, 89%; M ) Ir, 93%). The exclusive
formation of 5c and 6c in these transformations is unusual, given
the established propensity of [(COD)M]+ fragments for η5
binding to indenyl ligands.20 The connectivity in 5c and 6c that
was initially determined by use of 1H, 13C, and 31P NMR
methods was confirmed subsequently by use of X-ray crystal-
lographic techniques (Figure 1; Tables 1 and 2). Interestingly,
the metal-ligand contacts within 5c and 6c differ only modestly
from those found in the corresponding cations [5b]+X- and
[6b]+X-, thereby confirming the nearly isosteric relationship
between the κ2-P,N-3b and κ2-P,N-3c ligands in these complexes
with respect to the metal coordination sphere. However, while
pronounced bond length alternation is observed within the
indene portion of the κ2-P,N-3b ancillary ligands in these
cationic complexes, the indenide moiety of the κ2-P,N-3c ligand
in each of 5c and 6c exhibits a more delocalized structure,
consistent with a 10π-electron indenide framework. In fact, the
extent of delocalization observed within the carbocyclic back-
bone of the κ2-P,N-3c ligands in these zwitterionic complexes
1
Synthesis and Characterization of the Formally Zwitte-
rionic Complexes 5c and 6c. While the ability of the indenyl
fragment to coordinate metals in an η5 fashion is well-known,19
other indenyl-metal binding modes are also commonly ob-
-
(27) Attempts to prepare [6b]+B(C6F5)4 by employing Li(Et2O)2.5B-
(C6F5)4 as the halide-abstracting reagent under analogous conditions
produced dark solids that exhibited numerous 31P NMR resonances and
from which no pure Ir(I) complexes could be isolated.
(28) The shortest cation-anion contacts: (a) [5b]+BF4-, 2.78 Å between
F and C1(H)2; (b) [6b]+SO3CF3-, 2.44 Å between F and C1(H)2; (c)
[6b]+PF6-, 2.48 Å between F and C4-H.
(29) For a selection of crystallographically characterized Rh(I) and Ir(I)
[(COD)M(κ2-P,N)]+X- complexes, see: (a) Phillips, A. D.; Bolan˜o, S.;
Bosquain, S. S.; Daran, J.-C.; Malacea, R.; Peruzzini, M.; Poli, R.; Gonsalvi,
L. Organometallics 2006, 25, 2189. (b) Dahlenburg, L.; Go¨tz, R. Eur. J.
Inorg. Chem. 2004, 888. (c) Brauer, D. J.; Kottsieper, K. W.; Rossenbach,
S.; Stelzer, O. Eur. J. Inorg. Chem. 2003, 1748. (d) Hou, D.-R.; Reibenspies,
J.; Colacot, T. J.; Burgess, K. Chem.-Eur. J. 2001, 7, 5391. (e) Kataoka,
Y.; Imanishi, M.; Yamagata, T.; Tani, K. Organometallics 1999, 18, 3563.
(f) Yang, H.; Lugan, N.; Mathieu, R. Organometallics 1997, 16, 2089. (g)
Berger, H.; Nesper, R.; Pregosin, P. S.; Ru¨egger, H.; Wo¨rle, M. HelV. Chim.
Acta 1993, 76, 1520. (h) McKay, I. D.; Payne, N. C. Can. J. Chem. 1986,
64, 1930. (i) Reference 7b herein.
(30) (a) Bradley, C. A.; Lobkovsky, E.; Keresztes, I.; Chirik, P. J. J.
Am. Chem. Soc. 2005, 127, 10291. (b) Bradley, C. A.; Keresztes, I.;
Lobkovsky, E.; Young, V. G.; Chirik, P. J. J. Am. Chem. Soc. 2004, 126,
16937. (c) Zargarian, D. Coord. Chem. ReV. 2002, 233-234., 157. (d)
Stradiotto, M.; McGlinchey, M. J. Coord. Chem. ReV. 2001, 219-221, 311.

600 Organometallics, Vol. 26, No. 3, 2007
Cipot et al.
equals that observed in some discrete “naked” indenide salts.31
Although 5c and 6c formally lack a conventional resonance
contributor that places the anionic charge onto either of the
pnictogen donor groups, the short P-C3 and C1-C2 distances
in these complexes, when compared with those found in the
analogous cations [5b]+X- and [6b]+X-, suggest that a meth-
presence of AgPF6 generated a complex mixture of products.
While similarly complex product distributions were obtained
upon exposure of a THF solution of [5b]+PF6- to an atmosphere
of CO, clean conversion to a single phosphorus-containing
species (8, δ 31P ) 63) was achieved under similar conditions
in CH2Cl2, allowing for the subsequent isolation of 8 in 66%
yield (eq 2). The structure proposed for 8 is supported by
elemental analysis data as well as NMR spectroscopic (1H, 13C,
and 31P) data, with the exception that 13C NMR resonances
attributable to the CO groups were not observed, despite
prolonged data acquisition times. Moreover, while only two IR-
active CO vibrations are predicted for 8, three vibrations (2100,
2030, and 1990 cm-1) were detected for this complex. Although
definitive experimental data needed to unequivocally confirm
the connectivity in 8 and to rationalize such IR-spectroscopic
observations are currently lacking, it is possible that the third
IR band arises from the coexistence of a dimeric or other
oligomeric relative of 8 that possesses linear OC-M-CO
linkages.35 The absence of NMR evidence in support of multiple
Rh species in solution requires that the NMR spectroscopic
signature of such an oligomeric relative of 8 be identical to
that of 8 itself and/or that such monomeric and oligomeric
species must undergo rapid exchange on the NMR time scale.
i
ylene phosphide-type resonance structure featuring a Pr2Pd
C3 linkage and a formal anionic charge on phosphorus may
also figure prominently in 5c and 6c.32
Solubility Studies and Stoichiometric Reactivity with E-H
Bonds. The solubility of the discrete salts [5b]+X- and [6b]+X-,
and the neutral complexes 5c and 6c, in various organic media
was assessed, so as to determine the scope of solvents in which
such complexes may be utilized as catalysts.33 All complexes
were found to be fully soluble in either CH2Cl2 or THF with
the exception of [6b]+X-, which exhibited only limited solubil-
ity in the latter solvent. In contrast, while these cationic species
proved insoluble in all hydrocarbons, complexes 5c and 6c were
observed to dissolve fully in benzene or toluene, and even
proved soluble to a limited extent in hexanes or pentane.
The reactivity of the cationic and zwitterionic complexes
reported herein toward E-H bond containing small molecules
was surveyed prior to catalytic studies; unfortunately, such
reactions proved rather uninformative. Exposure of [5b]+X-,
[6b]+X-, 5c, or 6c to Et3SiH or Ph2SiH2 (1 or 10 equiv in THF)
at 24 or 60 °C led to the formation of a complex mixture of
products, including in some cases 3a and 3b (31P NMR).
Similarly, though exposure of 1 to an atmosphere of H2 is known
to generate the trinuclear cluster [Ir3H7(PCy3)3(Py)3]2+,2 treat-
ment of 6c in C6D6 with H2 produced an intractable dark brown
precipitate.
The preparation of the zwitterionic complex (CO)2Rh(κ2-P,N-
3c) was examined subsequently. Whereas the combination of
[3c]Li and 0.5 equiv of [(COD)RhCl]2 afforded 5c cleanly (vide
Pursuit of Rh Carbonyl Complexes. With an aim toward
utilizing infrared spectroscopic techniques to compare the
electron-donating abilities of the κ2-P,N-ligands 3b and 3c in
these cationic and zwitterionic complexes (respectively), we
sought to prepare carbonyl-containing relatives of [5b]+X- and
5c. In exploring the reactivity of 3a with 0.5 equiv of
[(CO)2RhCl]2 in THF, complete consumption of 3a was noted
after 10 min, along with the formation of multiple phosphorus-
containing products (31P NMR). However, subsequent treatment
of this reaction mixture with anhydrous K2CO3 resulted in the
clean formation of a single phosphorus-containing species (7,
supra), under similar conditions employing [(CO)2RhCl]2
a
complicated mixture of products was generated (31P NMR).
Similarly complex product distributions were observed upon
exposure of THF, CH2Cl2, or toluene solutions of 5c to an
atmosphere of CO, and all efforts to isolate Rh-containing
products from these reactions have thus far not been fruitful.
Treatment of [5b]+X- and 5c with PMePh2. In a further
effort to qualitatively assess the binding affinity of the neutral
-
κ2-P,N-3b ligand in [5b]+BF4 relative to the anionic κ2-P,N-
3c ligand in 5c, these Rh(I) complexes were each treated with
31P ) 72), which was isolated as an analytically pure yellow-
an excess of PMePh2. Addition of 10 equiv of PMePh2 to
δ
[5b]+BF4 resulted in the quantitative consumption of the
-
orange solid in 43% yield (eq 1). The crystallographic charac-
terization of 7 (see Supporting Information and Table 1) reveals
trans-disposed phosphorus and chloride ligands in this square-
planar complex, a bonding motif that has been observed in
related crystallographically characterized (κ2-P,N)Rh(CO)Cl
species.34
starting Rh(I) complex, with concomitant formation of [Rh-
(PMePh2)4]+,36 3a, and 3b as the only phosphorus-containing
products (eq 3). Whereas both the neutral κ2-P,N-3b ligand and
COD in [5b]+BF4 were displaced by PMePh2, under similar
-
experimental conditions employing the zwitterionic 5c, only the
COD ligand underwent substitution, affording (PMePh2)2Rh-
(κ2-P,N-3c), 9 (eq 4). Complex 9 was obtained subsequently as
an analytically pure light brown powder in 66% isolated yield.
In the absence of significant steric differences between κ2-P,N-
3b and κ2-P,N-3c (vide supra), and given the possible hemilabile
character of these P,N-ligands whereby the P-donor may serve
as an anchor to the group 9 metal,37 it is plausible that the
divergent reactivity of [5b]+BF4- and 5c with PMePh2 may be
attributable in part to the ease with which the N-donor can
dissociate from the formally cationic Rh(I) center, presuming
that such a process is of relevance to the reaction pathway
Initial attempts to prepare [(CO)2Rh(κ2-P,N-3b)]+PF6 (8)
-
via treatment of 3a with 0.5 equiv of [(CO)2RhCl]2 in the
(31) For crystallographically characterized indenide salts, see: (a) Boche,
G.; Ledig, B.; Marsch, M.; Harms, K. Acta Crystallogr. 2001, E57, m570.
(b) Marder, T. B.; Williams, I. D. Chem. Commun. 1987, 1478.
(32) Izod, K. Coord. Chem. ReV. 2002, 227, 153.
(33) A combination of 6 mg of complex and 0.8 mL of solvent was
utilized in these qualitative visual solubility tests. The solubility of [5b]+X-
and [6b]+X- complexes was not found to vary significantly for X ) SO3-
CF3, BF4, and PF6.
(34) For a recent example, see: Dennett, J. N. L.; Bierenstiel, M.;
Ferguson, M. J.; McDonald, R.; Cowie, M. Inorg. Chem. 2006, 45, 3705.
(35) Andrieu, J.; Camus, J.-M.; Richard, P.; Poli, R.; Gonslavi, L.; Vizza,
F.; Peruzzini, M. Eur. J. Inorg. Chem. 2006, 51.
(36) Schrock, R. R.; Osborn, J. A. J. Am. Chem. Soc. 1971, 93, 2397.
(37) Slone, C. S.; Weinberger, D. A.; Mirkin, C. A. Prog. Inorg. Chem.
1999, 48, 233.

Rh and Ir κ2-P,N Complexes
Organometallics, Vol. 26, No. 3, 2007 601
Table 3. Hydrogenation of Styrenea
leading to substitution of the P,N-ligand in these complexes. In
this vein, and consistent with the observed reluctance of the
κ2-P,N-3c ligand in 5c to undergo substitution by PMePh2, the
dissociation of a ligand arm in [5b]+ to give a [(COD)M(κ1-
P,N-3b)]+ intermediate is anticipated to be more facile than an
analogous dechelation process involving 5c, given the height-
ened attractive electrostatic forces that are likely to exist between
the formally anionic P,N-indenide ligand and the formally
cationic [(COD)M]+ fragment in 5c, relative to those in [5b]+.
The retention of the anionic ancillary ligand in 5c following
treatment with PMePh2 contrasts the previously documented loss
entry
catalyst
loading (mol %)
solvent
yield (%) t (h)b
-
3-1
3-2
3-3
3-4
3-5
3-6
3-7
3-8
[5b]+PF6-
1.0
1.0
5.0
1.0
1.0
5.0
5.0
5.0
0.5
0.5
0.5
0.5
0.5
5.0
0.5
0.5
0.5
CH2Cl2
THF
THF
CH2Cl2
THF
THF
10
97
24
24
24
24
24
24
4
24
0.5
4
4
4
4
24
4
4
4
[5b]+PF6-
[5b]+PF6
>99
5c
5c
5c
2
17
63
94
THF
>99
>99
>99
>99
>99
91
5c
benzene
CH2Cl2
CH2Cl2
THF
-
3-9
[6b]+PF6
3-10
3-11
3-12c
3-13
3-14
3-15
3-16
1
-
[6b]+PF6
-
of BPh4 from zwitterionic (η6-PhBPh3)RhL2 species upon
1
THF
exposure to phosphines.12
6c
6c
6c
6c
CH2Cl2
CH2Cl2
THF
benzene
hexanes
56
76
58
37
3-17c 6c
17
b
a Conditions: 24 °C and ∼1 atm H2. Time at which the quoted
conversion (yield) was achieved on the basis of GC-MS and GC-FID data,
c
rounded to the nearest percent. Catalyst was not completely soluble in this
solvent.
-
(I) cations [5b]+PF6 and [6b]+PF6-, with those of the
structurally related zwitterionic complexes 5c and 6c, the metal-
mediated hydrogenation of styrene (24 °C, ∼1 atm H2) was
examined; selected catalytic results are collected in Table 3.
The catalytic behavior of [5b]+PF6 and 5c was found to be
-
similar, with both of these Rh(I) catalysts exhibiting an induction
period40a that was not observed for 2 or the Ir(I) catalysts
examined in this reactivity survey (vide infra). While both
[5b]+PF6- and 5c exhibited only modest activity at the 1.0 mol
% catalyst loading level in CH2Cl2 (entries 3-1 and 3-4),
improved catalytic performance was noted in THF (entries 3-2
and 3-5), especially when the catalyst loading was increased to
5.0 mol % (entries 3-3 and 3-6). By comparison, Wilkinson’s
catalyst (2) proved superior to both [5b]+PF6- and 5c in THF,
affording ethylbenzene quantitatively after only 4 h (entry 3-7).
1
In the course of characterizing 9 at 300 K, the H and 13C-
{1H} NMR resonances associated with the PPh2, PiPr2, and
NMe2 fragments were found to be broadened significantly
relative to the indenide and the PMePh2 signals. Furthermore,
the observation of two unique P(CHMe2)2 resonances in both
the 1H and 13C{1H} NMR spectra at this temperature is
consistent with 9 adopting an effective C1-symmetric solution
structure on the NMR time scale, in contrast to the apparent Cs
symmetry exhibited by [5b]+X, [6b]+X-, 5c, and 6c. The H
1
NMR line shape changes that occur on going from 240 to 350 K,
including the apparent simplification of the PPh2 and PiPr2
regions of the spectrum as well as coalescence (Tc ) 298 K) of
the initially well-resolved NMe2 resonances at 240 K, are
consistent with a dynamic process involving restricted rotation
about the Rh-PMePh2 bonds in 9. Although signal overlap
Although the styrene conversions achieved by use of [5b]+PF6
-
in THF were consistently greater than those obtained by use of
5c as a catalyst under similar conditions, the latter zwitterionic
species provided quantitative styrene conversion in benzene
(entry 3-8), a solvent in which [5b]+PF6 is not soluble.
-
Whereas cationic square-planar Ir(I) complexes, including
Crabtree’s catalyst (1) and related [(COD)Ir(P,N)]+X- species,
are among the most active homogeneous catalysts for alkene
hydrogenation,5,22a effective neutral Ir(I) alkene hydrogenation
catalysts are still very rare. In this regard, we became interested
in benchmarking the catalytic abilities of the neutral and
zwitterionic complex 6c against those of the more traditional
within the PPh2 and PiPr2 regions of the H NMR spectrum
1
rendered these resonances too complex to interpret in terms of
line shape analysis, evaluation of the temperature-dependent 1H
NMR line shape changes for the NMe2 unit employing the
Gutowsky-Holm approximation38 yielded a value for ∆Gq
Tc
of ca. 14 kcal/mol. It is unclear whether the calculated ∆Gq
Tc
based on the coalescence of the NMe2 signals provides a
measure of the barrier associated with the proposed Rh-PMePh2
restricted rotation process, and/or the barrier to an alternative
dynamic process involving Rh-N dissociation, rotation about
the C2-NMe2 linkage, inversion at N, and recoordination to
Rh, which would also render equivalent the two N-Me environ-
ments in 9 if rapid on the NMR time scale.39 However, such a
process involving N-dissociation alone cannot account for the
temperature-dependent line shape changes that are observed
within the PPh2 and PiPr2 regions of the 1H NMR spectra of 9.
Alkene Hydrogenation Studies. In an initial effort to
compare and contrast the catalytic abilities of the Rh(I) and Ir-
cationic complexes [6b]+PF6 and 1.40b In keeping with the
-
well-documented affinity of [(COD)Ir(P,N)]+X- alkene hydro-
genation catalysts for chlorocarbon reaction media,2,22a the
reduction of styrene mediated by [6b]+PF6 or 1 (0.5 mol %)
-
occurred quantitatively in CH2Cl2 after 0.5 h or 4 h, respectively
(entries 3-9 and 3-10). The noteworthy catalytic activity
-
displayed by [6b]+PF6 in CH2Cl2 contrasts the apparent
deactivating effect that this solvent has on the related Rh(I)
species, [5b]+PF6- and 5c (vide supra). While [6b]+PF6- also
-
(40) (a) For styrene hydrogenation reactions employing either [5b]+PF6
or 5c as a catalyst, <1% conversion was observed after 0.5 h, and in the
case of reactions employing 5.0 mol % catalyst loadings, only 5-10%
conversion was achieved after 4 h. (b) In a preliminary reactivity survey,
(38) Gutowsky, H. S.; Holm, C. H. J. Chem. Phys. 1956, 25, 1228.
(39) A similar dynamic process has been proposed to account for the
temperature-dependent 1H NMR line shape changes observed for [(COD)M-
(κ2-1-P(S)iPr2-2-NMe2-indene)]+BF4- (M ) Rh, Ir), the analysis of which
also yielded ∆GqTc of ca. 14 kcal/mol: Wechsler, D.; Myers, A.; McDonald,
R.; Ferguson, M. J.; Stradiotto, M. Inorg. Chem. 2006, 45, 4562.
[6b]+PF6 proved to be the most effective alkene hydrogenation catalyst
-
of the three [6b]+X- salts reported herein (X ) SO3CF3, BF4, or PF6); see
ref 18c herein. (c) Initial rates (TOF) are defined as (moles ethylbenzene
produced/moles catalyst)/time and are measured at 0.25 h for reactions
employing 0.5 mol % catalyst.

602 Organometallics, Vol. 26, No. 3, 2007
Cipot et al.
Table 4. Addition of Triethylsilane to Styrenea
provides access to synthetically useful organosilicon com-
pounds,41 metal-catalyzed alkene hydrosilylation can serve as
a prototype for E-H addition reactions in which product
selectivity can be a challenge. In contrast to the relative
simplicity of alkene hydrogenation, the metal-mediated addition
of silanes to substituted olefins can generate a range of
organosilicon species; the predominant products of this reaction
in the case of triethylsilane and styrene are depicted in eq 5.
Beyond the two possible regioisomers that can arise from simple
Si-H addition (10a and 10c), vinylsilanes formed via dehy-
drogenative silylation (10b, 10d, and 10e) are also commonly
observed. Notably, 10b is often generated as the major product
in such metal-catalyzed reactions where an excess of styrene is
employed as a sacrificial H2 acceptor.41,42
entry
catalyst
solvent yield (%)b 10ac 10bc 10cc otherc
-
-
-
4-1
4-2
4-3
4-4
4-5
4-6
4-7
4-8
4-9d
4-10
4-11
[5b]+PF6
5c
DCE
DCE
DCE
THF
THF
toluene
DCE
DCE
DCE
DCE
<1
<1
>99
37
>99
98
66
>99
86
>99
>99
35
98
88
<1
<1
4
5
10
2
66
99
86
51
91
35
94
88
55
<1
<1
93
32
90
96
<1
<1
<1
40
1
<1
<1
3
<1
<1
<1
<1
<1
<1
7
<1
<1
<1
<1
<1
<1
<1
<1
<1
<1
<1
<1
2
2
[5b]+PF6
5c
5c
[6b]+PF6
6c
6c
1
-
[6b]+PF6- THF
8
4-12d [6b]+PF6
4-13 6c
4-14d 6c
4-15 6c
THF
THF
THF
toluene
<1
2
<1
2
<1
<1
<1
<1
<1
7
64
In keeping with the poor alkene hydrogenation activity noted
for [5b]+PF6 and 5c in CH2Cl2, no hydrosilylation chemistry
-
a Conditions: 60 °C (except where noted); 5.0 mol % catalyst; styrene-
to-silane ratio of 5:1; DCE ) 1,2-dichloroethane; control experiments
confirmed that solutions of [5b]+PF6-, [6b]+PF6-, 5c, or 6c are each stable
was observed for reactions conducted in 1,2-dichloroethane
employing either of these Rh(I) catalyst complexes (entries 4-1
and 4-2). However, the ability of Rh(I) species to mediate alkene
hydrosilylations under these conditions was confirmed through
the use of 2. Complete consumption of the silane was achieved
after 4 h in 1,2-dichloroethane, along with excellent selectivity
for the vinylsilane 10b (entry 4-3); similar results were obtained
in benzene. While upon changing the solvent to THF only
modest catalytic productivity was exhibited by [5b]+PF6- (37%,
entry 4-4), quantitative consumption of the silane was achieved
under similar conditions employing 5c, affording 10b along with
the anti-Markovnikov addition product 10a in a 90:10 ratio
(entry 4-5). Similarly high conversions and even greater
selectivity were achieved by use of 5c as a catalyst in toluene
(entry 4-6), a solvent in which the structurally related cationic
b
upon heating at 60 °C for a minimum of 24 h (31P NMR). Based on the
consumption of triethylsilane at 24 h, except for entry 4-3, where the yield
c
is quoted after 4 h. Product distribution based on GC-MS and GC-FID
data, rounded to the nearest percent; other silicon-containing products
d
including 10d and 10e. Reaction conducted at 24 °C.
outperformed both [5b]+PF6 and 5c in THF, affording ethyl-
-
benzene quantitatively after only 4 h at the 0.5 mol % catalyst
loading level (entry 3-11), incomplete conversion was noted
for analogous reactions employing 1 as a catalyst in THF, even
after 24 h (entry 3-12). Unlike [5b]+PF6 and 5c, which
-
exhibited comparable catalytic abilities, the zwitterionic 6c
proved to be a considerably less effective styrene hydrogenation
-
catalyst than both [6b]+PF6 and 1. When employing 0.5 mol
complex [5b]+PF6 is not soluble.
-
% 6c as a catalyst in CH2Cl2, a maximum conversion of 56%
was achieved after 4 h (entry 3-13), and only 76% conversion
was realized by use of 5.0 mol % 6c after a 24 h period (entry
3-14). However, given that reasonable initial styrene hydrogena-
tion rates are observed for 6c (355 h-1; cf. 740 h-1 for
[6b]+PF6-),40c it is feasible that the lower styrene conversions
obtained by use of 6c as a catalyst, as compared to reactions
employing [6b]+PF6-, may reflect the instability of 6c under
these experimental conditions, rather than the inherent catalytic
activity of this zwitterionic species. In exploring further the
effect of solvent on the catalytic performance of 6c, the
production of ethylbenzene in THF (entry 3-15) was found to
be similar to that obtained in CH2Cl2 at the same catalyst
loading. In addition, while both [6b]+PF6- and 1 are not soluble
in hydrocarbons, complex 6c exhibited modest catalytic activity
in benzene (entry 3-16), as well as in hexanes (entry 3-17). Other
In contrast to some of the other platinum-group metals, the
use of Ir complexes as alkene hydrosilylation catalysts has
received relatively little attention.43 However, evidence for the
selectivity advantages that can be brought about by use of Ir
catalysts in such transformations is provided in a report by
Uemura and co-workers.44 These workers observed that whereas
analogous chiral κ2-P,N Rh and Ir species both proved capable
of mediating the asymmetric hydrosilylation of acetophenone
in nearly quantitative yield and with high enantioselectivity, the
-
differences in the catalytic abilities of [6b]+PF6 and 6c were
brought to light through the study of more highly substituted
alkene substrates (0.5 mol % catalyst, 24 °C, ∼1 atm H2).
Whereas the cationic Ir(I) complex [6b]+PF6- in CH2Cl2 proved
capable of reducing cyclohexene quantitatively (0.5 h) and
1-methylcyclohexene to a significant extent (4 h, 83%), only
26% reduction of cyclohexene was achieved by use of 6c as a
catalyst after 4 h, and no zwitterion-mediated reduction of the
aforementioned trisubstituted alkene was observed in CH2Cl2
or benzene.
Styrene Hydrosilylation Studies. In continuing our head-
to-head reactivity comparison of the structurally related cationic
and zwitterionic Rh(I) and Ir(I) species reported herein, the
catalytic addition of triethylsilane to styrene mediated by
[5b]+PF6-, [6b]+PF6-, 5c, or 6c was chosen as a second test
reaction (5.0 mol % catalyst, 60 °C, excess styrene); selected
experimental results are provided in Table 4. As well as
representing a widely employed chemical transformation that
(41) For selected reviews, see: (a) Marciniec, B. Coord. Chem. ReV.
2005, 249, 2374. (b) Ojima, I.; Li, Z.; Zhu, J. In Chemistry of Organic
Silicon Compounds; Rappoport, Z., Apeloig, Y., Eds.; Wiley: New York,
1998; Vol. 2 p 1687. (c) Marciniec, B.; Gulin´ski, J. J. Organomet. Chem.
1993, 446, 15. (d) Hiyama, T.; Kusumoto, T. In ComprehensiVe Organic
Synthesis; Trost, B. M., Fleming, I., Eds.; Pergamon Press: Oxford, 1991;
Vol 8, p 763. (e) Speier, J. L. AdV. Organomet. Chem. 1979, 17, 407.
(42) For examples featuring Rh(I) catalysts, see: (a) Takeuchi, R.; Yasue,
H. Organometallics 1996, 15, 2098, and references therein. (b) Kakiuchi,
F.; Nogami, K.; Chatani, N.; Seki, Y.; Murai, S. Organometallics 1993,
12, 4748. (c) Onopchenko, A.; Sabourin, E. T.; Beach, D. L. J. Org. Chem.
1984, 49, 3389. (d) Onopchenko, A.; Sabourin, E. T.; Beach, D. L. J. Org.
Chem. 1983, 48, 5101.
(43) For examples of alkene hydrosilylation mediated by neutral Ir(I)
complexes, see: (a) Oro, L. A.; Fernandez, M. J.; Esteruelas, M. A.;
Jimenez, M. S. J. Mol. Catal. 1986, 37, 151. (b) Apple, D. C.; Brady, K.
A.; Chance, J. M.; Heard, N. E.; Nile, T. A. J. Mol. Catal. 1985, 29, 55.
(44) Nishibayashi, Y.; Segawa, K.; Ohe, K.; Uemura, S. Organometallics
1995, 14, 5486.

Rh and Ir κ2-P,N Complexes
Organometallics, Vol. 26, No. 3, 2007 603
chiral alcohol products obtained upon workup from the respec-
tive reactions were found to possess different absolute configu-
rations.
suggest that alterations to the ancillary P,N-ligand architecture
in 6c may provide access to more active and long-lived Ir(I)
hydrogenation catalysts that are hydrocarbon-soluble. Although
-
-
[6b]+PF6 proved superior to both [5b]+PF6 and 5c in terms
of styrene hydrogenation activity in CH2Cl2 or THF, high
conversion to ethylbenzene was achieved by use of 5c as a
catalyst in benzene, in which [5b]+PF6- and [6b]+PF6- are not
soluble. The catalytic utility of these zwitterionic Rh(I) and Ir-
(I) species was also demonstrated in hydrosilylation reactions
employing triethylsilane and styrene. Remarkably, while reac-
tions mediated by either 5c (toluene or THF) or 6c (1,2-
dichloroethane or THF) were observed to occur in nearly
quantitative yield and with high selectivity, a different orga-
nosilane product was obtained in each case; the use of 5c as a
catalyst afforded E-1-triethylsilyl-2-phenylethene (10b) selec-
tively, whereas 1-triethylsilyl-2-phenylethane (10a) was the only
organosilane product detected in reactions employing 6c as a
catalyst in 1,2-dichloroethane. The complete selectivity for 10a
exhibited by this zwitterionic Ir(I) complex in the presence of
excess styrene is noteworthy, especially in light of the observa-
tion that a nearly equimolar ratio of 10a and 10b was generated
in reactions employing Crabtree’s catalyst (1) under similar
The lack of catalytic activity observed for both [5b]+PF6
-
and 5c in 1,2-dichloroethane contrasts the substantial silane
conversion (66%) and complete selectivity for 10a that was
achieved by use of catalytic amounts of the Ir(I) cation
[6b]+PF6- in this solvent (entry 4-7). Even more remarkable is
the catalytic performance exhibited by the zwitterionic species
6c, which under analogous conditions afforded 10a quantita-
tively, despite the 4-fold excess of styrene employed (entry 4-8).
High yields of 10a were also obtained by use of 6c as a catalyst
in 1,2-dichloroethane on lowering the reaction temperature from
60 °C to 24 °C, with no loss in selectivity (entry 4-9). While
hydrosilylation reactions mediated by Crabtree’s catalyst (1) at
60 °C were also quantitative, very poor selectivity was observed
(entry 4-10).
In examining the effect of solvent on the catalytic behavior
of [6b]+PF6- and 6c, hydrosilylation reactions were conducted
in THF. Nearly quantitative silane conversions and good
selectivity for 10a were achieved by use of [6b]+PF6- or 6c in
this solvent at 60 °C (entries 4-11 and 4-13). While on lowering
the temperature to 24 °C both of these Ir(I) catalyst complexes
displayed complete selectivity for 10a in THF, the zwitterionic
6c outperformed the analogous cation in terms of yield (entries
4-12 and 4-14). In contrast to the excellent catalytic performance
that was exhibited by the zwitterionic Rh(I) complex 5c in
toluene (entry 4-6), the Ir(I) analogue 6c proved much less
effective in this solvent relative to the others surveyed (entry
4-15). The catalytic abilities of the discrete salts [6b]+PF6- and
1 in toluene could not be surveyed, due to the poor solubility
of these complexes in hydrocarbon media.
conditions.
On the basis of the results of this catalytic survey, [5b]+PF6
,
-
[6b]+PF6-, 5c, 6c, and their derivatives constitute a comple-
mentary family of catalyst complexes for the addition of E-H
bonds to alkenes, when issues of catalytic activity/selectivity
and solvent compatibility are considered. Notably, these catalytic
studies also confirm that zwitterionic species such as 5c and 6c
represent an effective class of neutral catalyst complexes for
the addition of E-H bonds to unsaturated substrates, whose
solubility and reactivity properties are in some cases divergent
from those of more traditional Rh(I) and Ir(I) [(COD)M(κ2-
P,N)]+X- salts. We are presently uncertain as to the specific
origins of the differing reactivity properties of structurally related
cationic group 9 complexes supported by κ2-P,N-3b and their
zwitterionic relatives featuring κ2-P,N-3c. The observation that
Summary and Conclusions
The versatility of 1-PiPr2-2-NMe2-indene (3a) has been
demonstrated through the use of this ligand in the synthesis of
neutral, cationic, and zwitterionic Rh(I) and Ir(I) complexes.
Whereas the facile rearrangement of 3a to 3-PiPr2-2-NMe2-
indene (3b) within the coordination sphere of these group 9
metals presented a challenge with regard to the isolation of κ2-
P,N-3a derivatives ([5a]+X- and [6a]+X-) in high yield, such
an isomerization process has been exploited successfully in the
preparation of κ2-P,N-3b complexes ([5b]+X- and [6b]+X-).
Zwitterionic complexes 5c and 6c featuring κ2-3-PiPr2-2-NMe2-
indenide (κ2-P,N-3c) ligands were prepared in high isolated
yield, either upon extrusion of HX from [5b]+X- or [6b]+X-
or by addition of [3c]Li to 0.5 equiv of [(COD)MCl]2. We view
5c and 6c as constituting an unusual new class of formally
charge-separated Rh(I) and Ir(I) species that are distinct from
more conventional η5-indenyl group 9 complexes.
Data obtained from X-ray crystallographic experiments
revealed that the steric characteristics of the metal coordination
environments in these cationic κ2-P,N-3b and zwitterionic κ2-
P,N-3c group 9 complexes are strikingly similar. Despite such
apparent similarities, divergent catalytic behavior both within
and between the cation/zwitterion pairs [5b]+X-/5c and [6b]+X-/
6c was observed. Among these four complexes, the Ir(I) cation
[6b]+PF6- proved to be the most effective styrene hydrogenation
catalyst in CH2Cl2 or THF, exhibiting catalytic activity rivaling
that of Crabtree’s catalyst (1) under similar conditions. While
the catalytic performance of 6c proved vastly inferior to that of
[6b]+PF6- in these solvents, the initial hydrogenation rates noted
for 6c in CH2Cl2, as well as the modest catalytic activity
exhibited by this zwitterionic species in hydrocarbon media,
-
[5b]+BF4- was cleanly transformed into [Rh(PMePh2)4]+BF4
upon exposure to excess PMePh2, while under similar conditions
only the COD ligand in 5c was displaced, suggests that tighter
binding of the anionic κ2-P,N-3c ligand to the [(COD)M]+
fragment in 5c and 6c may represent a potential source of such
divergent reactivity. In the absence of complete formal charge
separation within 5c and 6c, the electrophilicity of the associated
metal center in each of these complexes should be attenuated
relative to [5b]+PF6- and [6b]+PF6-, which in turn could also
influence the catalytic behavior displayed by these species. As
well, the participation of the backbone indenide unit in these
zwitterionic species during the course of catalytic transforma-
tions cannot be discounted, and we are intrigued by the
possibility of cooperative metal-ligand bifunctional activation
of substrate E-H bonds involving formally zwitterionic deriva-
tives of κ2-P,N-3c. Synthetic and mechanistic experiments
directed toward addressing these issues are currently underway
in our laboratory.45
Experimental Section
General Considerations. Except where noted, all manipulations
were conducted in the absence of oxygen and water under an
atmosphere of dinitrogen, either by use of standard Schlenk methods
or within an mBraun glovebox apparatus, utilizing glassware that
(45) Compounds 3a, [5b]+PF6-, 5c, [6b]+PF6-, and 6c are commercially
available from Strem Chemicals, Inc.

604 Organometallics, Vol. 26, No. 3, 2007
Cipot et al.
3
was oven-dried (130 °C) and evacuated while hot prior to use. Celite
(Aldrich) was oven-dried (130 °C) for 5 days and then evacuated
for 24 h prior to use. The nondeuterated solvents tetrahydrofuran,
dichloromethane, diethyl ether, toluene, benzene, hexanes, and
pentane were deoxygenated and dried by sparging with dinitrogen
gas, followed by passage through a double-column solvent purifica-
tion system purchased from mBraun Inc. Tetrahydrofuran, dichlo-
romethane, and diethyl ether were purified over two alumina-packed
columns, while toluene, benzene, hexanes, and pentane were
purified over one alumina-packed column and one column packed
with copper-Q5 reactant. The solvents used within the glovebox
were stored over activated 3 Å molecular sieves. Both CD2Cl2
(Cambridge Isotopes) and CDCl3 (Aldrich) were degassed by using
three repeated freeze-pump-thaw cycles, dried over CaH2 for 7
days, distilled in vacuo, and stored over 3 Å molecular sieves for
24 h prior to use. Purification of NEt3 was achieved by stirring
over KOH for 7 days, followed by distillation; the distilled NEt3
was then refluxed over CaH2 for 3 days under dinitrogen, followed
by distillation. All silver salts (Aldrich) and anhydrous K2CO3
(Aldrich) were dried in vacuo for 12 h prior to use. Hydrogen
(99.999%, UHP grade) and carbon monoxide gases (99.5%,
chemically pure grade) were obtained from Air Liquide and were
used as received. All other liquid solvents or reagents (Aldrich)
were degassed by using three repeated freeze-pump-thaw cycles
and then dried over 3 Å molecular sieves for 24 h prior to use.
Compounds 3a,18d [3c]Li,18d [(COD)RhCl]2,46a [(COD)IrCl]2,46b
[(CO)2RhCl]2,47 and (PPh3)3RhCl (2)48 were prepared using litera-
ture methods and dried in vacuo for 24 h prior to use, while
[(COD)Ir(PCy3)(Py)]+PF6- (1) was obtained from Strem (COD )
η4-1,5-cyclooctadiene). All 1H, 13C, and 31P NMR characterization
data were collected at 300 K on either a Bruker AC-250
spectrometer operating at 250.1, 62.9, and 101.3 MHz, for 1H, 13C,
and 31P (respectively) or a Bruker AV-500 spectrometer operating
at 500.1, 125.8, and 202.5 MHz (respectively) with chemical shifts
NMR (CDCl3, 250.1 MHz): δ 8.90 (d, JHH ) 7.3 Hz, 1H, Ar-
H), 7.20-7.04 (m, 3H, Ar-Hs), 5.78 (s, 1H, C3-H), 5.47-5.34
(m, 3H, COD and C1-H), 3.74 (br m, 2H, COD), 3.00 (m, 1H,
P(CHMe2)), 2.73 (s, 6H, NMe2), 2.38 (m, 5H, COD and P(CHMe2)),
1.96 (m, 4H, COD), 1.70 (d of d, 3JPH ) 15.8 Hz, 3JHH ) 6.8 Hz,
3
3
3H, P(CHMeMe)), 1.32 (d of d, JPH ) 14.1 Hz, JHH ) 7.3 Hz,
3H, P(CHMeMe)), 1.02 (d of d, JPH ) 11.5 Hz, JHH ) 7.3 Hz,
3
3
3
3
3H, P(CHMeMe)), 0.44 (d of d, JPH ) 13.7 Hz, JHH ) 7.3 Hz,
3H, P(CHMeMe)). 13C{1H} NMR (CDCl3, 62.9 MHz): δ 161.7
(d, J ) 9 Hz, C2), 144.9 (d, J ) 4 Hz, C3a or C7a), 139.7 (C7a or
C3a), 126.7 (Ar-C), 126.5 (Ar-C), 121.6 (2Ar-Cs), 118.3 (C3),
108.5 (COD), 101.0 (m, COD), 68.6 (m, COD), 43.8 (NMe2), 43.2
(m, C1), 33.2 (COD), 33.1 (COD), 28.3 (COD), 28.2 (COD), 24.0
(m, PCHMe2), 23.0 (d, J ) 16 Hz, PCHMe2), 21.9 (d, J ) 15 Hz,
PCHMeMe), 20.1 (PCHMeMe), 19.3 (PCHMeMe), 18.0 (m,
PCHMeMe). 31P{1H} NMR (CDCl3, 101.3 MHz): δ 48.2 (d, 1JRhP
) 144 Hz). Crystals of 4a‚CH2Cl2 suitable for use in a single-
crystal X-ray diffraction experiment were grown from a concen-
trated CH2Cl2 solution stored at -30 °C.
Synthesis of [(COD)Rh(κ2-P,N-3a)]+PF6 ([5a]+PF6-). To a
-
solution of [(COD)RhCl]2 (0.120 g, 0.244 mmol) in THF (2 mL)
was added a solution of AgPF6 (0.123 g, 0.487 mmol) in THF (2
mL) via Pasteur pipet. The vial was sealed with a PTFE-lined cap
and the mixture shaken vigorously for ∼30 s; the immediate
formation of a white precipitate was observed. After 5 min, a
solution of 3a (0.134 g, 0.487 mmol) in THF (3 mL) was added,
producing a yellow-orange solution. The reaction vial was sealed
and shaken vigorously for 30 s, after which the precipitate was
allowed to settle over 15 min. The reaction mixture was filtered
through Celite to remove the precipitated solids, which were
discarded. From the initially homogeneous filtrate a yellow micro-
crystalline precipitate was obtained. This solid was isolated by
decanting away the supernatant (which was found to contain a
mixture of [5a]+PF6 and [5b]+PF6 as identified by 31P NMR
analysis) and dried in vacuo to leave [5a]+PF6- as an analytically
pure, orange powder (0.041 g, 0.066 mmol, 14%). Anal. Calcd for
C25H38P2N1Rh1F6 (631.4 gmol-1): C 47.55; H 6.07; N 2.22.
Found: C 47.70; H 6.06; N 2.23. 1H NMR (CD2Cl2, 500.1 MHz):
δ 7.36-7.29 (m, 3H, Ar-Hs), 7.24 (m, 1H, Ar-H), 6.37 (br s,
1H, C3-H), 5.22 (br s, 1H, COD), 4.94 (br s, 1H, COD), 4.42 (d,
2JPH ) 12.0 Hz, 1H, C1-H), 4.32 (br s, 1H, COD), 4.15 (br s, 1H,
COD), 3.25 (s, 3H, NMeaMeb), 2.79 (s, 3H, NMeaMeb), 2.62-2.37
(m, 4H, COD), 2.32-2.15 (m, 3H, COD and P(CHMeaMeb)(CH-
MecMed)), 2.14-2.01 (m, 3H, COD and P(CHMeaMeb)(CHMec-
-
-
1
reported in parts per million downfield of SiMe4 (for H and 13C)
1
or 85% H3PO4 in D2O (for 31P). In some cases, H and 13C NMR
chemical shift assignments are made on the basis of data obtained
from 13C-DEPT, 1H-1H COSY, 1H-13C HSQC, 1H-13C HMBC,
and/or 1D 1H NOE NMR experiments. Variable-temperature NMR
studies involving 7 were conducted on a Bruker AC-250 spectrom-
eter, with temperature calibrations carried out using an external
MeOH/MeOD standard.49 GC-FID analyses were performed on a
Perkin-Elmer autosystem gas chromatograph, while GC-MS analy-
ses were performed on a Varian Saturn-2000 system. All GC
analyses were carried out using a 30 m × 0.53 mm J&W DB5
column, with helium as the carrier gas. Elemental analyses were
performed either by Desert Analytics, Tucson, AZ, or by Canadian
Microanalytical Service Ltd., Delta, British Columbia (Canada).
Synthesis of (COD)RhCl(κ1-P,N-3a) (4a). A solution of 3a
(0.348 g, 1.26 mmol) in CH2Cl2 (3 mL) was added via Pasteur
pipet over 1 min to a glass vial containing a small magnetic stir
bar and a light orange solution of [(COD)RhCl]2 (0.312 g, 0.633
mmol) in CH2Cl2 (3 mL). The vial was then sealed with a PTFE-
lined cap, and the resulting dark orange solution was magnetically
stirred for 1 h. 31P NMR data collected on an aliquot of this crude
reaction solution indicated the quantitative formation of 4a. The
reaction mixture was then filtered through Celite and the CH2Cl2
solvent and other volatile materials were removed in vacuo, yielding
4a as an analytically pure, bright yellow powder (0.633 g, 1.21
mmol, 96%). Anal. Calcd for C25H38N1P1Cl1Rh1 (521.91 gmol-1):
3
3
Med)), 1.71 (d of d, JPH ) 17.8 Hz, JHH ) 7.2 Hz, 3H,
P(CHMeaMeb)(CHMecMed)), 1.50 (d of d, 3JPH ) 12.5 Hz, 3JHH
)
6.9 Hz, 3H, P(CHMeaMeb)(CHMecMed)), 1.08 (d of d, 3JPH ) 15.0
Hz, 3JHH ) 7.4 Hz, 3H, P(CHMeaMeb)(CHMecMed)), 0.82 (d of d,
3JPH ) 17.8 Hz, 3JHH ) 6.9 Hz, 3H, P(CHMeaMeb)(CHMecMed)).
13C{1H} NMR (CD2Cl2, 125.8 MHz): δ 165.1 (C2), 143.2 (C3a
or C7a), 134.9 (d, J ) 6.7 Hz, C7a or C3a), 128.9, 126.4, 125.3,
122.9 (C4, C5, C6, and C7), 116.7 (d, J ) 9.4 Hz, C3), 108.4
(COD), 105.1 (COD), 76.0 (d, JRhC ) 12.8 Hz, COD), 72.1 (d,
JRhC ) 12.4 Hz, COD), 51.7 (NMeaMeb), 48.1 (NMeaMeb), 42.9
1
(d, JPC ) 10.8 Hz, C1), 33.2 (COD), 32.0 (COD), 28.6 (COD),
28.3 (COD), 25.7 (d, 1JPC ) 19.4 Hz, P(CHMeaMeb)(CHMecMed)),
23.8 (d, 1JPC ) 17.0, P(CHMeaMeb)(CHMecMed)), 20.5 (d, 2JPC
)
2
4.8 Hz, P(CHMeaMeb)(CHMecMed)), 19.1 (d, JPC ) 4.2 Hz,
P(CHMeaMeb)(CHMecMed)), 18.7 (d, 2JPC ) 6.1 Hz, P(CHMeaMe-
b)(CHMecMed)), 18.4 (P(CHMeaMeb)(CHMecMed)). 31P{1H} NMR
(CD2Cl2, 202.5 MHz): δ 25.7 (d, 1JRhP ) 147.0 Hz, P(CHMe2)2),
1
C 57.53, H 7.34, N 2.68. Found: C 57.18, H 7.22, N 2.39. H
1
-143.2 (septet, JPF ) 709.3 Hz, PF6).
(46) (a) Giordano, G.; Crabtree, R. H. Inorg. Synth. 1990, 28, 88. (b)
Herde, J. L.; Lambert, J. C.; Senoff, C. V. Inorg. Synth. 1974, 15, 18.
(47) Malbosc, F.; Chauby, V.; Berre, S-L; Etienne, M.; Daran, J.-C.;
Kalck, P. Eur. J. Inorg. Chem. 2001, 2689.
(48) Osborn, J. A.; Wilkinson, G. Inorg. Synth. 1990, 28, 77.
(49) Braun, S.; Kalinowski, H.-O.; Berger, S. 150 and More Basic NMR
Experiments; Wiley-VCH: Toronto, 1998; p 136.
Synthesis of [(COD)Rh(κ2-P,N-3b)]+SO3CF3- ([5b]+SO3CF3-).
To a solution of [(COD)RhCl]2 (0.048 g, 0.098 mmol) in THF (1
mL) was added via Pasteur pipet a solution of AgSO3CF3 (0.050
g, 0.20 mmol) in THF (2 mL). The vial was sealed with a PTFE-
lined cap and the mixture shaken vigorously for ∼30 s. A white

Rh and Ir κ2-P,N Complexes
Organometallics, Vol. 26, No. 3, 2007 605
-
-
precipitate formed immediately, which was allowed to settle over
15 min and was removed from the reaction mixture by filtration
over Celite and discarded. A solution of 3a (0.054 g, 0.20 mmol)
in THF (2 mL) was added to the filtrate, with immediate formation
of an orange solution. The reaction vial was sealed, manually shaken
for 30 s, and then left to stand for 2.5 h. 31P NMR data collected
on an aliquot of the reaction mixture at this stage revealed a ∼ 90:10
isomeric mixture of [5a]+SO3CF3- and [5b]+SO3CF3-. To facilitate
the rearrangement of [5a]+SO3CF3- to [5b]+SO3CF3-, NEt3 (0.08
mL) was added to the reaction mixture. The vial was sealed and
manually shaken for 10 s. After 5 min, the solvent and other volatile
materials were removed in vacuo to leave an orange solid. This
solid was dried in vacuo, affording [5b]+SO3CF3- as an analytically
pure, orange powder (0.11 g, 0.17 mmol, 87%). Anal. Calcd for
C26H38P1N1S1O3Rh1F3 (619.55 gmol-1): C 49.12, H 6.03, N 2.20.
Found: C 48.98, H 6.20, N 2.25. 1H NMR (CD2Cl2, 500.1 MHz):
δ 7.53 (m, 1H, Ar-H), 7.49 (m, 1H, Ar-H), 7.39-7.36 (m, 2H,
Ar-H), 5.14 (br s, 2H, COD), 4.49 (br s, 2H, COD), 3.72 (s, 2H,
C1(H)2), 3.03 (s, 6H, NMe2), 2.83 (m, 2H, P(CHMeaMeb)2), 2.54-
the presence of an isomeric mixture of [5a]+PF6 and [5b]+PF6
.
To the isolated supernatant was added NEt3 (0.070 mL) to facilitate
the rearrangement of [5a]+PF6- to [5b]+PF6-. The vial was sealed
and manually shaken for 10 s, and after standing for an additional
5 min, the solvent and other volatile materials were removed in
vacuo to leave a sticky dark yellow-orange solid. The solid was
washed with toluene (2 × 1 mL) and subsequently dried in vacuo
-
to leave [5b]+PF6 as an analytically pure, orange powder (0.189
g, 0.300 mmol, 62%). Anal. Calcd for C25H38P2N1Rh1F6 (631.4
gmol-1): C 47.55; H 6.07; N 2.22. Found: C 47.49; H 6.02; N
2.40. All 1H, 13C{1H}, and 31P{1H} NMR data for the cation were
identical to those reported for [5b]+SO3CF3-. NB: If precipitated
-
[5a]+PF6 is not separated from the supernatant in the aforemen-
-
tioned procedure, analytically pure [5b]+PF6 is obtained in 73%
isolated yield.
Synthesis of (COD)Rh(κ2-P,N-3c) (5c). A solution of [(COD)-
RhCl]2 (0.044 g, 0.090 mmol) in toluene (3 mL) was transferred
dropwise via Pasteur pipet to a glass vial containing a magnetically
stirred slurry of [3c]Li (0.050 g, 0.18 mmol) suspended in toluene
(3 mL). The reaction mixture immediately turned deep orange-
red, accompanied by the formation of a fine precipitate. The vial
was then sealed with a PTFE-lined cap and the solution was
magnetically stirred for 45 min, after which time 31P NMR data
collected on an aliquot of the crude reaction mixture indicated the
quantitative formation of 5c. The reaction mixture was then filtered
through Celite, the supernatant was concentrated in vacuo, and the
product was crystallized from the supernatant by storing at -30 °C.
After 24 h, crystals of 5c were isolated by transferring the
supernatant solution to a new glass vial by using a Pasteur pipet;
this solution was then concentrated in vacuo and stored at -30 °C
to induce further crystallization. After repeating this procedure, the
isolated crops of crystals were combined and dried in vacuo,
yielding 5c as an analytically pure, orange-red microcrystalline solid
(0.079 g, 0.16 mmol, 89%). Anal. Calcd for C25H37N1P1Rh1 (485.45
gmol-1): C 61.85, H 7.68, N 2.89. Found: C 61.68, H 7.62, N
3
2.49 (m, 4H, COD), 2.26-2.18 (m, 4H, COD), 1.45 (d of d, JPH
) 17.1 Hz, 3JHH ) 7.2 Hz, 6H, P(CHMeaMeb)2), 1.35 (d of d, 3JPH
3
) 15.9 Hz, JHH ) 7.0 Hz, 6H, P(CHMeaMeb)2). 13C{1H} NMR
(CD2Cl2, 125.8 MHz): δ 195.9 (SO3CF3), 177.3 (d, J ) 17.9 Hz,
1
C2), 143.5 (C3a or C7a), 137.7 (C7a or C3a), 128.8 (d, JPC
)
26.3, C3), 127.1, 126.9, 125.3, 123.0 (C4, C5, C6, or C7), 105.9
1
(m, COD), 73.3 (d, JRhC ) 12.4 Hz, COD), 51.1 (NMe2), 32.3
1
(COD), 31.2 (d, J ) 10.1 Hz, C1), 28.1 (COD), 24.9 (d, JPC
)
24.5 Hz, P(CHMeaMeb)2), 19.4 (P(CHMeaMeb)2), 18.5 (d, J ) 3.8
Hz, P(CHMeaMeb)2). 31P{1H} NMR (CD2Cl2, 202.5 MHz): δ 43.1
1
(d, JRhP ) 150.6 Hz).
Synthesis of [(COD)Rh(κ2-P,N-3b)]+BF4 ([5b]+BF4-). A
-
solution of 3a (0.20 g, 0.72 mmol) in THF (2 mL) was added via
Pasteur pipet to a glass vial containing a small magnetic stir bar
and a light orange solution of [(COD)RhCl]2 (0.18 g, 0.36 mmol)
in THF (2 mL), thereby producing a dark orange solution. The vial
was then sealed with a PTFE-lined cap, and the solution was
magnetically stirred for 5 min. The cap was then removed and a
solution of AgBF4 (0.14 g, 0.72 mmol) in THF (2 mL) was added
to the reaction mixture over 1 min by use of a Pasteur pipet; a
precipitate formed immediately. The vial was again sealed with
the PTFE-lined cap and the reaction mixture magnetically stirred
for 1 h. 31P NMR data collected on an aliquot of this crude reaction
mixture indicated the quantitative formation of [5b]+BF4-. The
reaction mixture was filtered through Celite and the THF solvent
and other volatile materials were removed in vacuo, yielding
1
2.76. H NMR (C6D6, 250.1 MHz): δ 7.97 (m, 1H, Ar-H), 7.75
(m, 1H, Ar-H), 7.32 (m, 2H, Ar-Hs), 6.31 (d, J ) 3.7 Hz, 1H,
C1-H), 4.23 (m, 2H, COD), 3.95 (m, 2H, COD), 2.66 (s, 6H,
NMe2), 2.48 (m, 2H, P(CHMe2)2), 2.09-1.45 (m, 8H, COD), 1.17
(d of d, 3JPH ) 14.6 Hz, 3JHH ) 7.3 Hz, 6H, P(CHMeaMeb)2), 1.05
(d of d, 3JPH ) 15.9 Hz, 3JHH ) 7.3 Hz, 6H, P(CHMeaMeb)2). 13C-
{1H} NMR (C6D6, 62.9 MHz): δ 181.1 (d, J ) 10 Hz, C2), 163.1
(m, C3, C3a, or C7a), 138.8 (m, C3, C3a, or C7a), 121.1, 120.0,
116.9, 116.6 (C4, C5, C6, C7), 99.8 (m, COD), 87.4 (d, J ) 12
Hz, C1), 69.1 (d, J ) 13 Hz, COD), 54.2 (NMe2), 32.6 (m, COD),
28.3 (COD), 26.4 (d, J ) 26 Hz, P(CHMe2)2), 19.7 (P(CHMea-
Meb)2), 19.5 (d, J ) 4 Hz, P(CHMeaMeb)2). 31P{1H} NMR (C6D6,
-
analytically pure [5b]+BF4 as a bright orange powder (0.39 g,
0.68 mmol, 94%). Anal. Calcd for C25H38N1P1B1F4Rh1 (573.27
gmol-1): C 52.38, H 6.68, N 2.44. Found: C 52.12, H 6.84, N
2.32. All 1H, 13C{1H}, and 31P{1H} NMR data for the cation were
identical to those reported for [5b]+SO3CF3-. Crystals suitable for
single-crystal X-ray diffraction analysis were obtained from a
1
101.3 MHz): δ 33.2 (d, JRhP ) 145 Hz). Slow evaporation of a
benzene solution of 5c within the glovebox produced crystals
suitable for single-crystal X-ray diffraction analysis.
Synthesis of [(COD)Ir(κ2-P,N-3a)]+SO3CF3- ([6a]+SO3CF3-).
To a solution of [(COD)IrCl]2 (0.080 g, 0.12 mmol) in THF (5
mL) was added a solution of AgSO3CF3 (0.061 g, 0.24 mmol) in
THF (2 mL) via Pasteur pipet. Immediate formation of a white
precipitate was observed. After 1 h, the reaction mixture was filtered
through Celite and the filtered solids were discarded. To the clear,
bright yellow filtrate solution was added a solution of 3a (0.065 g,
0.24 mmol) in THF (1.5 mL); the solution turned dark red
immediately. The reaction vial was sealed with a PTFE-lined cap,
shaken vigorously for 30 s, and allowed to sit undisturbed for 1 h.
Over this time, an orange crystalline solid had formed in the reaction
vial. The supernatant (which was found to contain a mixture of
[6a]+SO3CF3 and [6b]+SO3CF3 as identified by 31P NMR
analysis) was decanted away from the solid product. In turn, this
solid was dried in vacuo to afford [6a]+SO3CF3- as an analytically
pure, bright orange solid (0.071 g, 0.098 mmol, 41%). Anal. Calcd
for C26H38P1N1Ir1O3S1F3 (724.8 gmol-1): C 43.08; H 5.28; N 1.93.
-
concentrated THF solution of [5b]+BF4 at room temperature.
-
Synthesis of [(COD)Rh(κ2-P,N-3b)]+PF6 ([5b]+PF6-). To a
solution of [(COD)RhCl]2 (0.120 g, 0.244 mmol) in THF (2 mL)
was added a solution of AgPF6 (0.123 g, 0.487 mmol) in THF (2
mL) via Pasteur pipet. The vial was sealed with a PTFE-lined cap
and the mixture shaken vigorously for ∼30 s; the immediate
formation of a white precipitate was observed. After 5 min, a
solution of 3a (0.134 g, 0.487 mmol) in THF (3 mL) was added,
which resulted in the formation of a yellow-orange solution. The
reaction vial was sealed and shaken vigorously for 30 s, after which
the precipitate was allowed to settle over 15 min. The reaction
mixture was filtered through Celite to remove the precipitated solids,
which were discarded. From the initially homogeneous filtrate some
amount of yellow microcrystalline material precipitated ([5a]+PF6-),
which was separated from the supernatant. At this stage, 31P NMR
data collected on an aliquot of the supernatant solution indicated
-
-

606 Organometallics, Vol. 26, No. 3, 2007
Cipot et al.
Found: C 43.08; H 5.55; N 1.95. 1H NMR (CD2Cl2, 500.1 MHz):
was observed. After 1 h, the reaction mixture was filtered through
Celite and the filtered solids were discarded. To the clear, bright
yellow filtrate solution was added 3a (0.041 g, 0.15 mmol) in THF
(1.5 mL), with formation of an orange solution. After 1.5 h, the
solvent and other volatile materials were removed in vacuo to leave
a sticky, brown solid. The solid was washed with toluene (2 mL)
δ 7.43-7.32 (m, 4H, Ar-Hs), 6.57 (s, 1H, C3-H), 4.98 (m, 1H,
2
COD), 4.78 (m, 1H, COD), 4.62 (d, JPH ) 13.0 Hz, 1H, C1-H),
4.20 (m, 1H, COD), 3.94 (m, 1H, COD), 3.27 (s, 3H, NMeaMeb),
3.16 (s, 3H, NMeaMeb), 2.56-2.41 (m, 3H, COD and P(CHMea-
Meb)(CHMecMed) and P(CHMeaMeb)(CHMecMed)), 2.38-2.21 (m,
3H, COD), 2.13-2.03 (m, 2H, COD), 1.89-1.86 (m, 2H, COD),
and subsequently with pentane (2 mL) to leave [6b]+BF4 as an
-
3
3
1.73 (d of d, JPH ) 13.1 Hz, JHH ) 7.2 Hz, 3H, P(CHMea-
analytically pure, orange powder (0.037 g, 0.056 mmol, 37%). Anal.
3
3
Meb)(CHMecMed)), 1.59 (d of d, JPH ) 13.1 Hz, JHH ) 6.9 Hz,
Calcd for C25H38P1N1Ir1B1F4 (662.6 gmol-1): C 45.32; H 5.78; N
3
1
3H, P(CHMeaMeb)(CHMecMed)), 1.16 (d of d, JPH ) 15.3 Hz,
2.11. Found: C 45.39; H 5.70; N 2.15. All H, 13C{1H}, and 31P-
3JHH ) 7.3 Hz, 3H, P(CHMeaMeb)(CHMecMed)), 0.79 (d of d, 3JPH
) 17.9 Hz, 3JHH ) 7.0 Hz, 3H, P(CHMeaMeb)(CHMecMed)). 13C-
{1H} NMR (CD2Cl2, 125.8 MHz): δ 166.9 (C2), 143.5 (C3a or
C7a), 134.7 (C7a or C3a), 129.1, 126.9, 125.6, 123.6 (C4, C5, C6,
{1H} NMR data for the cation were identical to those reported for
[6b]+SO3CF3-. Whereas a much higher apparent isolated yield of
-
[6b]+BF4 is obtained by use of less toluene and pentane in the
washing procedure, the resulting solids thereby obtained did not
reliably provide satisfactory combustion analysis.
3
2
and C7), 118.1 (d, JPC ) 8.7 Hz, C3), 96.8 (d, JPC ) 11.5 Hz,
2
-
Synthesis of [(COD)Ir(κ2-P,N-3b)]+PF6 ([6b]+PF6-). To a
COD), 93.6 (d, JPC ) 11.1 Hz, COD), 61.4 (COD), 58.2 (COD),
1
53.0 (NMeaMeb), 49.1 (NMeaMeb), 42.9 (d, JPC ) 17.3 Hz, C1),
solution of [(COD)IrCl]2 (0.060 g, 0.090 mmol) in THF (3 mL)
was added a solution of AgPF6 (0.045 g, 0.18 mmol) in THF (2
mL) via Pasteur pipet. A white precipitate formed immediately.
After 5 min a solution of 3a (0.049 g, 0.18 mmol) in THF (3 mL)
was added to the reaction mixture, with immediate darkening of
the solution to deep red. The reaction vial was then sealed and
shaken vigorously for 30 s, after which the precipitate was allowed
to settle over 15 min. The reaction mixture was filtered through
Celite and the filtered solids were discarded. To the filtrate was
added NEt3 (0.075 mL), after which the reaction vial was sealed,
manually shaken for 10 s, and left to sit for 5 min. The solvent and
other volatile materials were then removed in vacuo to leave a
sticky, orange-brown solid. The product was extracted into CH2-
Cl2 (5 mL) and the mixture filtered over Celite in order to remove
any insoluble materials. The filtrate was dried in vacuo to leave an
orange solid, which was washed once with pentane (2 mL) and
subsequently dried in vacuo to leave [6b]+PF6- as an analytically
pure, orange powder (0.10 g, 0.14 mmol, 78%). Anal. Calcd for
C25H38P2N1Ir1F6 (720.7 gmol-1): C 41.66; H 5.31; N 1.94. Found:
34.3 (br s, COD), 32.3 (br s, COD), 29.7 (COD), 28.7 (COD), 26.2
1
1
(d, JPC ) 25.6 Hz, P(CHMeaMeb)(CHMecMed)), 24.3 (d, JPC
)
23.4, P(CHMeaMeb)(CHMecMed)), 20.7 (d, 2JPC ) 3.1 Hz, P(CH-
2
MeaMeb)(CHMecMed)), 19.3 (d, JPC ) 2.5 Hz, P(CHMeaMeb)-
(CHMecMed)), 19.1 (d, 2JPC ) 5.6 Hz, P(CHMeaMeb)(CHMecMed)),
18.5 (P(CHMeaMeb)(CHMecMed)). 31P{1H} NMR (CD2Cl2, 202.5
MHz): δ 15.0.
Synthesis of [(COD)Ir(κ2-P,N-3b)]+SO3CF3- ([6b]+SO3CF3-).
To a solution of [(COD)IrCl]2 (0.10 g, 0.15 mmol) in THF (3 mL)
was added a solution of AgSO3CF3 (0.076 g, 0.30 mmol) in THF
(2 mL) via Pasteur pipet. Immediate formation of a white precipitate
was observed. After 1 h, the reaction mixture was filtered through
Celite and the filtered solids were discarded. To the clear, bright
yellow filtrate solution was added a solution of 3a (0.082 g, 0.30
mmol) in THF (1.5 mL); a dark red solution immediately formed.
After magnetically stirring the mixture for 1 h, the supernatant was
decanted away from any precipitate that had formed (identified as
[6a]+SO3CF3- based on 31P NMR analysis) and NEt3 (0.4 mL) was
added to the solution to facilitate the rearrangement of [6a]+SO3CF3-
to [6b]+SO3CF3-. After 30 min, the solvent and other volatile
materials were removed in vacuo to leave a dark brown, sticky
solid. This solid was washed with toluene (2 × 2 mL) and pentane
(2 mL) to leave [6b]+SO3CF3- as an analytically pure, orange solid
(0.10 g, 0.14 mmol, 47%). Anal. Calcd for C26H38P1N1Ir1O3S1F3
(724.8 gmol-1): C 43.08; H 5.28; N 1.93. Found: C 42.96; H 5.09;
N 1.78. 1H NMR (CD2Cl2, 500.1 MHz): δ 7.52-7.48 (m, 2H, Ar-
Hs), 7.38-7.34 (m, 2H, Ar-Hs), 4.81-4.76 (m, 2H, COD), 4.26-
4.20 (m, 2H, COD), 3.80 (s, 2H, C1(H)2), 3.13 (s, 6H, NMe2),
3.06-2.99 (m, 2H, P(CHMe2)2), 2.32-2.26 (m, 4H, COD), 1.94-
1.90 (m, 4H, COD), 1.38 (d of d, 3JPH ) 17.5 Hz, 3JHH ) 7.5 Hz,
6H, P(CHMeaMeb)2), 1.30 (d of d, 3JPH ) 15.5 Hz, 3JHH ) 7.0 Hz,
6H, P(CHMeaMeb)2). 13C{1H} NMR (CD2Cl2, 125.8 MHz): δ 195.9
(SO3CF3), 179.5 (d, 2JPC ) 16.0 Hz, C2), 144.2 (d, 2JPC ) 6.0 Hz,
C 41.44; H 5.41; N 2.11. All 1H, 13C{1H}, and 31P{1H} NMR data
-
for the cation were identical to those reported for [6b]+SO3CF3
.
Crystals suitable for single-crystal X-ray diffraction analysis were
-
obtained from a concentrated THF solution of [6b]+PF6 at room
temperature.
Synthesis of (COD)Ir(κ2-P,N-3c) (6c). A solution of [(COD)-
IrCl]2 (0.099 g, 0.15 mmol) in toluene (3 mL) was added to a slurry
of [3c]Li (0.081 g, 0.29 mmol) in toluene (3 mL) via Pasteur pipet;
the solution instantly darkened to ruby red. After 40 min, the
reaction mixture was filtered through Celite to remove any
precipitate. At this stage, 31P NMR data collected on an aliquot of
the crude reaction mixture indicated the quantitative formation of
6c. The solvent and any other volatile materials were then removed
in vacuo to yield 6c as an analytically pure, dark orange solid (0.16
g, 0.27 mmol, 93%). Anal. Calcd for C25H37P1N1Ir1 (574.8 gmol-1):
1
1
C3a), 136.9 (C7a), 130.7 (d, JPC ) 34.8 Hz, C3), 127.4, 127.2,
C 52.24; H 6.49; N 2.44. Found: C 52.53; H 6.44; N 2.68. H
2
3
125.9, 122.6 (C4, C5, C6, C7), 93.4 (d, JPC ) 11.3 Hz, COD),
NMR (C6D6, 500.1 MHz): δ 7.98 (d, JHH ) 11.0, 1H, C7-H),
59.8 (COD), 52.2 (NMe2), 33.0 (d, 3JPC ) 2.6 Hz, COD), 31.6 (d,
7.80 (d, 3JHH ) 8.0, 1H, C4-H), 7.39-7.32 (m, 2H, C6-H and C5-
H), 6.30 (d, J ) 3.7 Hz, 1H, C1-H), 3.88-3.84 (m, 4H, COD),
2.75 (s, 6H, NMe2), 2.71 (m, 2H, P(CHMe2)2), 1.95-1.79 (m, 4H,
COD), 1.39 (m, 2H, COD), 1.29 (m, 2H, COD), 1.18 (d of d, 3JPH
) 15.1 Hz, 3JHH ) 7.0 Hz, 6H, P(CHMeaMeb)2), 1.05 (d of d, 3JPH
1
3JPC ) 9.5 Hz, C1), 28.8 (COD), 25.4 (d, JPC ) 30.4 Hz,
P(CHMe2)2), 19.7 (P(CHMeaMeb)2), 18.5 (P(CHMeaMeb)2). 31P-
{1H} NMR (CD2Cl2, 202.5 MHz): δ 35.9. Crystals suitable for
single-crystal X-ray diffraction analysis were obtained from a
concentrated THF solution of [6b]+SO3CF3- at room temperature.
3
) 15.9 Hz, JHH ) 7.2 Hz, 6H, P(CHMeaMeb)2). 13C{1H} NMR
-
NB: If in using the procedure above the precipitated [6a]+SO3CF3
2
(C6D6, 125.8 MHz): δ 164.5 (d, JPC ) 22.8 Hz, C2), 139.3 (d, J
is not separated from the supernatant containing a mixture of
) 10.2 Hz, C3, C3a, or C7a), 130.6 (C3, C3a, or C7a), 121.5, 119.8,
-
[6a]+SO3CF3 and [6b]+SO3CF3-, then NEt3 can be added to the
3
116.9, 116.7 (C4, C5, C6, and C7), 87.3 (d, JPC ) 11.2 Hz, C1),
mixture to afford analytically pure [6b]+SO3CF3- in 89% isolated
84.7 (d, JPC ) 11.8 Hz, COD), 54.8 (NMe2), 54.6 (COD), 32.8 (d,
1
yield.
JPC ) 2.3 Hz, COD), 28.7 (COD), 26.4 (d, JPC ) 33.8 Hz,
Synthesis of [(COD)Ir(κ2-P,N-3b)]+BF4 ([6b]+BF4-). To a
P(CHMe2)2), 19.6 (P(CHMeaMeb)2), 19.1 (P(CHMeaMeb)2). 31P-
{1H} NMR (C6D6, 202.5 MHz): δ 25.0. Crystals suitable for single-
crystal X-ray diffraction analysis were grown from a concentrated
toluene solution of 6c at room temperature.
-
solution of [(COD)IrCl]2 (0.050 g, 0.075 mmol) in THF (3 mL)
was added a solution of AgBF4 (0.032 g, 0.16 mmol) in THF (2
mL) via Pasteur pipet. Immediate formation of a white precipitate

Rh and Ir κ2-P,N Complexes
Organometallics, Vol. 26, No. 3, 2007 607
Synthesis of (CO)(Cl)Rh(κ2-P,N-3b) (7). To a solution of 3a
(0.050 g, 0.18 mmol) in THF (2 mL) was added via Pasteur pipet
a solution of [(CO)2RhCl]2 (0.016 g, 0.091 mmol) in THF (1 mL).
The resulting clear yellow solution was magnetically stirred for 10
min, upon which K2CO3 (0.050 g, 0.36 mmol) was added. After
magnetically stirring the heterogeneous mixture for 30 min, the
slurry was filtered through Celite and the filtered solid discarded,
to leave a clear orange filtrate solution. The solvent was removed
in vacuo to leave an oily residue, from which some product (7)
was extracted into Et2O (2 × 2 mL); yellow-orange crystalline plates
of 7 suitable for X-ray crystallographic analysis were obtained from
this fraction. Additional product was obtained by dissolution of
the remaining oily residue in benzene (2 mL), from which brown
impurities were precipitated out by addition of pentane (1 mL), to
leave a yellow solution. The solution was filtered through Celite
and the product isolated by removal of the solvent in vacuo. The
combined crops of analytically pure 7 were dried in vacuo for
18 h (0.034 g, 0.077 mmol, 43%). Anal. Calcd for C18H26-
PNOClRh: C 48.94, H 5.93, N 3.17. Found: C 49.07, H 6.26, N
Synthesis of (PPh2Me)2Rh(κ2-P,N-3c) (9). A dark red solution
of 5c (0.096 g, 0.20 mmol) in toluene (5 mL) was cooled to -35 °C,
after which PPh2Me (0.19 mL, 0.99 mmol) was added via
Eppendorf pipet. The solution immediately darkened to brown-red.
The vial was sealed with a PTFE-lined cap, and the mixture was
left to magnetically stir for 36 h. 31P NMR analysis of the reaction
at this stage confirmed the quantitative conversion of 5c to 9. The
solvent was removed in vacuo and the dark brown solid was washed
with pentane (3 × 2 mL). The resulting solid was dried in vacuo
for 16 h to leave 9 as an analytically pure, light brown powder
(0.11 g, 0.14 mmol, 66%). Anal. Calcd for C43H51P3N1Rh1 (777.7
gmol-1): C 66.41; H 6.61; N 1.80. Found: C 66.21; H 6.49; N
1.60. NB: Signals in the 1H NMR spectrum of 9 are broadened at
300 K, in keeping with the observed dynamic behavior of this
complex (see text). Some 13C NMR resonances are also broadened,
and we were unable to observe the NMe2 resonances in the 13C-
{1H} NMR spectrum of this complex. Data reported here were
1
obtained at 300 K, unless otherwise indicated. H NMR (C6D6,
3
500.1 MHz): δ 7.84 (d, JHH ) 7.6 Hz, 1H, C4-H), 7.73 (m, 1H,
1
3.29. H NMR (C6D6, 500.1 MHz): δ 7.11-7.06 (m, 2H, C5-H
C5-H or C6-H), 7.59 (m, 1H, C6-H or C5-H), 7.51 (d, 3JHH ) 7.5
Hz, C7-H), 7.39-7.26 (m, 5H, Ar-Hs), 7.16-6.92 (m, some
resonances obscured by reference solvent, Ar-Hs), 6.18 (s, 1H,
C1-H), 3.50-2.10 (very broad resonance at 300 K; upon cooling
to 240 K, this resolved into two singlets at 3.34 ppm (3H, NMea)
and 2.52 ppm (3H, NMeb)), 3.00 (m, 1H, Pc(CHMe2)(CHMe2)),
and C6-H), 7.03 (d, 3JHH ) 6.8 Hz, 1H, C4-H), 7.00 (d, 3JHH ) 6.8
Hz, 1H, C7-H), 2.59 (s, 6H, NMe2), 2.46 (s, 2H, C1(H)2), 2.06-
3
3
1.97 (m, 2H, P(CHMeaMeb)2), 1.06 (d of d, JPH ) 17.8 Hz, JHH
) 6.9 Hz, 6H, P(CHMeaMeb)2), 0.81 (d of d, 3JPH ) 16.7 Hz, 3JHH
) 6.9 Hz, 6H, P(CHMeaMeb)2). 13C{1H} NMR (C6D6, 125.8
1
2
2
MHz): δ 189.7 (d of d, JRhC ) 73.2 Hz, JPC ) 18.5 Hz, CO),
175.3 (d, J ) 18.9 Hz, C2), 143.4 (C3a or C7a), 138.3 (C7a or
C3a), 132.4 (d, 1JPC ) 29.2 Hz, C3), 126.8 (C5 or C6), 125.8 (C6
2.79 (m, 1H, Pc(CHMe2)(CHMe2)), 1.58 (d, JPH ) 6.2 Hz, 3H,
Pa(C6H5)2Me), 1.29-1.12 (m, 6H, Pc(CHMeaMeb)2), 1.10 (d, 2JPH
) 6.8 Hz, 3H, Pb(C6H5)2Me), 0.95-0.83 (m, 6H, Pc(CHMeaMeb)2).
13C{1H} NMR (C6D6, 125.8 MHz): δ 167.3 (C2), 147.6 (C3a or
C7a), 144.1, 143.4 (Pa(C(CH)5)2Me), 142.3 (C7a or C3a), 140.4,
139.4 (Pb(C(CH)5)2Me), 133.8, 133.6, 132.9, 132.7, 132.1, 131.9,
131.8 (P(C(CH)5)2Me), 123.7 (C4), 121.5 (C5 or C6), 118.2, 118.1
(C7 and C6 or C7 and C5), 103.2 (C1), 37.3 (d, 1JPC ) 112.3 Hz,
C3), 27.7 (br s, Pc(CHMe2)(CHMe2)), 26.9 (br s, Pc(CHMe2)-
or C5), 124.9 (C7), 121.6 (C4), 48.8 (NMe2), 29.8 (d, J ) 10.2
2
Hz, C1), 26.5 (d, J ) 30.8 Hz, P(CHMeaMeb)2), 19.7 (d, JPC
)
5.3 Hz, P(CHMeaMeb)2), 19.5 (P(CHMeaMeb)2). 31P{1H} NMR
(C6D6, 202.5 MHz): δ 72.2 (d, JRhP ) 161.5 Hz). FTIR (C6H6/
1
CsI; cm-1) ν(CO): 1977.
-
Synthesis of [(CO)2Rh(κ2-P,N-3b)]+PF6 (8). Compound
1
[5b]+PF6- (0.054 g, 0.085 mmol) was dissolved in CH2Cl2 (4 mL)
in a Schlenk flask equipped with a magnetic stir bar. The flask
was transferred to a Schlenk line, and the orange solution was
degassed by use of three successive freeze-pump-thaw cycles.
Magnetic stirring of the solution was initiated, and the flask was
backfilled with CO (∼1 atm). Within 30 s, the solution turned bright
yellow in color. After 90 min, the flask was purged with N2 for
20 min. The reaction vessel was then transferred into the glovebox,
whereupon pentane (2 mL) was added to the reaction mixture in
order to induce precipitation of the product (8) as a fluffy, yellow
solid. The solid product 8 was isolated by filtration through a
Bu¨chner funnel equipped with a fine-porosity filter paper. Ad-
ditional product was isolated in a similar manner by concentrating
the filtrate to approximately half the original volume in vacuo,
followed by adding an additional amount of pentane (2 mL) to
induce further precipitation. This process was repeated until the
filtrate was essentially colorless. All isolated crops of 8 were
combined and dried in vacuo for 5 h (0.033 g, 0.056 mmol, 66%).
Anal. Calcd for C19H26P2NO2F6Rh: C 39.39, H 4.52, N 2.42.
Found: C 39.23, H 4.88, N 2.38. 1H NMR (CD2Cl2, 500.1 MHz):
7.62 (m, 1H, C7-H), 7.46-7.44 (m, 3H, C4-H, C5-H, and C6-
H), 3.79 (s, 2H, C1(H)2), 3.30 (s, 6H, NMe2), 2.89-2.82 (m, 2H,
P(CHMeaMeb)2), 1.51 (d of d, 3JPH ) 19.4 Hz, 2JHH ) 7.0 Hz, 6H,
P(CHMeaMeb)2), 1.32 (d of d, 3JPH ) 18.1 Hz, 2JHH ) 7.0 Hz, 6H,
P(CHMeaMeb)2). 13C{1H} NMR (CD2Cl2, 125.8 MHz): δ 175.7
(d, J ) 16.5 Hz, C2), 144.2 (d, J ) 6.4 Hz, C3a or C7a), 136.2
(CHMe2)), 22.3 (Pc(CHMeaMeb)2), 21.5 (d, JPC ) 20.9 Hz,
1
Pa(C6H5)2Me), 20.1, 20.0 (Pc(CHMeaMeb)2), 14.5 (d, JPC ) 23.1
Hz, Pb(C6H5)2Me). 31P{1H} NMR (C6D6, 202.5 MHz): δ 35.8 (d
of d of d, 1JPaRh ) 113.8 Hz, 2JPaPb ) 31.6 Hz, 2JPaPc ) 237.4 Hz,
Pa(C6H5)2Me, cis to N), 24.2 (d of d of d, 1JPbRh ) 189.8 Hz, 2JPbPa
) 31.1 Hz, 2JPbPc ) 26.6 Hz, Pb(C6H5)2Me, trans to N), 15.6 (d of
1
2
2
d of d, JPcRh ) 177.0 Hz, JPcPb ) 26.0 Hz, JPcPa ) 237.2 Hz,
Pc(CHMe2)2).
General Protocol for Hydrogenation Experiments. The pro-
tocol used for hydrogenation reactions employing 0.5 mol % catalyst
loading in CH2Cl2 is provided as a representative procedure. A
solution of catalyst compound in CH2Cl2 (0.014 mmol in 7 mL to
give a 0.002 M solution) was allowed to equilibrate for 5 min, at
which point the alkene (2.8 mmol) was added by use of an
Eppendorf pipet. The vial was then sealed and shaken vigorously.
Subsequently, 1 mL aliquots of the mixture were placed in glass
reactor cells, which were each equipped with a magnetic stir bar
and sealed under nitrogen with a PTFE valve. The cells were
transferred immediately to a Schlenk line and degassed by use of
three freeze-pump-thaw cycles employing liquid nitrogen. Mag-
netic stirring of the solutions was initiated, and the evacuated cells
were backfilled with hydrogen gas (∼1atm, UHP grade). With the
exception of the noted experiments in which the catalyst was not
fully soluble, clear yellow homogeneous solutions were observed
throughout. At the desired sampling time, the reactor cell was
opened to air and 2 mL of hexanes (or pentane in the case of the
hydrogenation of cyclohexene) was added via Pasteur pipet. The
resultant slurries were then filtered through a short Al2O3 column
(2 cm), from which clear, colorless solutions eluted. These solutions
were transferred to GC vials and sealed. Products of each reaction
were identified by use of GC-MS, and quantitative data were
obtained from GC-FID analysis; tabulated data represent the average
of at least two runs.
1
(C7a or C3a), 130.0 (d, JPC ) 34.0 Hz, C3), 127.9, 127.7, 122.2
(C4, C5, and C6), 125.8 (C7), 54.8 (NMe2), 30.9 (d, J ) 10.8 Hz,
1
C1), 26.1 (d, JPC ) 29.6 Hz, P(CH(MeaMeb)2), 19.9 (P(CHMea-
Meb)2 or P(CHMeaMeb)2), 19.8 (d, 2JPC ) 3.7 Hz, P(CHMeaMeb)2
or P(CHMeaMeb)2). 31P{1H} NMR (CD2Cl2, 202.3 MHz): δ 62.7
1
1
(d, JRhP ) 123.5 Hz, P(CHMe2)2), -144.5 (septet, JPF ) 710.7
Hz, PF6-). FTIR (CH2Cl2/CsI; cm-1) ν(CO): 2100, 2030, 1990.

608 Organometallics, Vol. 26, No. 3, 2007
Cipot et al.
2
2
General Protocol for Hydrosilylation Experiments. The
protocol used for these reactions employing 5.0 mol % catalyst
loading (relative to Et3SiH) in benzene with a styrene to silane ratio
of 5:1 run at 60 °C is provided as a representative procedure. A
solution of catalyst compound in benzene (0.0225 mmol in 4.5 mL
to give a 0.005 M solution) was allowed to equilibrate for 5 min,
at which point the alkene (2.25 mmol) was added by use of an
Eppendorf pipet. The vial was then sealed and shaken vigorously.
Subsequently, Et3SiH (0.45 mmol) was added by use of an
Eppendorf pipet to the reaction mixture, and the vial was then sealed
and shaken as before. Aliquots (1 mL) of the mixture were placed
in glass reactor cells, which were each equipped with a magnetic
stir bar and sealed under nitrogen with a PTFE valve. The cells
were transferred immediately to a Schlenk line, and magnetic
stirring of the solutions was initiated. At the desired sampling time,
the reactor cell was opened to air and ∼1 mL of pentane was added
via Pasteur pipet. The resultant mixtures were then filtered through
a short Al2O3 column (2 cm), from which clear, colorless solutions
eluted. These solutions were transferred to GC vials and sealed.
Products of each reaction were identified by use of GC-MS, while
quantitative data were obtained from GC-FID analysis; tabulated
data represent the average of at least two runs.
Crystallographic Solution and Refinement Details. With the
exception of 7, crystallographic data were obtained at 193((2) K
on a Bruker PLATFORM/SMART 1000 CCD diffractometer using
graphite-monochromated Mo KR (λ ) 0.71073 Å) radiation,
employing samples that were mounted in inert oil and transferred
to a cold gas stream on the diffractometer. Programs for diffrac-
tometer operation, data collection, data reduction, and absorption
correction (including SAINT and SADABS) were supplied by
Bruker. The structures were solved by use of direct methods (except
for 4a‚CH2Cl2 and 5c, where in each case a Patterson search/
structure expansion was employed) and refined by use of full-matrix
least-squares procedures (on F2) with R1 based on Fo g 2σ(Fo )
2
2
and wR2 based on Fo g -3σ(Fo ). Anisotropic displacement
parameters were employed throughout for the non-hydrogen atoms,
and all hydrogen atoms were added at calculated positions and
refined by use of a riding model employing isotropic displacement
parameters based on the isotropic displacement parameter of the
attached atom. Structural disorder was noted during the refinement
of [6b]+PF6-, which was treated in a satisfactory manner by
employing a model featuring two half-occupied formula units of
the complex within the asymmetric unit. Data and experimental
details related to the crystallographic characterization of 7 (including
an ORTEP diagram) are provided in the Supporting Information.
Acknowledgment is made to the Natural Sciences and
Engineering Research Council (NSERC) of Canada (including
a Discovery Grant for M.S. and a Postgraduate Scholarship for
J.C.), the Canada Foundation for Innovation, the Nova Scotia
Research and Innovation Trust Fund, and Dalhousie University
for their generous support of this work. We also thank Dr.
Michael Lumsden (Atlantic Region Magnetic Resonance Center,
Dalhousie) for his assistance in the acquisition of NMR data,
as well as Deepa Abeysekera and David Wolstenholme for their
assistance in studying the structural and reactivity properties
of [5b]+BF4
.
-
Supporting Information Available: Single-crystal X-ray dif-
fraction data in CIF format for 4a‚CH2Cl2, [5b]+BF4-, 5c,
[6b]+SO3CF3-, [6b]+PF6-, 6c, and 7, as well as crystallographic
details and an ORTEP diagram of 7, are available free of charge
via the Internet at http://pubs.acs.org.
OM060758C