Ruthenated acetonitrile: Unusual bronsted acidity of a polar Aprotic solvent

Article abstract of DOI:10.1021/om800343x

Addition of acetonitrile to the complex [Ru(η⁵-indenyl) (PR₂)(PPh₃)] (1) gives the unusual metalation product [Ru(η⁵-indenyl)(CH₂CN)(HPR₂)(PPh ₃)] (2), which has been structurally characterized. This reaction clearly demonstrates high Bronsted basicity at the terminal phosphide ligand in 1. ³¹P{¹H} NMR studies show that less acidic N-donor solvents simply disrupt the Ru-P π-bond in 1 to give adducts [Ru(η⁵-indenyl)(L)(HPR₂)(PPh₃)] (L = benzonitrile (6) or pyridine (7)), which are in equilibrium with 1 and free L. The analogous acetonitrile adduct (4) was observed by NMR at 240 K during the formation of 2, but is quickly replaced by 2 at higher temperatures. NMR studies of an alternate route to the metalated complex 2, starting from the cationic N-bound acetonitrile adduct [Ru(η⁵-indenyl)(NCCH ₃-(HPCy₂)(PPh₃)][PF₆] (3a), along with the demonstrated lability of the benzonitrile and pyridine adducts, suggest that the metalation of acetonitrile by 1 proceeds via an intermolecular C-H addition across the Ru=P double bond, rather than the intramolecular C-H activation of N-bound acetonitrile. This is confirmed by the observation, by ³¹P[¹H} NMR, of multiple product isotopomers in the reaction of 1 with a 1:1 mixture of d₃- and d₀- acetonitrile. O-Donor solvents also deprotonated by 1 include water, alcohols, and acetone, which give the complexes [Ru(η⁵-indenyi)(X) (HPCy₂)(PPh₃)], where X = OR (8), CH₂(=O) CH₃ (10).

Full text of DOI:10.1021/om800343x

Organometallics 2008, 27, 5025–5032
5025
Ruthenated Acetonitrile: Unusual Brønsted Acidity of a Polar
“Aprotic” Solvent
Eric J. Derrah,† Karina E. Giesbrecht,† Robert McDonald,‡ and Lisa Rosenberg*,†
Department of Chemistry, UniVersity of Victoria, P.O. Box 3065, Victoria, BC, Canada V8W 3V6, and
X-ray Crystallography Laboratory, Department of Chemistry, UniVersity of Alberta, Edmonton,
AB, Canada T6G 2G2
ReceiVed April 16, 2008
Addition of acetonitrile to the complex [Ru(η5-indenyl)(PR2)(PPh3)] (1) gives the unusual metalation
product [Ru(η5-indenyl)(CH2CN)(HPR2)(PPh3)] (2), which has been structurally characterized. This
reaction clearly demonstrates high Brønsted basicity at the terminal phosphido ligand in 1. 31P{1H} NMR
studies show that less acidic N-donor solvents simply disrupt the Ru-P π-bond in 1 to give adducts
[Ru(η5-indenyl)(L)(HPR2)(PPh3)] (L ) benzonitrile (6) or pyridine (7)), which are in equilibrium with
1 and free L. The analogous acetonitrile adduct (4) was observed by NMR at 240 K during the formation
of 2, but is quickly replaced by 2 at higher temperatures. NMR studies of an alternate route to the metalated
complex 2, starting from the cationic N-bound acetonitrile adduct [Ru(η5-indenyl)(NCCH3)-
(HPCy2)(PPh3)][PF6] (3a), along with the demonstrated lability of the benzonitrile and pyridine adducts,
suggest that the metalation of acetonitrile by 1 proceeds via an intermolecular C-H addition across the
RudP double bond, rather than the intramolecular C-H activation of N-bound acetonitrile. This is
confirmed by the observation, by 31P{1H} NMR, of multiple product isotopomers in the reaction of 1
with a 1:1 mixture of d3- and d0-acetonitrile. O-Donor solvents also deprotonated by 1 include water,
alcohols, and acetone, which give the complexes [Ru(η5-indenyl)(X)(HPCy2)(PPh3)], where X ) OR
(8), CH2(dO)CH3 (10).
metalation of acetonitrile5,6 to give M-CH2CN complexes are
surprisingly rare, although such C-H bond activation is
sometimes implicit from product analysis.7 Complexes resulting
from the oxidative addition of acetonitrile include [Ir(PMe3)4-
(CH2CN)(H)]Cl8 and [M(dmpe)2(CH2CN)(H)] (M ) Fe, Ru).9
Dinuclear lanthanide complexes containing µ-κ2-CH2CN ligands
form via silane elimination when acetonitrile is added to
trimethylsilyl precursor complexes.10 Other apparent examples
of the direct metalation of acetonitrile result from the oxidative
addition of the C-Cl bond of chloroacetonitrile (e.g.,
[Pt(PPh3)2(CH2CN)(Cl)]11 and [Ni(dippe)(CH2CN)(Cl)]4) and
subsequent reduction of the metal-chloride bond, or from
metathesis of metal halide precursors with “pre-metalated”
acetonitrile (e.g., [Ni(η5-Cp)(CH2CN)(PPh3)], prepared using
n-BuLi/CH3CN).12
Introduction
Acetonitrile is an extremely useful solvent in coordination
chemistry. Its polarity allows dissolution of the salts of complex
ions and other polar transition metal complexes. Its donor ability,
via the nitrile N, allows stabilization of low-coordinate species
in solution and their subsequent crystallization as nitrile adducts.
As a ligand, N-bound acetonitrile is sufficiently labile toward
substitution by other donor ligands that it finds widespread use
in transition metal precursor complexes.1 Thus, mechanisms for
both stoichiometric and catalytic reactions carried out in
acetonitrile frequently invoke this neutral N-donor as a benign
placeholder ligand. That being said, there is an appreciable
literature concerned with metal-mediated reduction of the
acetonitrile C-N triple bond, which typically involves the
participation of N-bound or η2-CN-bound acetonitrile ligands,
leading to formation of new C-H, C-O, C-C, and/or N-H
bonds.2,3 Activation of the C-C bond in acetonitrile, via similar
intermediate complexes, is also known.4 Examples of the
(5) Stewart, R. The Proton: Applications to Organic Chemistry;
Academic Press: Orlando, 1985; Chapter 2. The pKa of 24 for acetonitrile
was measured in DMSO and converted to a measurement in H2O. All other
pKa values mentioned in this article were measured in H2O.
(6) The C-H bond strength in acetonitrile is ∼97 kcal/mol. Lide, D. R.,
Ed. CRC Handbook of Chemistry and Physics; CRC Press: Boca Raton,
FL, 1987; pp 9-61.
(7) See for example: (a) Cobalt-mediated H/D exchange from solvent
CD3CN: Fujita, E.; Creutz, C. Inorg. Chem. 1994, 33, 1729. (b) Formation
ofnovelcrotonitrileamidoligandsfromacetonitrilemetalationatyttrium:Duchateau,
R.; vanWee, C. T.; Teuben, J. H. Organometallics 1996, 15, 2291. (c) R-C-
H activation of nitriles: Murahashi, S.; Takaya, H. Acc. Chem. Res. 2000,
33, 225.
(8) English, A. D.; Herskovitz, T. J. Am. Chem. Soc. 1977, 99, 1648.
(9) Ittel, S. D.; Tolman, C. A.; English, A. D.; Jesson, J. P. J. Am. Chem.
Soc. 1978, 100, 7577.
(10) Heeres, H. J.; Meetsma, A.; Teuben, J. H. Angew. Chem., Int. Ed.
Engl. 1990, 29, 420.
(11) (a) Ros, R.; Bataillard, R.; Roulet, R. J. Organomet. Chem. 1976,
118, C53. (b) Ros, R.; Renaud, J.; Roulet, R. HelV. Chim. Acta 1975, 58,
133.
* Corresponding author. E-mail: lisarose@uvic.ca.
† University of Victoria.
‡ University of Alberta.
(1) Collman, J. P.; Hegedus, L. S.; Norton, J. R.; Finke, R. G. Principles
and Applications of Organotransition Metal Chemistry; University Science
Books: Sausalito, CA, 1987; p 199.
(2) Storhoff, B. N.; Lewis, H. C. Coord. Chem. ReV. 1977, 23, 1.
(3) See for example:Pittard, K. A.; Cundari, T. R.; Gunnoe, T. B.; Day,
C. S.; Petersen, J. L. Organometallics 2005, 24, 5015. Bianchini, C.; Meli,
A.; Moneti, S.; Vizza, F. Organometallics 1998, 17, 2636. Joshi, A. M.;
MacFarlane, K. S.; James, B. R.; Frediana, P. Chem. Ind. (Dekker) 1994,
53, 497. Chin, C. S.; Lee, B. Catalysis 1992, 14, 135. Rhodes, L. F.;
Venanzi, L. M. Inorg. Chem. 1987, 26, 2692. Grey, R. A.; Pez, G. P.; Wallo,
A. J. Am. Chem. Soc. 1981, 103, 7536.
(4) Atesin, T. A.; Li, T.; Lachaize, S.; Brennessel, W. W.; Garcia, J. J.;
Jones, W. D. J. Am. Chem. Soc. 2007, 129, 7562.
10.1021/om800343x CCC: $40.75
2008 American Chemical Society
Publication on Web 09/10/2008

5026 Organometallics, Vol. 27, No. 19, 2008
Derrah et al.
Scheme 1
We recently isolated the terminal phosphido complex [Ru(η5-
indenyl)(PR2)(PPh3)] (1), in which a novel RudPR2 double bond
offsets coordinative unsaturation at the five-coordinate Ru
center.13 Although complex 1 is soluble in nonpolar hydrocarbon
solvents such as hexanes and benzene, we needed to identify a
more polar solvent for mechanistic studies of a range of 2+2-
cycloaddition reactions at the Ru-P double bond in 1.14 Of the
solvents we screened, acetonitrile gave the most surprising result.
We report here its reaction with 1 to give the unusual metalated
product [Ru(η5-indenyl)(CH2CN)(HPR2)(PPh3)] (2, Scheme 1),
an overall 1,2-addition of the acetonitrile C-H bond across the
Ru-P double bond in 1. We place this reaction in the context
of the interactions of 1 with other N- and O-donor solvents.
The chemistry described here and the clean formation of 2
highlight the potential utility of the terminal phosphido group
in metal-mediated chemistry: it is a ligand of variable coordina-
tion mode, which also acts as a potent internal base.
The metalation of acetonitrile by 1 was a surprise on several
counts. We had anticipated that 1 might react with acetonitrile
to form an N-bound adduct (4) (Scheme 2),18 analogous to the
deep red, six-coordinate carbonyl adduct [Ru(η5-indenyl)(PR2)-
(CO)(PPh3)] (5),19 in which the terminal phosphido ligand
persists, but has pyramidal, instead of planar, geometry at
phosphorus and has a much longer Ru-P single bond.20
Analogous N-bound adducts (6) result from the addition of
benzonitrile to d8-toluene solutions of 1. Pyramidal geometry
at the phosphido ligands in complexes 6a,b is evident from their
2
upfield 31P chemical shifts and extremely small JPP coupling
(<5 Hz) to coordinated PPh3 (Table 1).13,21 Small amounts of
1 remain in equilibrium with adducts 6 in these mixtures, even
in the presence of excess benzonitrile.22 The addition of excess
pyridine to d8-toluene solutions of 1a,b gives similar equilibria,
but N-bound adducts 7a,b are observed only at low temperature:
at temperatures above 290 K the mixtures decompose to a
number of as-yet unidentified products. Despite the fact that
pyridine is a stronger donor than benzonitrile,23 the relative
amounts of 1 in these samples are higher than in the benzonitrile
samples, and the equilibria shift from 7 toward (1 + pyridine)
as the temperature is raised from 230 K to 290 K. Consistent
with this shift is a color change from deep red to blue as the
sealed NMR samples are thawed and warmed. Van’t Hoff
Results and Discussion
The addition of excess acetonitrile to dark blue solutions of
1 in d8-toluene gives an immediate color change to yellow-
orange. 31P{1H} NMR spectroscopy indicates disappearance of
the downfield peaks corresponding to planar phosphido ligands
(276.3 (1a), 288.3 (1b) ppm) and the appearance of new peaks
1
at 71.4 (2a) and 78.4 (2b) ppm, which show large JPH
∼324-330 Hz in the corresponding proton-coupled 31P spectra
1
(Table 1). H NMR spectra also confirm the P-H bond in 2,
showing doublets of multiplets at 3.86 (2a) and 3.98 (2b) ppm
with matching 1JPH splittings. 1H peaks due to the diastereotopic
methylene protons of the CH2CN ligand are seen at 1.14, 0.80
(2a) and 1.05, 0.79 (2b) ppm; these multiplets become simple
(18) Another possible outcome was the 2+2-cycloaddition of the CN
triple bond in acetonitrile and the RudPR2 bond in 1. Although previously
observed for a bent phosphinidene fragment (MdPR; Hou, Z. M.; Breen,
T. L.; Stephan, D. W. Organometallics 1993, 12, 3158) and for a Schrock
carbene complex (MdCR2; Wood, C. D.; McLain, S. J.; Schrock, R. R.
J. Am. Chem. Soc. 1979, 101, 3210), we see no evidence for this reaction.
(19) We reported 5a previously (ref 13); 5b was prepared similarly, by
sealing an NMR sample of 1b under an atmosphere of carbon monoxide.
For 5b, 31P{1H} (145.78 MHz, d8-toluene, δ): 67.7 (d, 2JPP ) 9 Hz, PPri2),
51.6 (d, PPh3).
(20) (a) This switch in phosphido coordination mode (planar vs
pyramidal) is similar to the well-established hemilability of the nitrosyl
ligand (linear vs bent), as has been noted previously: Bohle, D. S.; Jones,
T. C.; Rickard, C. E. F.; Roper, W. R. Organometallics 1986, 5, 1612.
Similar adduct formation is observed for other “operationally unsaturated”,
five-coordinate ruthenium complexes: (b) Johnson, T. J.; Folting, K.; Streib,
W. E.; Martin, J. D.; Huffman, J. C.; Jackson, S. A.; Eisenstein, O.; Caulton,
K. G. Inorg. Chem. 1995, 34, 488.
2
1
doublets with JHH ) 2.0 Hz in the corresponding H{31P}
spectra. These assignments are confirmed by correlation to
13C{1H} peaks for the unique, Ru-bound methylene carbon at
1
-25.6 (2a) and -25.9 (2b) ppm in the H/13C HSQC spectra
and to the nitrile carbon peak (133.1 (2a), 133.0 (2b) ppm) in
the 1H/13C HMBC spectra. Infrared spectroscopy also supports
the metalated structure of 2: in addition to νPH (2332 (2a) and
2296 (2b) cm-1), we observe νCN at 2179 (2a) and 2181 (2b)
cm-1. These are lower frequencies than for N-bound acetonitrile
in [Ru(η5-indenyl)(NCCH3)(HPR2)(PPh3)]PF6 (3, vide infra, νCN
) 2272 cm-1),15 although they are considerably higher than
those reported for the corresponding alkali metal carbanions,16
suggesting a less ionic metal-carbon interaction in 2.17 Finally,
X-ray diffraction analysis of a single crystal of 2a gives
unequivocal evidence for its metalated structure in the solid state
(Figure 1).
(21) (a) Planas, J. G.; Hampel, F.; Gladysz, J. A. Chem.-Eur. J. 2005,
11, 1402. (b) Given the importance of strong donor ancillary ligands in the
P-basic complexes described in (a), it is perhaps not surprising that complex
5, a comparable 18e- complex with a π-acidic CO ligand, does not cause
isotope scrambling when added to a 1:1 mixture of d0- and d3-acetonitrile.
(See Experimental Section.)
(22) (a) This equilibrium is evident in 31P{1H} spectra of sealed samples
containing 1 and excess benzonitrile (see Experimental Section and
Supporting Information). While pure 1a,b in d8-toluene slowly decompose
at room temperature in solution via the orthometalation of a phenyl group
on the coordinated triphenylphosphine ligand, to give [Ru(η5-indenyl)
(HPR2)(κ2-o-C6H4PPh2)] (ref 13), the absence of this product in samples
of 1 and excess benzonitrile that have stood for 24 h or longer suggests
that reVersible formation of 6 is fast relative to the orthometalation reaction
(i.e., lifetimes of 1 in the equilibrium mixtures are sufficiently short to
preclude this slow thermal decomposition). (b) We see no evidence for the
analogous, reversible dissociation of CO from complex 5 in solution. (See
Experimental Section.)
(12) Davidson, J. G.; Barefield, E. K.; Vanderveer, D. G. Organome-
tallics 1985, 4, 1178.
(13) Derrah, E. J.; Pantazis, D. A.; McDonald, R.; Rosenberg, L.
Organometallics 2007, 26, 1473.
(14) Derrah, E. J.; Pantazis, D. A.; McDonald, R.; Hall, S. A.; Rosenberg,
L., manuscript in preparation.
(15) N-Coordinated nitriles often exhibit a shift of νCN to higher
frequency than that for the free nitrile (ref 2). For free acetonitrile, νCN
)
2254 cm-1: SDBSWeb: http://riodb01.ibase.aist.go.jp/sdbs/ (National In-
stitute of Advanced Industrial Science and Technology, July 25, 2007).
(16) For MCH2CN, where M ) Li, Na, K, νCN ) 2049-2051 cm-1
.
(23) A useful measure of “donicity” (Lewis basicity) of the solvents
studied is their donor number, “DN”. (a) Gutmann, V.; Schmid, R. Coord.
Chem. ReV. 1974, 12, 263. (b) For pyridine, acetonitrile, and benzonitrile,
DN ) 33.1, 14.1, and 11.9, respectively.
Juchnovski, I. N.; Dimitrova, J. S.; Binev, I. G.; Kaneti, J. Tetrahedron
1978, 34, 779.
(17) See ref 9 for further discussion of νCN for M-CH2CN complexes.

Ruthenated Acetonitrile
Organometallics, Vol. 27, No. 19, 2008 5027
2
Table 1. 31P{1H} NMR Data for New Compounds: δ (ppm) (multiplicity, JPP or ω1/2 (Hz))
a
complex
PR2/HPR2
PPh3
1JPH
[Ru(η5-indenyl)(CH2CN)(HPCy2)(PPh3)]
[Ru(η5-indenyl)(CH2CN)(HPPri2)(PPh3)]
[Ru(η5-indenyl)(NCCH3)(HPCy2)(PPh3)]PF6
[Ru(η5-indenyl)(NCCH3)(HPPri2)(PPh3)]PF6
[Ru(η5-indenyl)(PCy2)(NCCH3)(PPh3)]
[Ru(η5-indenyl)(PPri2)(NCCH3)(PPh3)]
[Ru(η5-indenyl)(PCy2)(py)(PPh3)]
2ab
72.4 (d, 38)
78.3 (d, 33)
57.1 (d, 36)
67.0 (d, 36)
65.2 (s)
58.8 (d)
330
324
344
340
2bb
58.9 (d)
52.8 (d)
52.2 (d)
54.4 (s)
3ac
3bc
4ad,e
4bd,e
74.0 (s)
53.8 (s)
6ad,e
82.9 (br s, 29)
95.8 (br s, 18)
68.9 (s)
51.0 (br s, 20)
48.2 (br s, 19)
54.1 (s)
[Ru(η5-indenyl)(PPri2)(py)(PPh3)]
6bd,e
[Ru(η5-indenyl)(PCy2)(NCPh)(PPh3)]
[Ru(η5-indenyl)(PPri2)(NCPh)(PPh3)]
[Ru(η5-indenyl)(OH)(HPCy2)(PPh3)]
[Ru(η5-indenyl)(OH)(HPPri2)(PPh3)]
[Ru(η5-indenyl)(OMe)(PPh3)(HPCy2)]
[Ru(η5-indenyl)(OMe)(PPh3)(HPPri2)]
[Ru(η5-indenyl)(OBut)(HPCy2)(PPh3)]
[Ru(η5-indenyl)(OBut)(HPPri2)(PPh3)]
[Ru(η5-indenyl)(CH2C(O)CH3)(HPCy2)(PPh3)]
[Ru(η5-indenyl)(CH2C(O)CH3)(HPPri2)(PPh3)]
7ad,f
7bd,f
79.0 (s)
53.6 (s)
8a-Hd
8b-Hd
8a-Med,f
8b-Med,f
8a-Butd
8b-Butd
10ad,g
10bd,e
66.8 (d, 46)
75.9 (d, 44)
65.3 (d, 48)
75.5 (d, 48)
68.1 (d, 44)
77.2 (d, 44)
67.9 (d, 42)
77.2 (d, 44)
55.0 (d)
54.8 (d)
55.6 (d)
53.8 (d)
55.0 (d)
54.7 (d)
54.7 (d)
54.1 (d)
343
325
340
348
341
315
329
338
a Measured in Hz from corresponding 31P spectrum or (3a,b) 1H NMR spectrum. b 202.4 MHz, samples in d6-benzene. c 202.4 MHz, samples in
1
d1-chloroform; for both 3a,b, δPF6 ) -143.6 ppm, JPF ) 713 Hz. d 145.8 MHz, samples in d8-toluene. e Recorded at 240 K. f Recorded at 250 K.
g Recorded at 260 K.
Scheme 2
analysis of the temperature dependence of Keq for (1 + pyridine)
) 7 gave the following thermodynamic parameters (listed for
a, b): ∆H° (kcal/mol) ) -8.4, -7.6; ∆S° (cal/K/mol) ) -14,
-10; ∆G°(240 K) (kcal/mol) ) -4.8, -5.0. These data point
to entropic control of the equilibrium between 1 and 7: loss of
entropy causes the equilibrium to shift away from adduct
Figure 1. Molecular structure of 2a. Non-hydrogen atoms are
represented by Gaussian ellipsoids at the 20% probability level.
The hydrogen atoms attached to P1 and C10 are shown with
formation at higher temperatures, although adduct 7 is enthal-
pically favored over (1 + free pyridine).24 We presume the
decreased stability of 7, relative to 6, stems from the consider-
able crowding at Ru in these N-bound adducts.
arbitrarily small thermal parameters; all other hydrogens are not
shown. Selected interatomic distances (Å) and bond angles (deg)
(C* denotes the centroid of the plane defined by C(7A)-C(1)-C(2)-
C(3)-C(3A)): Ru-P(1) ) 2.2512(7), Ru-P(2) ) 2.2969(7),
Ru-C10 ) 2.178(3), Ru-C* ) 1.925, ∆ ) 0.115, C(10)-C(11)
) 1.452(4), C(11)-N ) 1.141(4); P(1)-Ru-P(2) ) 91.57(2),
When the reaction of acetonitrile with 1 is monitored at low
temperature by 31P{1H} NMR, we do observe transient signals
due to the N-bound acetonitrile adducts, 4 (Table 1). This raises
the question of whether 4 is an intermediate in the formation
of the metalation product, 2. Intramolecular proton abstraction
by the phosphido ligand from N-bound acetonitrile seems
unlikely, since the linear, N-bound nitrile functionality places
the acetonitrile C-H bonds quite far from the Ru-P bond in
4. Also, the equilibria we observe between 1 and its benzonitrile
(6) and pyridine (7) adducts suggest that the acetonitrile ligand
in 4 is probably quite labile, so our observation of 4 does not
preclude intermolecular activation by 1 of the C-H bond in
free acetonitrile. To gain further insight into the possible
involvement of 4 in the formation of 2, we monitored by
31P{1H} NMR the reaction of KOBut with the cationic, N-bound
acetonitrile complex 3a (Scheme 3, an alternate synthetic route
P(1)-Ru-C(10)
) 86.40(7), P(2)-Ru-C(10) ) 90.05(7),
P(1)-Ru-C* ) 127.8, P(2)-Ru-C* ) 125.9, C(10)-Ru-C* )
123.3, Ru-C(10)-C(11) ) 112.81(19), C(10)-C(11)-N )
179.2(3). Indenyl slip distortion: ∆ ) d[Ru-C(7A),C(3A)] -
d[Ru-C(1),C(3)] ) 0.115 Å.
to 2a). Initial spectra indicate the rapid, complete consumption
of 3a, with formation of both 1a (minor) and 2a (major), but
there is no sign of 4a.25 This is consistent with facile dissociation
of acetonitrile early in the reaction, followed by formation of
1a and its intermolecular reaction with free acetonitrile, but does
not rule out either (i) rapid consumption of transient 4a, whose
concentration would be low in the absence of excess acetonitrile
in the mixture, or (ii) a competitive pathway in which the
acetonitrile, rather than the secondary phosphine in 3a, is
(24) The exchange of pyridine between the bulk mixture and 7 was
sufficiently slow on the NMR time scale to allow line-shape analysis of
31P{1H} NMR spectra recorded at different temperatures and calculation
of activation parameters for this exchange (conversion of 1a,b to 7a,b,
respectively). Eyring data: ∆Hq(kcal/mol) ) 8.8, 6.7; ∆Sq(cal/K/mol) )-
16,-23; ∆Gq(240 K) (kcal/mol) ) 13, 13. Arrhenius data: Ea(kcal/mol) )
9.3, 7.2. See Supporting Information for full details.
(25) The identity of these products early in the reaction is certain;
however, low solubility of 3a and KOBut in d8-toluene and the production
of gelatinous ButOH and solid KPF6 in the NMR sample render this
experiment a bit messy and difficult to monitor to completion.

5028 Organometallics, Vol. 27, No. 19, 2008
Derrah et al.
deprotonated by KOBut. Initial 31P{1H} NMR spectra of the
analogous reaction using the CD3CN analogue of 3a show
mostly unreacted d3-3a, no 4a, a small amount of 1a, and just
a tiny amount of 2a.26 While this experiment points qualitatively
to a deuterium isotope effect on the rate of metalation of
acetonitrile, again it does not preclude the intermediacy of 4a
in the reaction. More compelling evidence for an intermolecular
pathway from 1 to 2 comes from the addition of a 1:1 mixture
of d3- and d0-acetonitrile to 1a,b. The relative amount of
deuterium incorporated into the P-H position in the products
2a,b in these reactions indicates kinetic isotope effects (2.0 and
3.2, respectively) consistent with the importance of the C-H
bond breaking in the transition state of the reaction. 31P{1H}
NMR spectra of these product mixtures clearly indicate the
presence of at least four isotopomers (attributable to d0,1,2,3-2),
whereas only two (pure d3- and d0-2) would be expected for
the intramolecular metalation of acetonitrile via adduct 4.
Scheme 4
Scheme 3
to this intermediate are currently under computational investiga-
tion. If kinetically accessible, it provides a convincing picture
for a competition between reversible binding of acetonitrile and
its intermolecular activation by 1 (Scheme 4), in which the
outcome is dictated by high stability of the C-H metalated
product, 2.
The formation of complex 2 demonstrates much higher
Brønsted basicity at the phosphido ligand in 1 than we expected.
Although P-nucleophilicity/basicity is an established feature of
terminal phosphido complexes of Ru and other late metals,21,27
and we previously observed protonation of the planar PR2 group
in 1a by HCl and NH4Cl (and its methylation, by MeI);13
nevertheless, the observed deprotonation of acetonitrile suggests
that pKa values for conjugate acids of 1a,b are extraordinarily
high, given the participation of the lone pair on this phosphido
ligand in a π-bond with ruthenium. They approach those
reported for the conjugate acids of pyramidal phosphido
ruthenium complexes [Ru(η5-Cp)(PR2)(PR′3)2] (estimated pKa
(H2O or THF) ∼20-25), in which PR3′ are highly Lewis basic
trialkylphosphines.21 It is possible, in these reactions of 1, that
a 16-electron intermediate with a single Ru-P bond and
pyramidal geometry at the phosphido P is key to the high
basicity we observe, as well as providing the open coordination
site that leads to adduct formation at Ru (Scheme 4). Preliminary
DFT calculations,28 examining the trajectory of formation of
5a from 1a, have uncovered such an intermediate, at ap-
proximately 6.7 kcal/mol higher energy than 1a, with the sum
of angles at PCy2 of 340.6° and a Ru-PCy2 distance of 2.211
Å, relative to 2.166 Å calculated for 1a.13 Hypothetical pathways
In the context of the high P-basicity implied by the formation
of 2, it is not surprising that protic O-donor solvents H2O and
MeOH (pKa ) 15.74 and 15.5, respectively5) are also quanti-
tatively deprotonated by 1 (Scheme 5), giving the corresponding
hydroxide (8-H) and methoxide (8-Me) complexes (Table 1).
(The latter rapidly decarbonylates, even at low temperature, to
give the hydride complex 9.29) Addition of excess ButOH (pKa
) 19.25) to 1a,b gives only partial conversion (15% (a), 36%
(b)) to the corresponding tert-butoxide complex 8-But (Scheme
5, Table 1), along with several other unidentified products, as
determined by 31P{1H} NMR. This result is surprising, since
the reaction of 1 with acetonitrile (a weaker acid) goes to
completion, but it can be rationalized by the size of the tert-
butoxy group, which presumably disfavors formation of the
Ru-O bond at this crowded ruthenium center. We see no
evidence in these reactions for the formation of O-bound adducts
of the neutral ROH. This may reflect the relatively high rate of
proton abstraction from these reagents by 1, but it is also
consistent with our observation that diethyl ether and THF,
which have comparable O-donor ability30 to water and methanol
(and are presumably much weaker C-H acids), do not react
with 1, as long as they are rigorously dried. Acetone also has
comparable O-donor ability to water,30 but is a relatively strong
C-H acid (pKa ) 19):5 it reacts with 1 to give ∼80% of the
C-H metalated product 10, identified by 31P NMR at low
(26) When the reaction of d3-3a with KOBut is carried out on a synthetic
scale and allowed to go to completion, the solid 2a isolated is a mixture of
d3- and d2-isotopomers, the latter containing P-H instead of P-D bonds
(see Experimental Section and Supporting Information). This may indicate
the non-innocence of ButOH initially formed or the direct abstraction of
deuterium by KOBut from d3-acetonitrile.
(27) See Chan, V. S.; Stewart, I. C.; Bergman, R. G.; Toste, F. D. J. Am.
Chem. Soc. 2006, 128, 2786 and Scriban, C.; Glueck, D. S. J. Am. Chem.
Soc. 2006, 128, 2788 and references therein.
(29) Complex 9a was prepared previously by the addition of methanolic
sodium methoxide to [Ru(η5-indenyl)Cl(HPR2)(PPh3)] (ref 32). Complex
9b was identified spectroscopically; 31P{1H} (145.78 MHz, d8-toluene, δ):
2
(28) Pantazis, D. A., personal communication.
72.2 (d, JPP ) 28 Hz, HPPri2), 70.0 (d, PPh3).

Ruthenated Acetonitrile
Organometallics, Vol. 27, No. 19, 2008 5029
Scheme 5
Deuterated solvents were purchased from Canadian Isotope Labo-
ratories (CIL), freeze-pump-thaw degassed, and vacuum trans-
ferred from sodium/benzophenone (d6-benzene, d8-toluene) or
calcium hydride (d-chloroform, d3-acetonitrile) before use. [RuCl(η5-
indenyl)(HPCy2)(PPh3)],32[RuCl(η5-indenyl)(HPPri2)(PPh3)],13[Ru(η5-
indenyl)(PCy2)(PPh3)] (1a),13 [Ru(η5-indenyl)(PPri2)(PPh3)2] (1b),13
and [Ru(η5-indenyl)(PCy2)(CO)(PPh3)] (5a)13 were prepared as
previously reported. NMR spectra were recorded on a Bruker
1
AVANCE 500 operating at 500.13 MHz for H, 125.77 MHz for
13C, and 202.46 MHz for 31P, or on a Bruker AVANCE 360
1
2
operating at 360.13 MHz for H, 55.28 MHz for H, and 145.78
MHz for 31P. Chemical shifts are reported in ppm at ambient
temperature unless otherwise noted. H chemical shifts are refer-
1
enced against residual protonated solvent peaks at 7.16 ppm
(C6D5H), 2.09 ppm (PhCD2H), and 7.24 ppm (CHCl3). 2H chemical
shifts are referenced against trace C6H5D in benzene. 13C chemical
shifts are referenced against d6-benzene at 128.4 ppm and d-
Scheme 6
chloroform at 77.5 ppm. All H, H, and 13C chemical shifts are
reported relative to tetramethylsilane, and 31P chemical shifts are
reported relative to 85% H3PO4(aq). Microanalysis was performed
by Canadian Microanalytical Service Ltd., Delta, BC, Canada. IR
spectra were recorded on a Perkin-Elmer FTIR Spectrum One
spectrophotometer using KBr pellets under a nitrogen atmosphere.
Mass spectrometry was carried out by Dr. David McGillivray at
the Department of Chemistry, University of Victoria.
1
2
temperature (Scheme 6, Table 1).31 At temperatures above 280
K, mixtures of 1 and 10 decompose to a number of unidentified
products.
[Ru(η5-indenyl)(CH2CN)(HPCy2)(PPh3)] (2a). To a Schlenk
flask containing a blue solution of [Ru(η5-indenyl)(PCy2)(PPh3)2]
(100 mg, 0.15 mmol) in toluene (5 mL) was added acetonitrile (0.8
mL, 0.6 g, 20 mmol). The resulting orange solution was allowed
to stir for 1.5 h before the solvent was evaporated under vacuum
to give a dark orange oil. Recrystallization from toluene (2 mL)
and pentane (20 mL) at -30 °C gave a dark brown crystalline solid
(2a, 56 mg, 0.074 mmol, 53% yield) along with two unidentified
complexes as minor impurities (<5%). The sample was recrystal-
lized a second time from toluene (1-2 mL) by the slow vapor
diffusion of pentane (10 mL) to give 2a (44 mg, 0.061 mmol, 79%
yield) in >99% purity. IR (KBr, cm-1): 2332 (w, νP-H), 2179
(s, νCN). FAB-MS (+LSIMS matrix mNBA; m/z (relative intensity):
717.3 (8%) [M+], 677.3 (83%) [M+ - CH2CN], 479.1 (100%)
[M+ - CH2CN - HPCy2]. HR-MS (+LSIMS matrix mNBA):
exact mass (monoisotopic) calcd for C41H47NP2Ru 717.2227; found
717.2227 ( 0.0020 (average of 3 trials). Anal. Calcd for
C41H47NP2Ru: C, 68.70; H, 6.61. Found: C, 67.52; H, 6.47 (see
Supporting Information for 1H and 31P{1H} NMR). Dec pt:
196-198 °C.
Conclusions
The above reactions of a range of N- and O-donor solvents
demonstrate a delicate balance between Lewis acidity at
ruthenium and Brønsted basicity at the phosphido phosphorus
in 1. This is highlighted by the fact that we are able to observe
the N-bound acetonitrile adduct 4 at low temperatures and “trap”
the methylenenitrile anion in 2. We are currently investigating
reactions of 2 that might allow further transformation of this
functional anion via the Ru-C bond. Future work will focus
on more generally exploiting the cooperation between the inner
and outer coordination sphere at ruthenium in 1, which is
provided by the variable coordination mode and basicity of the
terminal phosphido ligand in this system.
Experimental Section
[Ru(η5-indenyl)(CH2CN)(HPPri2)(PPh3)] (2b). To a Schlenk
flask containing a blue solution of [Ru(η5-indenyl)(PPri2)(PPh3)2]
(100 mg, 0.17 mmol) in toluene (5 mL) was added acetonitrile (0.8
mL, 0.6 g, 20 mmol). The resulting orange solution was allowed
to stir for 1.5 h before the solvent was evaporated under vacuum
to give a dark yellow oil. Pentane (10 mL) was added, and the
mixture was allowed to stir for 3 h. The resulting yellow suspension
was filtered and washed with pentane (3 × 10 mL) to give [Ru(η5-
indenyl)(CH2CN)(HPPri2)(PPh3)2] (2b) (76 mg, 0.12 mmol, 71%
yield) along with two unidentified complexes as minor impurities
(<5%). Crude 2b (61 mg, 0.10 mmol) was recrystallized by the
vapor diffusion of pentane (10 mL) into a toluene solution (1 mL)
to give an orange crystalline solid (30 mg, 0.05 mmol, 49% yield).
IR (KBr, cm-1): 2296 (m, νPH), 2181 (s, νCN). FAB-MS (+LSIMS
matrix mNBA; m/z (relative intensity): 637.2 (8%) [M+], 597.2
(70%) [M+ - CH2CN], 479.1 (100%) [M+ - CH2CN - HPPri2].
HR-MS (+LSIMS matrix mNBA): exact mass (monoisotopic) calcd
for C35H39NP2Ru 637.1601; found 637.1591 ( 0.0008 (average of
3 trials). Anal. Calcd for C35H39NP2Ru: C, 66.02; H, 6.17. Found:
C, 65.88; H, 6.17. Dec pt: 168-169 °C.
General Details. Unless otherwise noted, all reactions and
manipulations were performed under nitrogen in an MBraun Unilab
1200/780 glovebox or using conventional Schlenk techniques. All
solvents were sparged with nitrogen for 25 min and dried using an
MBraun solvent purification system (SPS) except: benzene was
freeze-pump-thaw degassed and vacuum transferred from sodium/
benzophenone prior to use; acetone was dried over K2CO3 prior to
distillation under nitrogen; pyridine was dried over MgSO4 prior
to distillation under nitrogen; benzonitrile was predried with K2CO3,
then fractionally distilled from P2O5 under dynamic vacuum; ButOH
was dried over Na and a small amount of benzene was added prior
to fractional distillation under nitrogen to remove remaining water
as a tertiary azeotrope; H2O was freeze-pump-thaw degassed;
and Et2O was dried over CaH2 prior to distillation under nitrogen.
(30) O-Donor strengths vary as DN ) 20 (THF) > 19.2 (Et2O) > 19.0
(MeOH) > 18.0 (water, although this is much higher, 33, for bulk water)
>17 (acetone). In ref 23, the authors note that, although overall trends in
DN for any given heteroatom donor type are valid, comparisons between
different heteroatom donors are not reliable.
(31) The C-H metalation of acetone at ruthenium has been reported
for: (a) [RuH (naphthyl)(dmpe)2], with loss of naphthalene (ref 9); (b)
[Ru(η5-Cp*)(OH)(PMe3)2], with loss of H2O: Bryndza, H. E.; Fong, L. K.;
Paciello, R. A.; Tam, W.; Bercaw, J. E. J. Am. Chem. Soc. 1987, 109, 1444.
(32) Derrah, E. J.; Marlinga, J. C.; Mitra, D.; Friesen, D. M.; Hall, S. A.;
McDonald, R.; Rosenberg, L. Organometallics 2005, 24, 5817.

5030 Organometallics, Vol. 27, No. 19, 2008
Derrah et al.
[Ru(η5-indenyl)(NCCH3)(HPCy2)(PPh3)][PF6] (3a). To a
Schlenk flask containing an orange suspension of [Ru(η5-
indenyl)Cl(HPCy2)(PPh3)] (215 mg, 0.30 mmol) and [NH4][PF6]
(148 mg, 0.90 mmol, 3 equiv) in methanol (5 mL) was added
acetonitrile (2.0 mL, 1.6 g, 38 mmol). No immediate reaction was
observed, but after the mixture had stirred for 2 days a yellow
precipitate formed, which was isolated by filtration, then redissolved
in CH2Cl2 (5 mL). The resulting yellow solution was filtered to
remove unreacted [NH4][PF6], and the volume was reduced (50%)
under vacuum. Hexanes (40 mL) was added, which gave 3a (218
mg, 0.25 mmol, 84% yield) as an orange crystalline solid. IR (KBr,
cm-1): 2334 (w, νPH), 2272 (w, νCN). FAB-MS (+LSIMS matrix
mNBA; m/z (relative intensity): 718.2 (8%) [M+], 677.0 (100%)
[M+ - CH3CN], 478.9 (33%) [M+ - CH3CN - HPCy2]. HR-MS
(+LSIMS matrix mNBA): exact mass (monoisotopic) calcd for
C41H48NP2Ru 718.2306; found 718.2308 ( 0.0012 (average of 3
trials). Anal. Calcd for C41H48F6P3Ru · 0.17CH2Cl2: C, 55.56; H,
5.49. Found: C, 55.31; H, 5.39 (see Supporting Information for 1H
NMR). Dec pt: 144-146 °C.
valve adaptor, removed from the glovebox, and connected to a
Schlenk line. Each sample was degassed using three freeze-
pump-thaw cycles, then frozen in N2(l). Approximately 0.1 mL
of the appropriate reagent (1-5 mmol) was transferred under
vacuum (acetonitrile, ButOH, methanol, ether, THF, acetone) or
delivered by syringe (1:1 acetonitrile/d3-acetonitrile, pyridine,
benzonitrile, H2O), and the tube was flame-sealed. The thawed
solution was shaken to mix the reagents before the tube was placed
in the NMR spectrometer. For reactions monitored at low temper-
ature, the solutions were thawed sufficiently to allow mixing and
kept cold in an acetone/dry ice bath before being placed in the
precooled spectrometer.
(A) Addition of Acetonitrile Monitored at Room Tempera-
ture. The mixed reagents gave a dark yellow-orange solution,
indicative of the formation of [Ru(η5-indenyl)(CH2CN)(HPR2)-
(PPh3)] (2a,b), prior to being placed in the NMR spectrometer.
For 1a: 31P{1H} NMR shows 2a as the major product (93%),
along with six or seven minor, unidentified products (7%).
For 1b: 31P{1H} NMR shows 2b as the major product (93%),
Alternate Preparation of 2a via 3a. To a Schlenk flask
containing a yellow suspension of [Ru(η5-indenyl)(NCCH3)-
(HPCy2)(PPh3)][PF6] (3a) (52 mg, 0.040 mmol) in toluene (5 mL)
was added KOtBu (8 mg, 0.072 mmol, 1.2 equiv). The resulting
green solution was stirred for 24 h, during which time it turned
black-yellow. After 5 days the solution turned yellow, at which
point the solvent was removed under vacuum. The resulting yellow
oil was dissolved in a minimal amount of hexanes (5 mL), and a
yellow precipitate formed after 24 h. A yellow powder (ap-
proximately 10 mg) was isolated by filtration, dried under vacuum,
and confirmed to be 2a by NMR spectroscopy. This reaction was
repeated using d3-3a (see below), giving a similar yield of a yellow
powder whose 1H and 2H NMR spectra are consistent with a
mixture of isotopomers of 2a.26
along with five or six minor, unidentified products (7%).
(B) Addition of Acetonitrile Monitored at Low Tempera-
tures. The mixed reagents gave a dark red solution, prior to being
placed in the NMR spectrometer, which was precooled to 240 K.
When the sample was removed from the spectrometer at room
temperature, the solution was dark yellow-orange.
For 1a: 31P{1H} NMR spectrum at 240 K shows 2a as the major
product (∼45%), along with unreacted 1a (∼33%), and peaks
consistent with the minor product [Ru(η5-indenyl)(NCCH3)(PCy2)-
(PPh3)] (4a, ∼22%) (Table 1).
For 1b: 31P{1H} NMR spectrum at 240 K shows 2b as the major
product (∼50%), along with unreacted 1b (∼18%), and peaks
consistent with the minor product [Ru(η5-indenyl)(NCCH3)-
(PPri2)(PPh3)] (4b, ∼32%) (Table 1).
[Ru(η5-indenyl)(NCCD3)(HPCy2)(PPh3)][PF6] (d3-3a). This
complex was prepared as described above for 3a, using [Ru(η5-
indenyl)Cl(HPCy2)(PPh3)] (315 mg, 0.44 mmol) and [NH4][PF6]
(231 mg, 1.42 mmol, 3.2 equiv) in methanol (10 mL), and
d3-acetonitrile (1.0 mL, 0.8 g, 20 mmol). The product, d3-3a, was
isolated as a yellow powder (312 mg, 0.36 mmol, 82% yield), and
For both samples, 31P{1H} NMR spectra above 240 K showed
rapid conversion of mixtures of 1 and 4 to the metalated prod-
uct 2.
(C) Addition of 1:1 Acetonitrile/d3-Acetonitrile at Room
Temperature. Two samples were prepared: one in d8-toluene and
one in d0-toluene for analysis by 2H NMR. In both cases, the mixed
reagents gave a dark yellow-orange solution, indicative of the
formation of [Ru(η5-indenyl)(CH2CN)(HPR2)(PPh3)] (2a,b) and/
or its deuterated derivative(s), prior to being placed in the NMR
1
its identity was confirmed by H and 31P{1H} NMR.
[Ru(η5-indenyl)(NCCH3)(HPPri2)(PPh3)][PF6] (3b). To a
Schlenk flask containing an orange solution of [Ru(η5-
indenyl)Cl(HPPri2)(PPh3)] (205 mg, 0.32 mmol) and [NH4][PF6]
(158 mg, 0.97 mmol, 3 equiv) in MeOH (5 mL) was added
acetonitrile (2 mL, 1.6 g, 38 mmol). No immediate reaction was
observed, but after the mixture had stirred for 2 days a yellow
solution formed. The solvent was removed under vacuum, and the
resulting yellow oil was redissolved in CH2Cl2 (5 mL), which was
filtered to remove unreacted [NH4][PF6]. The volume was reduced
(50%) under vacuum, and hexanes (20 mL) was added to give 182
mg (0.23 mmol, 72% yield) of 3b as an orange crystalline solid. A
portion (100 mg, 0.13 mmol) was recrystallized a second time from
CH2Cl2 (2 mL) and hexanes (20 mL) to give analytically pure 3b
(68 mg, 0.086 mmol, 68% yield). IR (KBr, cm-1): 2343 (w, P-H),
2272 (w, CN); FAB-MS (+LSIMS matrix mNBA; m/z (relative
intensity): 638.2 (8%) [M+], 597.2 (100%) [M+ - CH3CN], 479.1
(65%) [M+ - CH3CN - HPPri2]. HR-MS (+LSIMS matrix
mNBA): exact mass (monoisotopic) calcd for C35H40NP2Ru 638.1680;
found 638.1698 ( 0.0008 (average of 3 trials). Anal. Calcd for
1
2
spectrometer. The H, H{1H}, and 31P{1H} NMR spectra con-
firmed that a mixture of deuterated and nondeuterated 2a,b had
1
formed. The H NMR spectra were used to determine kH/kD (2a,
2.0; 2b, 3.2) by comparing the ratio between the P-H (2a, 3.86
ppm; 2b, 3.98 ppm) and the indenyl H2 (5.56, 5.45 ppm) peaks.
(D) Addition of Pyridine Monitored at Low Temperatures.
Hexamethylbenzene (a, 35.1 mg, 0.22 mmol; b, 34.8 mg, 0.21
mmol) and triphenylphosphine oxide (a, 6.4 mg, 0.023 mmol; b,
6.7 mg, 0.024 mmol) were added to the sealable NMR tube as 1H
and 31P NMR standards, respectively. Prior to the addition of
pyridine the samples were thoroughly mixed by sonication to ensure
all species dissolved. For both 1a,b, the flame-sealed sample was
dark red when removed from the acetone/dry ice bath but quickly
(∼15 s) turned blue as it was placed in the NMR spectrometer,
which was precooled to 210 K (a) or 230 K (b). The temperature
was immediately lowered to 190 K, at which only peaks due to
the coordination products [Ru(η5-indenyl)(pyridine)(PR2)(PPh3)]
(6a,b) were observed by 31P{1H} NMR. As the temperature was
increased in 10 K intervals, these peaks decreased in intensity, while
those due to the phosphido complexes (1a,b) increased in intensity,
until the temperature was raised to 290 K, at which point the
samples decomposed to 10-15 unidentified products. The resulting
data were used to calculate Keq at 230-260 K (see Supporting
Information).
1
C35H40F6NP3Ru · 0.20CH2Cl2 (see Supporting Information for H
NMR): C, 52.87; H, 5.09. Found: C, 53.23; H, 5.30. Mp: 160-162
°C.
NMR-Scale Experiments Using 1a,b.33 In the glovebox solid
1 (a: 15 mg, 0.022 mmol; b: 15 mg, 0.025 mmol) was placed in a
sealable NMR tube with d8-toluene (0.6 mL). In some cases, internal
standard hexamethylbenzene (1H) or triphenylphosphine oxide (31P)
was also added. The tube was then capped with a Teflon needle

Ruthenated Acetonitrile
Organometallics, Vol. 27, No. 19, 2008 5031
For 1a: 31P{1H} NMR shows peaks due to a major product
(93%), assigned as [Ru(η5-indenyl)(OH)(HPCy2)(PPh3)] (8a-H), a
small amount of unreacted 1a (2%), and three or four other
unidentified products (5%).
For 1b: 31P{1H} NMR shows peaks due to a major product
(90%), assigned as [Ru(η5-indenyl)(OH)(HPiPr2)(PPh3)] (8b-H), a
small amount of unreacted 1b (2%), and two other unidentified
products (8%).
(G) Addition of Methanol Monitored at Low Tempera-
tures. For both samples, the thawed solution turned red, before
being placed in the NMR spectrometer, which was precooled to
240 K. Complete conversion to a mixture of [Ru(η5-indenyl)-
(OMe)(HPR2)(PPh3)] (8-Me) and [Ru(η5-indenyl)(H)(HPR2)-
(PPh3)] (9) was immediate, with no residual 1 observed by 31P{1H}
NMR. As the temperature was increased in 10 K increments to
RT, signals for 9 increased in intensity at the expense of those for
8-Me, until only 9 remained, along with ∼2% unidentified minor
products. Both solutions were yellow when the warmed samples
were removed from the spectrometer.
(H) Addition of tert-Butanol Monitored at Room Tempera-
ture. For 1a: The mixed reagents retained the blue color of 1a.
31P{1H} NMR shows mainly unreacted 1a (77%), with peaks due
to a minor product (15%) assigned as [Ru(η5-indenyl)(OBut)-
(HPCy2)(PPh3)] (8a-But), and five or six other unidentified products
(8%).
For 1b: The mixed reagents gave a red solution. 31P{1H} NMR
shows mainly unreacted 1b (50%), with peaks due to a minor
product (36%) assigned as [Ru(η5-indenyl)(OBut)(HPPri2)(PPh3)]
(8b-But), and three or four other unidentified products (14%).
(I) Additions of Diethyl Ether and THF to 1a,b Monitored
at Room Temperature. The mixed reagents retained the blue color
of 1a,b. 31P{1H} NMR confirmed the continued presence of 1, with
only minor shift deviations (<0.5 ppm) arising from the added
solvent.
(J) Addition of Acetone to 1a,b Monitored at Low Temper-
atures. In both cases, there was no immediate change in color as
the mixture was thawed and placed in the precooled NMR
spectrometer. 31P{1H} NMR showed that, at 240 K, the samples
were mainly unreacted 1a,b (82-86%), along with peaks assigned
to [Ru(η5-indenyl)(CH2COCH3)(HPR2)(PPh3)] (10a,b, 14-15%),
and one or two other unidentified products (1-3%). As the
temperature was increased to 280 K in 10 K intervals, the intensity
of peaks due to 10a,b increased to 70-77%, at the expense of those
due to 1a,b. Upon further warming, the sample rapidly decomposed,
giving (by 31P{1H} NMR) nine or 10 unidentified products at room
temperature. After 1-2 days the sample was dark red and had
further decomposed to a number of unidentified products, including
some major species with peaks in the µ-phosphido region (240-100
ppm), with no residual 1a,b or 10a,b.
Table 2. Crystallographic Experimental Details for 2a
formula
C41H47NP2Ru
716.81
fw
cryst dimens (mm)
cryst syst; space group
unit cell params
0.32 × 0.17 × 0.14
orthorhombic; P212121 (No. 19)
a (Å) 12.3804(9)
b (Å) 15.0022(11)
c (Å) 18.9812(13)
V (Å3) 3525.4(4)
Z 4
1.351
0.565
54.88
30 397 (-16 e h e 15, -19 e
k e 19, -24 e l e 24)
8047 (Rint ) 0.0443)
7183
0.9251-0.8399
8047/0/410
-0.03 (2)
1.053
0.0301
0.0675
Fcalcd (g cm-3
µ (mm-1
)
)
data collection 2θ limit (deg)
total no. of data collected
no. of indep reflns
no. of obsd reflns [I g 2σ(I)]
range of transmn factors
no. of data/restraints/params
Flack absolute struct parama
goodness-of-fit (S) [all data]b
R1 [I g 2σ(I)]b
wR2 [all data]b
largest diff peak and hole (e Å-3
)
0.560 and -0.200
a Flack, H. D. Acta Crystallogr. 1983, A39, 876–881; Flack, H. D.;
Bernardinelli, G. Acta Crystallogr. 1999, A55, 908–915; Flack, H. D.;
Bernardinelli, G. J. Appl. Cryst. 2000, 33, 1143–1148. b S ) [Σw(Fo
-
2
Fc ) /(n - p)]1/2 (n ) number of data; p ) number of parameters
2 2
varied; w ) [σ2(Fo2) + (0.0350P)2]-1 where P ) [Max(Fo2, 0) + 2Fc ]/
2
2
2 2
3); RI ) Σ||Fo| - |Fc||/Σ|Fo|; wR2 ) [Σw(Fo - Fc ) /Σw(Fo4)]1/2
.
For 1a: 31P{1H} NMR spectrum at 240 K shows 68% 1a and
32% 6a.
For 1b: 31P{1H} NMR spectrum at 240 K shows 48% 1b and
52% 6b.
(E) Addition of Benzonitrile Monitored at Low Tempera-
tures. Hexamethylbenzene (a, 37.4 mg, 0.23 mmol; b, 34.0 mg,
0.21 mmol) and triphenylphosphine oxide (a, 11.6 mg, 0.042 mmol;
1
b, 6.6 mg, 0.024 mmol) were used as internal H and 31P NMR
standards, respectively. Prior to the addition of benzonitrile the
samples were thoroughly mixed by sonication to ensure all species
dissolved. The flame-sealed samples were red when removed from
the acetone/dry ice bath but rapidly turned burgundy while being
transferred to the spectrometer, which was precooled to 230 K.
For both samples at 230 K, peaks due to the benzonitrile adducts
[Ru(η5-indenyl)(PR2)(NCPh)(PPh3)] (7a,b) dominated the 31P{1H}
NMR spectra, but as the temperature was increased in 10 K intervals
to room temperature, small peaks due to the phosphido complex
(1a,b) appeared. No decomposition was observed at room temper-
ature. The changes in relative intensities of 1a,b and 6a,b over
this temperature range were not sufficiently large to calculate
reliably Keq for the adduct formation.
For 1a: 31P{1H} NMR at 250 K shows 60% 7a, 16% 1a, and
24% [Ru(η5-indenyl)(NH2CH2Ph)(HPCy2)(PPh3)].34
For 1a at 260 K: 31P{1H} NMR shows unreacted 1a (∼52%),
10a (∼43%), and one or two unidentified products (∼5%).
For 1b at 250 K: 31P{1H} NMR shows 10b (∼60%), unreacted
1b (∼42%), and an unidentified product (∼10%).
For 1b: 31P{1H} NMR at 250 K shows 68% 7b, 18% 1b, and
14% [Ru(η5-indenyl)(NH2CH2Ph)(HPCy2)(PPh3)].34
(F) Addition of Water Monitored at Room Temperature. For
both samples, the mixed reagents gave a dark red solution.
NMR-Scale Reactions of 3a and d3-3a with KOBut. Solid 3a
(15 mg, 0.018 mmol), KOBut (5 mg, 0.045 mmol, 2.5 equiv), and
OdPPh3 (7.2 mg, 0.026 mmol, internal standard) were placed in a
sealable NMR tube. d8-Toluene (0.8 mL) was transferred under
vacuum, and the tube was flame-sealed. The thawed solution was
shaken to mix the reagents before being placed in the NMR
spectrometer, where the progress of reaction was monitored by
31P{1H} NMR spectroscopy at 15 min intervals for 1 h.
NMR-Scale Experiments Using 5a. These sealed samples were
prepared using the same general method described above for 1a,b.
(A) Addition of Acetonitrile. Solid 5a (15 mg, 0.021 mmol)
was used. At room temperature, over a 25 h period, the mixed
reagents maintained the dark red color of 5a and no change was
observed in the 31P{1H} NMR spectrum. Even heating of the
(33) For these reactions, amounts of unreacted 1a,b include 12% or 16%
that exist as the phosphaalkene structural isomers (see ref 13).
(34) Peaks due to
a
species tentatively assigned as [Ru(η5-
indenyl)(NH2CH2Ph)(HPR2)(PPh3)] probably resulted from 1,2-addition to
1a,b of R-aminotoluene, a known impurity in benzonitrile (Perrin, D. D.;
Armarego, W. L. F. Purification of Laboratory Chemicals, 3rd ed.;
Pergamon Press: Toronto, 1988; p 94). Consistent with this explanation is
the observed decrease in relative intensity of these peaks (to ∼3%) when
the reactions were repeated (at room temperature) with doubly distilled
benzonitrile. However the 13C-DEPT 135 NMR spectra of the reaction
mixtures showed no benzylic carbons, and no R-aminotoluene was observed
in the 1H NMR spectra of purified benzonitrile: 1a,b is apparently
scavenging the trace amounts of R-aminotoluene present in the excess
benzonitrile added.

5032 Organometallics, Vol. 27, No. 19, 2008
Derrah et al.
sample at 65 °C for 4.5 h gave no spectroscopic changes, consistent
with the irreversible coordination of CO in this complex.
completed using the program SHELXL-97.37 The hydrogen atom
attached to P1 was located and freely refined, while hydrogen atoms
attached to carbons were assigned positions based on the sp2 or
sp3 hybridization geometries of their attached atoms and were given
thermal parameters 20% greater than those of the parent atoms.
See Table 2 for a summary of crystallographic experimental
parameters.
(B) Addition of 1:1 Acetonitrile/d3-Acetonitrile. Portions of
solid 5a (12 mg, 0.017 mmol) were used to prepare two sealed
samples: one in d8-toluene and one in d0-toluene for analysis by
1
2
2H NMR. The H and H NMR spectra were monitored over a 4
day period for signs of proton/deuteron exchange within the
acetonitrile. Only a trace amount of CHD2CN was observed
(a triplet and a doublet in the 1H and 2H{1H} NMR spectra,
respectively), consistent with the standard 0.2% residual proton
impurity in the commercial d3-acetonitrile used.
Acknowledgment. This work was supported by the
Natural Sciences and Engineering Research Council (NSERC)
of Canada.
Crystallographic Study of 2a. Orange crystals of 2a were grown
from diffusion of n-pentane into a toluene solution of the compound.
Data were collected on a Bruker PLATFORM/SMART 1000 CCD
diffractometer35 using Mo KR radiation (λ ) 0.71073 Å) at -80
°C. Unit cell parameters were obtained from a least-squares
refinement of the setting angles of 6093 reflections from the data
collection with 4.78° < 2θ < 51.80°. The space group was
determined to be P212121 (No. 19). The data were corrected for
absorption through use of the SADABS procedure. The structure
of 2a was solved using the Patterson serach and structure expansion
facilities within the DIRDIF-99 program system.36 Refinement was
Supporting Information Available: 1H and 13C NMR data for
2a,b and 3a,b; spectra illustrating sample purity, structure diagnosis
and isotope labeling; details of line shape analyses and van’t Hoff
calculations, and a crystallographic information file (CIF). This
materialisavailablefreeofchargeviatheInternetathttp://pubs.acs.org.
OM800343X
(36) Beurskens, P. T. ; Beurskens, G.; de.Gelder, R.; Garcia.Granda,
S.; Israel, R.; Gould, R. O.; Smits, J. M. M. The DIRDIF-99 Program
System; Crystallography Laboratory, University of Nijmegen: The Neth-
erlands, 1999.
(35) Programs for diffractometer operation, data collection, data reduc-
tion, and absorption correction were those supplied by Bruker.
(37) Sheldrick, G. M. Acta Crystallogr. 2008, A64, 112–122.