Chiral and Achiral Diphosphine Complexes of Ruthenium(II) Incorporating Labile Nitrile Ligands: Synthesis and Solution Chemistry of Mono- and Dinuclear Derivatives of Ru2Cl4(PP)2 (PP = Chelating Diphosphine)

Article abstract of DOI:10.1021/ic960735y

A family of nitrile complexes has been prepared by reaction of Ru₂Cl₄(PP)₂ or RuCl₂(PP)(PPh₃) (PP = Ph₂P(CH₂)₄-PPh₂ (dppb), Ph₂PCH₂CHOCMe₂OCHCH₂PPh₂ (diop), 2,2′-bis(diphenylphosphino)-1,1′-binaphthyl (binap)) with MeCN or PhCN, the product formed depending strongly on the reaction conditions. At high nitrile concentrations, the principal species present is RuCl(PP)(RCN)₃⁺X^- (X = Cl); the cation can generally be isolated (as the PF₆ salt) by abstraction of the chloride counterion with NH₄PF₆. Use of 2 equiv of NH₄PF₆ generates Ru(PP)(RCN)₄²⁺(PF₆^-)₂ (PP = dppb). In the absence of a halide-abstracting agent, addition of hexanes or diethyl ether precipitates neutral RuCl₂(PP)(RCN)₂ species, which undergo further loss of nitrile in the solid state (R = Me) or in solution (R = Me, Ph). Redissolution of the neutral species in chlorocarbon solvents gives Ru₂Cl₃(PP)₂(RCN)₂⁺X ^- (X = Cl) and, in benzene, Ru₂Cl₄(PP)₂(RCN). The dinuclear cation (X = PF₆) is also accessible by reaction of RuCl(PP)(RCN)₃⁺PF₆^- with CH₂Cl₂ or CDCl₃, while the mononitrile can be obtained directly by reaction of Ru₂Cl₄(PP)₂ or RuCl₂(PP)(PPh₃) with small amounts of nitrile in benzene.

Full text of DOI:10.1021/ic960735y

Inorg. Chem. 1997, 36, 1961-1966
1961
Chiral and Achiral Diphosphine Complexes of Ruthenium(II) Incorporating Labile Nitrile
Ligands: Synthesis and Solution Chemistry of Mono- and Dinuclear Derivatives of
Ru2Cl4(PP)2 (PP ) Chelating Diphosphine)
Deryn E. Fogg and Brian R. James*
Department of Chemistry, The University of British Columbia, Vancouver, BC V6T 1Z1, Canada
ReceiVed June 19, 1996X
A family of nitrile complexes has been prepared by reaction of Ru2Cl4(PP)2 or RuCl2(PP)(PPh3) (PP ) Ph2P(CH2)4-
PPh2 (dppb), Ph2PCH2CHOCMe2OCHCH2PPh2 (diop), 2,2′-bis(diphenylphosphino)-1,1′-binaphthyl (binap)) with
MeCN or PhCN, the product formed depending strongly on the reaction conditions. At high nitrile concentrations,
the principal species present is RuCl(PP)(RCN)3+X- (X ) Cl); the cation can generally be isolated (as the PF6
salt) by abstraction of the chloride counterion with NH4PF6. Use of 2 equiv of NH4PF6 generates
Ru(PP)(RCN)42+(PF6-)2 (PP ) dppb). In the absence of a halide-abstracting agent, addition of hexanes or diethyl
ether precipitates neutral RuCl2(PP)(RCN)2 species, which undergo further loss of nitrile in the solid state (R )
Me) or in solution (R ) Me, Ph). Redissolution of the neutral species in chlorocarbon solvents gives
Ru2Cl3(PP)2(RCN)2+X- (X ) Cl) and, in benzene, Ru2Cl4(PP)2(RCN). The dinuclear cation (X ) PF6) is also
-
accessible by reaction of RuCl(PP)(RCN)3+PF6 with CH2Cl2 or CDCl3, while the mononitrile can be obtained
directly by reaction of Ru2Cl4(PP)2 or RuCl2(PP)(PPh3) with small amounts of nitrile in benzene.
Introduction
in this area were directed principally at complexes of mono-
dentate phosphines,10 we were interested, for the purposes of
asymmetric catalysis, in extending this chemistry to complexes
of chelating chiral diphosphines. The susceptibility of the nitrile
ligands themselves in such systems to H2-hydrogenation has
been examined in related work from this laboratory,13,14 and a
detailed assessment of the comparative catalytic activity toward
imine hydrogenation of the nitrile complexes described herein
has been reported.15 The present work involved initially the
optimized syntheses and solution behavior using the achiral
diphosphine dppb (Ph2P(CH2)4PPh2) as a model for the later
used chiral diphosphines diop and binap, which likewise form
seven-membered chelate rings. Several of the dppb complexes
of MeCN have been briefly described.13,16,17 Routes to some
of the mononuclear binap complexes, via thermolysis of
RuCl(arene)(binap)+X- or Ru2Cl4(binap)2(NEt3) in the presence
of nitriles, were reported by other groups during the course of
this present work;11,12 one of these papers11 noted the variation
of the catalytic activity and stereoselectivity of a Ru(binap)
system with solvent composition, especially if MeCN was used
as cosolvent. Together with improved routes to a range of
nitrile-containing complexes with Ru(PP) moieties, also dis-
cussed here is their surprisingly intricate solution behavior
Utilization of Ru complexes in homogeneous catalysis appears
to grow exponentially. The use of Ru catalysts containing
chelating bis(tertiary phosphine) ligands (PP) for asymmetric
hydrogenation was established about 20 years ago,1,2 and there
have been recent spectacular advances in the area for a wide
range of unsaturated organics.3 More generally, besides being
active for the “more classic” catalytic organic reactions (such
as hydrogenation, isomerization, hydrosilylation, decarbonyla-
tion, dehydration, H/D exchange, homologation of amines, etc.),4
Ru phosphine complexes also find increasing utilization in a
much wider range of catalytic reactions including, for example,
C-H/olefin coupling,5 oxidations,6 aldol and Michael additions
via C-H activation of nitriles,7 and addition of carboxylic acids
to alkynes.8
The use of nitriles as ancillary ligands in potential catalysts
is attractive because of their general lability and ease of
replacement,9 for example by an organic moiety, and thus we
and others10-12 chose to study the interaction of dichlororuthe-
nium(II) phosphine compexes with nitriles. While early studies
X Abstract published in AdVance ACS Abstracts, April 1, 1997.
(1) James, B. R.; McMillan, R. S.; Morris, R. H.; Wang D. K. W. AdV.
Chem. Ser. 1978, No. 167, 122.
(2) Joshi, A. M.; Thorburn, I. S.; Rettig, S. J.; James, B. R. Inorg. Chim.
Acta 1992, 198-200, 283.
(3) (a) Noyori, R. Acta Chem. Scand. 1996, 50, 380. (b) Uematsu, N.;
Fujii, A.; Hashiguchi, S.; Ikariya, T.; Noyori, R. J. Am. Chem. Soc.
1996, 118, 4916. (c) Fujii, A.; Hashigushi, S.; Uematsu, N.; Ikariya,
T.; Noyori, R. J. Am. Chem. Soc. 1996, 118, 2521.
(10) (a) Ashworth, T. V.; Singleton, E. J. Organomet. Chem. 1974, 77,
C31. (b) Gilbert, J. D.; Rose, D.; Wilkinson, G. J. Chem. Soc. A
1970, 2765. (c) Gilbert, J. D.; Wilkinson, G. J. Chem. Soc. A 1969,
1749.
(11) Mashima, K.; Hino, T.; Takaya, H. J. Chem. Soc., Dalton Trans. 1992,
2099.
(4) (a) Jardine, F. H. Prog. Inorg. Chem. 1984, 31, 265. (b) Hoffman, P.
R.; Caulton, K. G. J. Am. Chem. Soc. 1975, 97, 4221.
(5) Murai, S.; Kakiuchi, F.; Sekina, S.; Tanaka, Y.; Kamatau, A.; Sonada,
M.; Chatani, N. Pure Appl. Chem. 1994, 66, 1527.
(6) Murahashi, S.-I.; Naota, T.; Hirai, N. J. Org. Chem. 1993, 58, 7318.
(7) Murahashi, S.-I.; Naota, T.; Taki, H.; Mizuno, M.; Takaya, H.; Komiya,
S.; Mizuho, Y.; Oyasato, N.; Hiraoka, M.; Hirano, M.; Fukuoka, A.
J. Am. Chem. Soc. 1995, 117, 12436.
(12) Shao, L.; Takeuchi, K.; Ikemoto, M.; Kawai, T.; Ogasawara, M.;
Takeuchi, H.; Kawano, H.; Saburi, M. J. Organomet. Chem. 1992,
435, 133.
(13) Thorburn, I. S.; Rettig, S. J.; James, B. R. J. Organomet. Chem. 1985,
296, 103.
(14) Joshi, A. M.; MacFarlane, K. S.; James, B. R.; Frediani, P. Chem.
Ind. 1994, 53, 497.
(15) (a) Fogg, D. E.; James, B. R.; Kilner, M. Inorg. Chim. Acta 1994,
222, 85. (b) Fogg, D. E.; James, B. R. Inorg. Chem. 1995, 34,
2557.
(8) Doucet, H.; Martin-Vaca, B.; Bruneau, C.; Dixneuf, P. J. Org. Chem.
1995, 60, 7247.
(9) Endres, H. In ComprehensiVe Coordination Chemistry; Wilkinson, G.,
Gillard, R. D., McCleverty, J. A., Eds.; Pergamon Press: Oxford, U.K.,
1987; Vol. 2, p 261.
(16) James, B. R.; Pacheco, A.; Rettig, S. J.; Thorburn, I. S.; Ball, R. G.;
Ibers, J. A. J. Mol. Catal. 1987, 41, 147.
(17) Fogg, D. E.; James, B. R. Can. J. Chem. 1995, 73, 1084.
S0020-1669(96)00735-5 CCC: $14.00 © 1997 American Chemical Society

1962 Inorganic Chemistry, Vol. 36, No. 9, 1997
Fogg and James
vacuum. Yield: 29 mg (71%). 31P{1H} NMR (C6D6): δ 54.7, 54.4
(ABq, J ) 44.2); 52.4, 45.7 (ABq, J ) 35.8). IR: υ(CtN) 2230 (w).
PP ) diop (3b). Precursor 2b (108 mg, 0.115 mmol) was stirred
in C6H6 (5 mL) with a small excess of PhCN (8 µL, 1.4 equiv per Ru2)
for 1 h. The solution was then concentrated to 0.25 mL and Et2O added
to precipitate the product, which was reprecipitated from C6H6-hexanes.
Yield: 71 mg (85%). 31P{1H} NMR (C6D6): δ 47.1, 44.1 (ABq, J )
42.1); 44.7, 35.8 (ABq, J ) 34.7); 44.6, 39.0 (ABq, J ) 35.6); ∼44-
45 (concealed; inferred from integration). IR: υ(CtN) 2232 (w).
PP ) binap (3c). The complex was prepared from 5c in 4 h, in the
manner described for 3a, but with a further reprecipitation from C6H6-
hexanes (83% yield). 31P{1H} NMR (C6D6): δ 62.5, 56.3 (ABq, J )
42.0); 62.3, 58.4 (ABq, J ) 41.0); 57.0, 51.2 (ABq, J ) 33.5); 55.3,
53.8 (ABq, J ) 35.1). IR: υ(CtN) 2229 (w).
Scheme 1. Summary of Relationships between Nitrile
Complexes (Stereochemistry Omitted; See eq 1 and Figures
1 and 2)a
Ru2Cl4(dppb)2(MeCN) (4a). The complex was prepared from 2a
in 92% yield in the manner described for 3b but using 1:2 MeCN-
C6H6. Alternatively, a suspension of 1a (400 mg, 0.33 mmol) in MeCN
(10 mL) was stirred for 2 h, and the solvent then removed under
vacuum. The residue was dissolved in hot C6H6, and Et2O was added
to precipitate an orange solid, which was filtered off, washed with Et2O
(3 × 5 mL), and dried under vacuum. Yield: 0.29 g (71%). 31P{1H}
NMR (C6D6): δ 53.5, 52.2 (ABq, J ) 44.3); 51.0, 45.8 (ABq, J )
36.2); (CD2Cl2) δ 52.3 (s); 50.8, 46.5 (ABq, J ) 36.4). IR: υ(CtN)
2282 (w).
RuCl2(PP)(PhCN)2. PP ) dppb (5a). Precursor 2a (233 mg, 0.270
mmol) was stirred in a mixture of PhCN (0.5 mL) and C6H6 (8 mL)
for 12 h. The yellow product was filtered off, washed with hexanes
(3 × 10 mL), and dried under vacuum. Yield: 182 mg (83%). 31P-
{1H} NMR (C6D6): δ 50.5 (s). IR: υ(CtN) 2241 (w).
PP ) diop (5b). The complex was prepared from 2b in a manner
similar to that given for 5a but without the benzene cosolvent; the
mixture was stirred for 15 min before precipitating the product with
hexanes (77% yield). 31P{1H} NMR (C6D6): δ 49.7, 36.0 (dd, J )
38.2). IR: υ(CtN) 2237 (w).
PP ) binap (5c). The complex was prepared from 2c in the same
manner as 5a (91% yield). 31P{1H} NMR (C6D6): δ 50.3 (s, immediate,
t-5c); 55.3, 53.6 (ABq, J ) 35.0, grows in rapidly, c,c,c-5c); (10:1
PhCN-C6D6) δ 50.1 (s, t-5c); 47.9, 44.3 (ABq, J ) 35.0; c,c,c-5c).
IR: υ(CtN) 2232 (w). Slightly different 31P{1H} NMR parameters
in neat PhCN were described for t- and c,c,c-5c in a report which
appeared during the course of this present work12 (see text).
RuCl2(PP)(MeCN)2. PP ) dppb (6a). A solution of 1a (74 mg,
0.060 mmol) in MeCN (10 mL) was stirred for 15 h and then
concentrated to ∼4 mL and treated with Et2O to precipitate the yellow
product (48 mg, 69% yield). Samples of the isolated product turned
orange on drying under air or vacuum overnight and were identified
by 31P{1H} NMR (C6D6) as a mixture of 6a and 4a. Loss of nitrile
from isolated 6a also occurred rapidly in benzene solution. Attempts
to prepare 6a in the same manner as its PhCN analog 5a were
unsuccessful. 31P{1H} NMR (C6D6): δ 50.3 (s).
a Conditions: (a) benzene, RCN:Ru2 ) 1.0; (b) RCN; (c) RCN,
NH4PF6:Ru ) 1.0; (d) RCN; (e) vacuum (R ) Me) or benzene; (f)
CH2Cl2 or CDCl3. PP: In the text, a, b, and c refer to dppb, diop, and
binap species, respectively; in one case (10d), d refers to a bis PPh3
complex.
(Scheme 1), which can be exploited to gain access to any
member of this network from virtually any other.
Experimental Section
Synthetic and spectroscopic work was performed under Ar at room
temperature (room temperature, ∼20 °C). Solvents were distilled from
CaH2 (CH2Cl2) or sodium-benzophenone (C6H6, ether, hexanes).
Acetonitrile and benzonitrile were distilled prior to use and stored under
Ar in the dark. NMR spectra were recorded on a Varian XL300
spectrometer, using the residual proton of the solvent (1H) or free PPh3
(31P: C6D6, -5.06; CDCl3, -5.46; CD2Cl2, -5.64; CD3CN, -6.49 ppm
with respect to 85% external H3PO4) as internal standards [s ) singlet,
dd ) doublet of doublets, q ) quartet, sept ) septet; all J values are
in Hz]. Infrared spectra (cm-1) were measured as Nujol mulls between
KBr plates on a Bomem Michelson 100 FT-IR spectrophotometer (w
) weak). Elemental analyses were performed by Mr. P. Borda of the
UBC Microanalytical Service; all new complexes gave satisfactory
analyses (Table S1, Supporting Information) unless otherwise indicated.
The Ru precursors Ru2Cl4(PP)2 (1a-c),2,16 RuCl2(PPh3)3,4 and RuCl2-
(PP)(PPh3) (2a-c)6,18,19 (PP: a, dppb; b, diop; c, binap) were prepared
as previously described. We reported recently the syntheses of the
Ru(dppb)(RCN)42+(PF6-)2 species (R ) Ph, 11a, or Me, 12a), including
the X-ray structure of the MeCN complex.17
PP ) binap (6c). The complex was observed in situ. A sharp 31P-
{1H} AB pattern assigned to an all-cis isomer of 6c, as well as the
broad resonances due to fac- and mer-RuCl(binap)(CD3CN)3+Cl- (cf.
10c-Cl), was evident in spectra of precursor 2c dissolved in CD3CN.
A sample of 10c-PF6 in CDCl3 showed immediately on dissolution only
a singlet at δ 51.9 for a trans-nitrile isomer of 6c (assigned by analogy
-
to data for 5c). Reprecipitation of RuCl(binap)(MeCN)3+PF6 (10c-
PF6) from CH2Cl2-C6H6 gave a yellow product which on dissolving
in C6D6 showed an AB pattern tentatively assigned to another all-cis
isomer of 6c; no PF6 septet was evident. 31P{1H} NMR (CD3CN): δ
46.9, 41.3 (ABq, J ) 38.0; c,c,c-6c); (CDCl3) δ 51.9 (s, t-6c; this gives
way over 24 h to the AB pattern of 8c-Cl and a singlet at δ 45.5 of
uncertain identity); (C6D6) δ 47.7, 45.1 (ABq, J ) 31.9; c,c,c-6c). These
NMR data contrast with values reported for 6c in a paper which
appeared during the course of this work11 (for details see text).
Ru2Cl3(PP)2(PhCN)2+X-. PP ) dppb, X ) Cl (7a-Cl). A solution
of 5a (44 mg, 0.046 mmol) in CH2Cl2 (0.5 mL) was stirred for 1 h and
then treated with Et2O (∼2 mL) to precipitate a bright yellow powder,
which was filtered off and dried overnight under vacuum. Yield: 26
mg (68%). The NMR data agree with those for the PF6 salt.
The 31P{1H} NMR data are listed with the individual, synthesized
1
complexes (see below); H NMR data for 3a-c, 4a, 5a-c, and 6a,
and the PF6- salts of 7a, 8a, 9a,b, and 10a,d are given in the Supporting
Information, Table S2.
Ru2Cl4(PP)2(PhCN). PP ) dppb (3a). A suspension of 5a (51
mg, 0.063 mmol) in C6H6 (10 mL) was stirred for 24 h. The solution
was concentrated, and Et2O was added to obtain an orange precipitate,
which was filtered off, washed with Et2O (3 × 5 mL), and dried under
X ) PF6 (7a-PF6). A solution of 5a (88 mg, 0.109 mmol) in CH2-
Cl2 (5 mL) was treated with a solution of NH4PF6 (8.9 mg, 0.054 mmol)

Diphosphine Complexes of Ru(II)
Inorganic Chemistry, Vol. 36, No. 9, 1997 1963
in acetone (2 × 1 mL). The mixture was stirred for 1 h and then filtered
through Celite. The filtrate was concentrated to ∼1 mL and Et2O added
to precipitate the yellow product, which was filtered off and dried
overnight under vacuum. Yield: 69 mg (83%). 31P{1H} NMR (CD2-
Cl2): δ 49.3, 46.0 (ABq, J ) 36.2), -144.5 (sept, J ) 713, PF6). IR:
υ(CtN) 2234 (w).
PP ) diop, X ) Cl (7b-Cl). The complex was prepared in situ
(from 5b) as described for 7a-Cl above. 31P{1H} NMR spectra obtained
immediately showed only signals for 7b-Cl; the phosphine shift
positions were identical to those of 7b-PF6.
X ) PF6 (7b-PF6). The complex was prepared in situ by reaction
of 9b-PF6 with CDCl3, stirring the solution for 10 h before measuring
the NMR spectrum. 31P{1H} NMR (CDCl3): δ 43.5, 36.4 (ABq, J )
35.6), -144.3 (sept, J ) 713, PF6).
PhCN in the solid state. 31P{1H} NMR (10:1 PhCN-CDCl3): δ 46.0,
44.4 (ABq, J ) 29.8); -144.4 (sept, J ) 712, PF6). IR: υ(CtN)
2234 (w).
RuCl(PP)(MeCN)3+X-. PP ) dppb, X ) Cl (10a-Cl). This
species was stable only in the presence of MeCN; after its formation
in situ by dissolving 2a in CD3CN, the NMR spectra for the cation
were identical to those of the PF6 salt.
X ) PF6 (10a-PF6). The complex was prepared in 76% yield from
precursor 2a in MeCN, in a manner otherwise similar to that described
for the PhCN analog. Alternatively, a solution of 1a (154 mg, 0.129
mmol) in MeCN (5 mL) was treated with NH4PF6 (41.9 mg, 0.257
mmol) in acetone (2 × 1 mL). The solution was stirred for 5 h and
then filtered through Celite. The filtrate was concentrated to ∼1 mL
and diluted slightly with CH2Cl2 (1 mL), and the product was
precipitated by addition of Et2O. Reprecipitation from a 1:1 mixture
of MeCN and CH2Cl2 gave a fine yellow powder; in the absence of
MeCN, clean products (free of 8a-PF6) could not be obtained. The
precipitate was washed with Et2O and hexanes and dried under vacuum.
Yield: 158 mg (74%). 31P{1H} NMR (CDCl3): δ 41.7 (s), -144.2
(sept, J ) 713, PF6); (CD3CN) δ 40.6 (s, fac); 42.5, 35.7 (ABq, J )
34.4, mer), -145.3 (sept, J ) 706, PF6). IR: υ(CtN) 2268 (w).
PP ) diop, X ) Cl (10b-Cl). The complex was prepared in situ
by dissolving 1b in CD3CN. The 31P{1H} NMR parameters agree with
those for the PF6 salt in CD3CN.
X ) PF6 (10b-PF6). Attempts to prepare this complex from 2b in
the manner described for the dppb analog gave a yellow precipitate
(73% yield). Although this material showed the expected 31P{1H} NMR
pattern, loss of MeCN occurred on exposure to vacuum: a color change
to white and then pink was observed on drying overnight, and a complex
series of NMR peaks was apparent on dissolving the pink solid in C6D6.
Addition of MeCN (∼0.2 mL) to the NMR sample regenerated the
original spectrum. 31P{1H} NMR (CD3CN): δ 37.1, 30.5 (ABq, J )
40.2); -145.1 (sept, J ) 707, PF6).
PP ) binap, X ) PF6 (7c-PF6). The complex was prepared in
68% yield (from 5c) as described for 7a-PF6. 31P{1H} NMR (CDCl3):
δ 54.1, 52.0 (ABq, J ) 35.0), -144.2 (sept, J ) 713, PF6).
Ru2Cl3(PP)2(MeCN)2+X-. PP ) dppb, X ) Cl (8a-Cl).
A
solution of 1a (74 mg, 0.062 mmol) in MeCN (10 mL) was stirred for
10 h and then concentrated to ∼2 mL and treated with Et2O to
precipitate a yellow product, which was washed well with Et2O (4 ×
5 mL) to remove MeCN. Reprecipitation from CH2Cl2-Et2O gave
yellow 8a-Cl; yield 40 mg (51%). 1H NMR and 31P{1H} NMR spectra
agree with those for the PF6 salt.
X ) PF6 (8a-PF6). The complex was not prepared by the method
we originally reported13 but by in situ reaction of 10a-PF6 with CH2-
Cl2 in 90% yield (as described for 7b-PF6 from 9b-PF6). Alternatively,
a solution of 2a (151 mg, 0.175 mmol) in MeCN (5 mL) was treated
with NH4PF6 (14.3 mg, 0.088 mmol) in MeCN (2 × 1 mL). The
mixture was stirred for 2 h and then stripped of solvent. The residue
was redissolved in CH2Cl2 (5 mL), filtered through Celite, and the
filtrate concentrated to ∼1 mL. The yellow product was precipitated
with C6H6 and dried under vacuum; yield 107 mg (88%). 31P{1H}
NMR (CD2Cl2): δ 49.4, 46.6 (ABq, J ) 37.2), -144.5 (sept, J ) 712,
PF6). IR: υ(CtN) 2275 (w).
PP ) binap, X ) Cl (10c-Cl). The complex was prepared in situ
by dissolving 1c in CD3CN. The 31P{1H} NMR parameters agree with
those for the PF6 salt.
PP ) diop, X ) PF6 (8b-PF6). The complex was prepared in situ
X ) PF6 (10c-PF6). A solution of 2c (108 mg, 0.102 mmol) in
MeCN (5 mL) was treated with NH4PF6 (16.9 mg, 0.104 mmol) and
stirred for 48 h. The solution was filtered through Celite, the filtrate
stripped to a yellow oil, and C6H6 added to the residue, giving a yellow
precipitate which was filtered off and washed with benzene (4 × 1
mL). Yield after drying under vacuum: 74 mg (70%). Microanalysis
indicated a low nitrogen content and deteriorated further when a sample
was reprecipitated from MeCN-C6H6. 31P{1H} NMR (CD3CN): δ
47.5, 45.5 (br ABq, J unobservable); 47.5, 44.6 (ABq, J ) 40.2);
-145.1 (sept, J ) 706, PF6). IR: υ(CtN) 2280 (w).
PP ) 2PPh3, X ) PF6 (10d-PF6). A solution of RuCl2(PPh3)3 (252
mg, 0.263 mmol) in MeCN (5 mL) was treated with NH4PF6 (42.9
mg, 0.263 mmol) and stirred for 48 h. The mixture was filtered through
Celite, the filtrate concentrated to ∼1 mL, and a pale yellow powder
precipitated with Et2O. The product was filtered off, washed with C6H6
(3 × 5 mL), and dried under vacuum. Yield: 175 mg (72%). 31P{1H}
NMR (CD3CN): δ 43.8 (m), 40.4 (m), -145.3 (sept, J ) 706, PF6).
IR: υ(CtN) 2279 (w).
-
from crude (undried) RuCl(diop)(MeCN)3+PF6 (10b-PF6) in CDCl3.
31P{1H} NMR (CDCl3): δ 43.8, 37.4 (ABq, J ) 36.0), -144.5 (sept,
J ) 713, PF6).
PP ) binap, X ) Cl (8c-Cl). The complex was observed by
-
31P{1H} NMR in spectra of crude RuCl(binap)(MeCN)3+PF6 (10c-
PF6) in CDCl3. 31P{1H} NMR (CDCl3): δ 54.3, 53.0 (ABq, J ) 36.1).
RuCl(PP)(PhCN)3+PF6-. PP ) dppb (9a-PF6). To a suspension
of 2a (121 mg, 0.141 mmol) in PhCN (2 mL) under Ar was added
CH2Cl2 (5 mL), giving a yellow solution which was immediately
subjected to vacuum for 10 min to remove CH2Cl2. The mixture was
stirred for 2 h and then treated with a solution of NH4PF6 (22.9 mg,
0.141 mmol) in acetone (2 × 1 mL). The mixture was stirred for 8 h
and then filtered through Celite; the filtrate was concentrated to a yellow
oil, diluted with C6H6 (1 mL), and treated with Et2O to precipitate the
product. Yield: 102 mg (71%) after washing with Et2O and C6H6 (3
× 5 mL each) and drying under vacuum. 31P{1H} NMR (CD3CN): δ
40.3 (s, fac); 41.7, 34.4 (ABq, J ) 33.5, mer), -145.3 (sept, J ) 706,
PF6). IR: υ(CtN) 2236 (w).
Results and Discussion
PP ) diop (9b-PF6). To a solution of 2b (78.5 mg, 0.0842 mmol)
in PhCN (0.25 mL) was added a solution of NH4PF6 (13.7 mg, 0.0840
mmol) in acetone (2 × 1 mL). The mixture was immediately diluted
with CH2Cl2 (2 mL) and filtered through Celite, and the filtrate was
concentrated to ∼0.25 mL. The pale yellow product precipitated by
addition of Et2O was filtered off, washed with hexanes (3 × 5 mL),
and dried under vacuum. Yield: 76.3 mg (83%). 31P{1H} NMR
(CDCl3): δ 37.2, 28.2 (ABq, J ) 39.3), -144.2 (sept, J ) 713, PF6).
IR: υ(CtN) 2265, 2247 (w).
The doubly chloride-bridged dimer [RuCl2(dppb)]2 (1a) reacts
with a wide range of two-electron donor ligands to give
monosubstituted products of the type RuCl(dppb)(µ-Cl)3Ru-
(dppb)(L), eq 1; such species are readily detected by the
PP ) binap (9c-PF6). The complex was prepared in a manner
similar to that described for the diop analog but stirring the precursor
2c in PhCN for 24 h before adding NH4PF6. The mixture was then
stirred for 2 h, diluted with C6H6 (5 mL), and filtered through Celite.
The filtrate was concentrated and treated with Et2O to precipitate the
yellow product, which was filtered off and washed with warm hexanes
(4 × 2 mL). Yield: 47 mg (61%). The microanalytical data are in
poor agreement with the proposed structure, probably owing to loss of
appearance of two AB quartet patterns in the 31P{1H} NMR
spectra.2,15 The corresponding L ) nitrile derivatives 3 and 4
can be isolated (see below), but nitrile ligands can also
easily cleave the chloride bridges to generate mononuclear
RuCl(PP)(RCN)3+Cl- (R ) Ph, Me), probably via the neutral

1964 Inorganic Chemistry, Vol. 36, No. 9, 1997
Fogg and James
Table 1. 31P{1H} NMR Data Outlining the Solution Behavior of
RuCl2(binap)(PhCN)2 (5c)
species RuCl2(PP)(RCN)2. Scheme 1 summarizes this chemistry
and the related conversions to be discussed here. The five-
coordinate complexes RuCl2(PP)(PPh3) (2; PP ) dppb, diop,
binap), which in CDCl3 or C6H6 provide an in situ source of
Ru2Cl4(PP)2 (1),2,18 also act as direct precursors to the cationic
species by displacement of PPh3 and chloride. The mildness
of these reaction conditions (room temperature, nitrile solvent),
possible because of the high substitutional lability of the Ru
precursors, is notable in view of the high temperatures typically
employed in synthetic and catalytic chemistry of Ru-phosphine
species.10-12 Use of the mixed-phosphine complexes 2a-c,
accessible in two steps and quantitative yield from commercially
available RuCl3,2,18 provides a particularly attractive route to
nitrile derivatives.
a
solvent
precursor
chem shift (JAB
)
assgnt
C6D6
isolated 5c
50.3
55.3, 53.6 (35)
50.3
48.4, 44.2 (35)
50.1
47.9, 44.3 (35)
54.1, 52.0 (35)
50.1
t-5c
c,c,c-5c
t-5c
c,c,c-5c
t-5c
c,c,c-5c
7c-Cl
t-5c
1:7 PhCN-C6D6
10:1 PhCN-C6D6
10:1 PhCN-CDCl3
3c
isolated 5c
2c
47.1, 43.9 (35)
46.0, 44.4 (30)
c,c,c-5c
9c-Cl
a J values in Hz.
The simultaneous presence in RuCl(PP)(RCN)3+Cl- of a
labile nitrile ligand and the potentially coordinating Cl- anion
results in a product of low stability, and isolation of the cationic
product requires removal of the counterion with, for example,
NH4+. The stoichiometry of this reaction is critical; any excess
of NH4PF6 leads to dicationic Ru(PP)(RCN)42+(PF6-)2, which
has been crystallographically characterized (PP ) dppb, R )
Me).17 Synthesis of the corresponding binap complex (11c) has
also been reported.11 Interestingly, the monocationic complex
RuCl(PP)(PhCN)3+PF6- (9c-PF6) could not be isolated in a pure
state, as earlier found for the MeCN analog.11 Nitrile loss from
such binap species may be facilitated by steric pressures exerted
by the bulky and rigid diphosphine ligand; a high degree of
lability or fluxionality was found in the present work to be
characteristic of the binap complexes.
The dppb species 9a and 10a exist as the fac isomer in C6D6
or chlorocarbon solvents, as judged by the presence of a 31P-
{1H} NMR singlet (though reaction to give other products via
loss of nitrile occurs, as discussed below). An accompanying
AB quartet of variable intensity in neat or C6D6-spiked nitrile
is attributed to isomerization to the mer species; the intensity
of this signal increases over time. Measurements were made
in CD3CN for the PhCN species 9a as well as 10a; displacement
of PhCN by CD3CN did not interfere if measurements were
made soon after dissolution because the nitrile exchange process
is rather slow, being ca. 50% complete after 24 h.
For the diop and binap complexes, disruption of molecular
symmetry by the chirality of the phosphine ligands should give
rise to an AB pattern for each isomer; assignment as fac or mer
is therefore made only by extrapolation from the dppb chemistry.
Interestingly, neither 9b nor 9c exhibits more than one AB
quartet in PhCN, probably owing to increased steric pressure
at the axial positions exerted by the greater rigidity and/or bulk
of the diphosphine ligand, relative to dppb. This tends to
support identification of the species present as the fac isomer,
for which interaction of the phosphine phenyl rings with the
comparatively bulky nitrile ligand is minimized. Both isomers
of the binap complex 10c are formed in CD3CN, though further
assignment cannot be made with certainty. Two AB quartets
are observed, the downfield halves of which (centered at δ 47.5)
overlap; the upfield peaks, though broad, are distinguishable
(δ 45.5, 44.6). Also initially present is an AB quartet due to
all-cis-6c (see discussion below and Table 2). Only one AB
pattern was reported for 10c in a study of the binap-MeCN
chemistry which appeared during the course of this work (1:1
MeCN-CDCl3, -30 °C: δ 47.7, 44.9, J ) 32).11 The absence
of signals for the second isomer may be due to any of a number
of factors: Duration of time in solution is probably a key factor,
but temperature or (given the usual absence of isomerization in
C6D6 or halogenated solvents) the concentration of MeCN may
also be significant. Rather surprisingly, only one isomer was
observed for the diop analog 10b in the present work, even on
prolonged exposure to CD3CN at room temperature. As this
species proved unexpectedly susceptible to decomposition with
loss of nitrile, it was not investigated in detail.
Attack of the chloride counterion on the Ru center, with
displacement of a nitrile ligand, occurs during attempts to isolate
solid RuCl(PP)(RCN)3+Cl-, and the neutral RuCl2(PP)(RCN)2
complexes (5, R ) Ph; 6, R ) Me) precipitate instead. In the
case of the MeCN species 6a, the neutral complex in turn
undergoes loss of MeCN on drying under vacuum and gives
orange Ru2Cl4(dppb)2(MeCN) (4a). The benzonitrile complexes
5a-c are less susceptible to nitrile loss under vacuum, owing
to the lower volatility of the aromatic nitrile, and the PhCN
derivatives were therefore investigated in greater detail. In the
solid state, 5a-c probably exist as the trans-nitrile isomer, as
judged from the single ν(CtN) in the infrared. The dppb
species 5a retains this geometry in solution (though loss of nitrile
is promoted in the absence of RCN, as discussed below); the
diop and binap complexes, in contrast, appear to undergo facile
isomerization. The 31P{1H} NMR spectrum of the diop
derivative 5b, obtained within 30 min of dissolution in C6D6,
consists solely of a pair of doublets assigned to an all-cis isomer;
the expected singlet for the trans isomer is not observed. For
the binap species, a singlet at δ 50.3 observed immediately on
dissolution of isolated, analytically pure 5c in C6D6 is assigned
to trans-5c. The solution behavior of 5c is complex (Table 1):
While an AB quartet (δ 55.3, 53.6; J ) 35) grows in rapidly in
neat C6D6, a different AB pattern (δ 48.4, 44.2; J ) 35) is found
in 1:7 PhCN-C6D6. The latter signal is shifted slightly upfield
as the proportion of PhCN is increased and is then accompanied
by that due to cationic 9c-Cl. The location of the upfield AB
quartet is in reasonable agreement with values found in neat
PhCN by other workers (δ 47.0, 43.3; J ) 35) and assigned to
all-cis 5c; this group recorded also a singlet at δ 48.6 for trans-
5c.12
Two distinct cis isomers of 5c can in fact be formed, owing
to the chirality of the binap ligand; while the chemical shift
difference between the two AB sets of signals described above
seems large for this apparently subtle difference in structural
form, precedent exists for such a distinction.20 More perplexing
is the implied diastereoselectivity of isomerization in a particular
solvent mixture: In no case are both AB patterns observed
simultaneously. An alternative possibility for assignment of the
downfield AB pattern is the doubly chloride-bridged structure
[RuCl(PP)(PhCN)]2(µ-Cl)2. No evidence for such a species is
(19) Wang, D. K. W. Ph.D. Thesis, The University of British Columbia,
1978.
(18) Jung, C. W.; Garrou, P. E.; Hoffman, P. R.; Caulton, K. G. Inorg.
Chem. 1984, 23, 726.
(20) Morandini, F.; Consiglio, G.; Ciani, G.; Sironi, A. Inorg. Chim. Acta
1984, 82, L27.

Diphosphine Complexes of Ru(II)
Inorganic Chemistry, Vol. 36, No. 9, 1997 1965
Table 2. 31P{1H} NMR Data Outlining the Solution Behavior of
RuCl2(binap)(MeCN)2 (6c)
a
solvent precursor
chem shift (JAB
47.7, 45.1 (32)
10c-PF6 51.9
54.3, 53.0 (36)b
45.5b
)
assgnt
c,c,c-6c
t-6c
8c-Cl
uncertain (see text)
C6D6
6c
CDCl3
CD3CN
2c
47.5, 45.5 (J unobservable), 10c-Cl
47.5, 44.6 (40)
46.9, 41.3 (38)
c,c,c-6c
a J in Hz. b Not immediately observed.
observed in the chemistry of dppb or diop, though perhaps its
stability relative to the usual, highly stable Ru2(µ-Cl)3 entity
may be enhanced by the bulk and rigidity of the binap ligand.
It should be noted that no coupling through the bridging
chlorines has been observed in the Ru2(µ-Cl)3 complexes
described below, although through-bridge P-coupling has been
reported for complexes such as (PPh3)2(H)Ru(µ-H)(µ-Cl)2Ru-
(η2-H2)(PPh3)2.21,22
Figure 1. Diastereomers of Ru2Cl4(PP)2(RCN), illustrated for PP )
In the corresponding binap-MeCN chemistry, no downfield
AB pattern corresponding to that observed for 5c in neat C6D6
was found. A single AB quartet appears (C6D6: δ 47.7, 45.1;
J ) 32) near the location described above for 5c in PhCN-
C6D6 and is assigned to all-cis-6c. A similar resonance (δ 46.9,
41.3; J ) 38) present immediately following dissolution of 2c
in CD3CN (Table 2) is completely replaced within a few hours
by signals due to cationic 10c-Cl. In CDCl3, a singlet at δ 51.9,
assigned to a trans isomer of 6c, was the sole phosphine signal
initially present; no AB pattern was observed. Conflicting data
appear in the literature11 for an isolated RuCl2(binap)(MeCN)2
complex (CDCl3): δ 54.8, 53.7; J ) 35.2; δ 51.7, 50.9; J )
35.2; ratio of AB quartets ) 9:1. The singlet for a trans isomer
was not observed in the literature work, possibly because of
delays between dissolving the sample and measuring the
spectrum; our investigations clearly show that the product
distribution is sensitive to both solvent and the time in solution.
Of the two AB quartets reported, however, neither corresponds
to that found in C6D6. The lower-field pattern is almost certainly
due to Ru2Cl3(binap)2(MeCN)2+Cl- (8c-Cl), which is formed
slowly from 6c in chlorinated solvents (see below). The other
pattern may be due to an all-cis isomer of 6c, as the authors
propose, or to a doubly chloride-bridged dimer, as suggested
above for the PhCN species. While the reported microanalytical
data support the formulation of the solid precursor as
RuCl2(binap)(MeCN)2 (with 0.5 CH2Cl2), the lability of the
nitrile ligands complicates the solution structure.
The remainder of the solution chemistry of the binap
complexes is largely consistent with that of the dppb and diop
species. Loss of nitrile from RuCl2(PP)(RCN)2 (5 or 6) is
promoted in solution, as noted above; in C6H6, the product is
Ru2Cl4(PP)2(RCN) (R ) Ph, 3a-c; R ) Me, 4a), as indicated
by elemental analysis and, in the case of 3b and 4a, by
independent synthesis from 2b and 2a, respectively, via reaction
with nitrile. The rate of conversion of 5 or 6 is variable,
depending probably on the concentration of residual free nitrile
in the sample. In the absence of nitrile, signals for the dinuclear
species are discernible by 31P{1H} NMR immediately following
dissolution in C6D6, and are predominant within 24 h. A pattern
of four AB quartets of equal integrated intensity is observed
for each of the chiral complexes 3b,c, probably reflecting the
low-energy difference between a pair of diastereomers either
binap. The phenyl rings of the binap ligands are omitted for clarity.
-
Figure 2. Neutral and ionic isomers of Ru2Cl4(PP)2(PhCN)2.
formed without discrimination or capable of interconverting in
solution. The NMR spectra of 3c in C7D8 remain essentially
unchanged down to -80 °C, tending to support the former
possibility. Figure 1 shows the possible structures for the
diastereomers, illustrated with the binap species 3c. A structure
of type a/b (the two are enantiomeric if the phosphine is achiral)
is clearly adopted by the corresponding dppb complex, for which
two 31P{1H} AB patterns are observed; isomer c, in which the
nitrile and the terminal chloride ligand are coplanar, would give
rise to two singlets.
A similar process of nitrile loss and dimerization almost
certainly occurs with the PPh3 analog of 6. Reprecipitation of
RuCl2(PPh3)2(MeCN)2 from toluene was in early work reported
to generate the doubly chloride-bridged product [RuCl2(PPh3)2-
(MeCN)]2(µ-Cl)2,23 but the accompanying analytical data sug-
gest that this should be reformulated as RuCl(PPh3)2(µ-
Cl)3Ru(PPh3)2(MeCN). This work was not reinvestigated in
the present study, the PPh3 chemistry being limited to prepara-
tion of the cationic species 10d-PF6.
Nitrile loss also occurs on dissolving mononuclear 5 or 6 in
CH2Cl2 or CDCl3, but the product is now the dinuclear cation
Ru2Cl3(PP)2(RCN)2+Cl- (R ) Ph, 7-Cl; R ) Me, 8-Cl). The
PF6 salts of 7a,c were readily prepared by addition of 1 equiv
of NH4PF6 per Ru2. The counterion has no effect on the
phosphine shift positions, the values for the Cl and PF6 salts
being identical, and the structure of 7a-Cl, earlier represented
as neutral [RuCl(dppb)(PhCN)]2(µ-Cl)2,2,15b is now redefined
on this basis as {[Ru(dppb)(PhCN)]2(µ-Cl)3}+Cl- (Figure 2).
Synthesis (from the mixed-valence species Ru2Cl5(dppb)2) and
conductivity studies of the dppb-MeCN complex 8a-PF6 were
previously reported by our group.13 Reaction of RuCl-
-
(PP)(RCN)3+PF6 (9, 10) with the chlorinated solvents CH2-
Cl2 or CDCl3 provided an unexpected additional route to the
dinuclear, cationic complexes, and quantitative conversion of
(21) Hampton, C.; Dekleva, T. W.; James, B. R.; Cullen, W. R. Inorg.
Chim. Acta 1988, 145, 165.
(22) Dekleva, T. W.; Thorburn, I. S.; James, B. R. Inorg. Chim. Acta 1985,
100, 49.
(23) Cole-Hamilton, D. J.; Wilkinson, G. J. Chem. Soc., Dalton Trans.
1979, 1283.

1966 Inorganic Chemistry, Vol. 36, No. 9, 1997
Fogg and James
the dppb and diop species to the dinuclear PF6 salts was
observed. Several experiments tend to implicate direct reaction
of the mononuclear cations with solvent, rather than with solvent
photochemical decomposition products such as HCl. For
example, efforts to block conversion of 10a-PF6 into 8a-PF6
by neutralization of possible HCl impurity were unsuccessful;
similarly, the extent of reaction to give 8a-PF6 was independent
of the batch of solvent used but varied with the batch of starting
cation (again, probably depending on the amount of residual
free nitrile present in the sample). Reports on the implication
of CH2Cl2 and CHCl3 in the quantitative halogenation of
inorganic species are increasingly common. The nature of the
chlorinating species is frequently uncertain, but accumulating
reports of oxidative-addition products (in which both chloride
and hydrocarbyl ligands are retained)24 show that direct reaction
between late transition-metal complexes and the halocarbon
solvent is not unusual, while crystallographic studies have also
shown CH2Cl2 in both chelating (at an Ag(I) center)25 and
bridging (at an Ru3 site)26 modes.
The behavior of the binap complex 10c-PF6 is again
anomalous within this chemistry. Dissolution of this species
in CDCl3 yields neutral t-RuCl2(binap)(MeCN)2 (6c) as the sole
initial product, as judged by in situ 31P{1H} NMR (δ 51.9, s).
The solution instability of 10c-PF6 was noted by Takaya’s group,
and a process of disproportionation into 6c and dicationic
Ru(binap)(MeCN)42+(PF6-)2 (12c) was proposed.11 A signal
near the location reported for 12c (δ 46.6)11 does in fact develop
in the 31P{1H} NMR spectrum of 10c-PF6 after 24 h in CDCl3:
the principal product after this time is the dinuclear cation 8c,
but accompanying the AB quartet for this species is a new
singlet at δ 45.5, of about a third of the integrated intensity.
The assignment of this peak remains somewhat uncertain despite
the near-correspondence in chemical shift, however, as our
extensive study of the solution chemistry of isolated RuCl2-
(PP)(RCN)2 complexes has shown no precedent for the im-
plied equilibration to 12 (R ) MeCN) or 11 (R ) PhCN)
species.
Addition of nitrile to 7 or 8 species (Scheme 1) cleaves the
triple chloride bridge to give the mononuclear cations RuCl-
(PP)(RCN)3+X-. In the presence of nitrile, reaction of 9-PF6
or 10-PF6 with chlorocarbons is suppressed, and MeCN-CH2-
Cl2 mixtures were in fact successfully used to reprecipitate 10a-
PF6. Owing to the weaker ligating character of PhCN, caution
must be employed in using this means of purification for the
corresponding PhCN complexes; 31P{1H} NMR signals for 7c,
for example, are observed even in 10:1 PhCN-CDCl3 im-
mediately following dissolution of 9c-PF6. The facile formation
of the dinuclear cations 7 and 8 from neutral or cationic,
mononuclear precursors in chlorocarbon solvents in the absence
of nitrile should be emphasized, given the ubiquity of CDCl3
and CH2Cl2 as NMR and reaction solvents. This reactivity pat-
tern is readily overlooked, and in the recently reported synthesis
of RuCl2(binap)(PhCN)2,12 the emergence of an AB quartet now
known to be due to 7c-Cl in the 31P{1H} NMR spectrum in
CH2Cl2 was tentatively and incorrectly attributed to a five-
coordinate species RuCl2(binap)(PhCN).27 Similarly, one of the
two sets of AB patterns reported11 in spectra of RuCl2(binap)-
(MeCN)2 in CDCl3 corresponds rather closely to that found for
8c, as noted above, and should probably be reassigned to Ru2-
Cl3(binap)2(MeCN)2+Cl-. The appearance of two methyl sing-
lets in the corresponding 1H NMR spectrum, cited as evidence
for the mononuclear formulation,11 is consistent with the pres-
ence of coordinated and displaced (free) MeCN in a 1:1 ratio.
The solution behavior of these nitrile derivatives of Ru(II) is
clearly intricate. This complexity provides, however, a flexible
network of synthetic relationships, which can be exploited by
judicious choice of reaction conditions to gain access to a wide
range of products.
Acknowledgment. We thank the Natural Sciences and
Engineering Research Council of Canada for financial support
and Johnson Matthey Ltd. and Colonial Metals Inc. for a loan
.
of RuCl3 3H2O. D.E.F. thanks the Killam Foundation for a
postgraduate fellowship, administered by the University of
British Columbia.
Supporting Information Available: Tables S1 and S2, containing
microanalytical and 1H NMR data (4 pages). Ordering information is
given on any current masthead page.
(24) (a) Ball, G. E.; Cullen, W. R.; Fryzuk, M. D.; James, B. R.; Rettig, S.
J. Organometallics 1991, 10, 3767. (b) Leoni, P. Organometallics
1993, 12, 2432 and references therein. (c) Freedman, D. A.; Mann,
K. R. Inorg. Chem. 1991, 30, 836. (d) Chang, J.; Bergman, R. G. J.
Am. Chem. Soc. 1987, 109, 4298. (e) Marder, T. B.; Fultz, W. C.;
Calabrese, J. C.; Harlow, R. L.; Milstein, D. J. Chem. Soc., Chem.
Commun. 1987, 1543.
(25) Newbound, T. D.; Colsman, M. R.; Miller, M. M.; Wulfsberg, G. P.;
Anderson, O. P.; Strauss, S. H. J. Am. Chem. Soc. 1989, 111, 3762.
(26) Bown, M.; Waters, J. M. J. Am. Chem. Soc. 1990, 112, 2442.
IC960735Y
(27) No evidence for such a coordinatively unsaturated species was in fact
observed in this chemistry. We recently reassigned17 the 31P{1H} NMR
parameters assigned to RuCl(dppb)(MeCN)2+ in an earlier report13 to
Ru(dppb)(MeCN)42+(PF6-)2.