Inhibiting δ → π Isomerization of Aryloxide Ligands in Late Transition-Metal Complexes

Article abstract of DOI:10.1021/om049394j

The room-temperature reaction of thallium perfluorophenoxide with RuHCl(PPh₃)₃ (4a) yields RuH(σ-OC₆X ₅(PPh₃)₃ (X = F, 5^F), the structure of which was confirmed by spectroscopic, microanalytical, and X-ray analysis. The analogous phenoxide complex (X = H, 5^H) is known to be unstable toward δ → π isomerization, and reaction of phenoxide ion with 4a affords RuH(η⁵-C₆H₅O)(PPh₃) ₂ (6a). We conclude that the driving force for isomerization in 5^F is minimized by the electron deficiency of the perfluorophenoxide ring. In contrast, incorporation of the aryloxide entity within a chelating framework proves insufficient to inhibit isomerization. Thus, reaction of 4a or RuCl₂(PPh₃)₃ (4b) with the sodium salt of 2′-(diphenylphosphino)-l,l′-binaphthyl-2-ol (NaO-MOP) yields RuX(O-MOP)(PPh₃) (7a, X = H; 7b, X = Cl). Detailed 1- and 2-D NMR analysis of 7b provides unequivocal evidence that the O-MOP ligand is bound to Ru in a σ fashion through phosphorus and in a π fashion via the aryloxide ring. The crystal structure of 5^F·THF is reported.

Full text of DOI:10.1021/om049394j

Organometallics 2005, 24, 103-109
103
Inhibiting σ f π Isomerization of Aryloxide Ligands in
Late Transition-Metal Complexes
Jennifer L. Snelgrove, Jay C. Conrad, Melanie D. Eelman, Maeve M. Moriarty,
Glenn P. A. Yap, and Deryn E. Fogg*
Centre for Catalysis Research and Innovation, Department of Chemistry,
University of Ottawa, 10 Marie Curie, Ottawa, Ontario, Canada K1N 6N5
Received August 5, 2004
The room-temperature reaction of thallium perfluorophenoxide with RuHCl(PPh3)3 (4a)
yields RuH(σ-OC6X5)(PPh3)3 (X ) F, 5F), the structure of which was confirmed by
spectroscopic, microanalytical, and X-ray analysis. The analogous phenoxide complex (X )
H, 5H) is known to be unstable toward σ f π isomerization, and reaction of phenoxide ion
with 4a affords RuH(η5-C6H5O)(PPh3)2 (6a). We conclude that the driving force for
isomerization in 5F is minimized by the electron deficiency of the perfluorophenoxide ring.
In contrast, incorporation of the aryloxide entity within a chelating framework proves
insufficient to inhibit isomerization. Thus, reaction of 4a or RuCl2(PPh3)3 (4b) with the sodium
salt of 2′-(diphenylphosphino)-1,1′-binaphthyl-2-ol (NaO-MOP) yields RuX(O-MOP)(PPh3)
(7a, X ) H; 7b, X ) Cl). Detailed 1- and 2-D NMR analysis of 7b provides unequivocal
evidence that the O-MOP ligand is bound to Ru in a σ fashion through phosphorus and in
a π fashion via the aryloxide ring. The crystal structure of 5F‚THF is reported.
Chart 1
Introduction
Robust “pseudohalide” ligands based on the phenoxide
ion offer considerable promise as sterically and elec-
tronically tunable anionic donors in transition-metal
catalysis. In the area of olefin metathesis, remarkable
selectivities have been achieved in asymmetric ring-
closing metathesis (RCM) using group 6 complexes of
biphenolate and binaphtholate ligands (e.g. 1; Chart
1).1,2 Within the more robust ruthenium catalysts,
impressive selectivity was reported for complexes con-
taining a binaphtholate-derived ligand (e.g. 2),3,4 while
our own group recently described exceptionally high
turnover numbers (>40 000) in RCM via perfluorophe-
noxide derivative 3, the first long-lived Ru-pseudoha-
lide catalyst for olefin metathesis.5
A stable ligand architecture can be critical in deter-
mining catalyst lifetime and selectivity and, ultimately,
catalyst utility. Several important calorimetric and
reactivity studies6-9 have led to reevaluation of the
assumption that late-transition-metal-oxygen bond
energies are inherently low, as the Pearson “hard-soft”
formalism would suggest. Recent work from our group,10
however, suggests that the well-established stability
of the Ru-OAr bond within complexes of the type6
Ru(σ-OAr)2(PMe3)4 may be a function of the nonlability
of the ancillary ligands. Thus, related complexes con-
taining labile PPh3 groups (e.g. RuCl(σ-OAr)(PPh3)3,
RuH(σ-OAr)(PPh3)3; Ar ) Ph, p-tBuC6H4) are not isol-
able, but undergo facile σ f π isomerization, enabled
by loss of phosphine, to generate π-aryloxide derivatives
(e.g. 6, Scheme 1).10-12 We speculated that the sta-
bility of the Ru(σ-OAr) bond in 3 could arise from
the attenuated π-donor capacity of the perfluorophe-
* To whom correspondence should be addressed. E-mail: dfogg@
science.uottawa.ca. Fax: (613) 562-5170.
(1) Hoveyda, A. H.; Schrock, R. R. Chem. Eur. J. 2001, 7, 945.
(2) Tsang, W. C. P.; Hultzsch, K. C.; Alexander, J. B.; Bonitatebus,
P. J.; Schrock, R. R.; Hoveyda, A. H. J. Am. Chem. Soc. 2003, 125,
2652.
(3) Van Veldhuizen, J. J.; Garber, S. B.; Kingsbury, J. S.; Hoveyda,
A. H. J. Am. Chem. Soc. 2002, 124, 4954.
(4) Van Veldhuizen, J. J.; Gillingham, D. G.; Garber, S. B.; Kataoka,
O.; Hoveyda, A. H. J. Am. Chem. Soc. 2003, 125, 12502.
(5) Conrad, J. C.; Amoroso, D.; Czechura, P.; Yap, G. P. A.; Fogg,
D. E. Organometallics 2003, 22, 3634.
(6) Fulton, J. R.; Holland, A. W.; Fox, D. J.; Bergman, R. G. Acc.
Chem. Res. 2002, 35, 44.
(7) Bryndza, H. E.; Tam, W. Chem. Rev. 1988, 88, 1163.
(8) Holland, P. L.; Andersen, R. A.; Bergman, R. G.; Huang, J.;
Nolan, S. P. J. Am. Chem. Soc. 1997, 119, 12800.
(9) Huang, J.; Li, C.; Nolan, S. P.; Petersen, J. L. Organometallics
1998, 17, 3516.
(10) Snelgrove, J. L.; Conrad, J. C.; Yap, G. P. A.; Fogg, D. E. Inorg.
Chim. Acta 2003, 345, 268.
(11) Other examples of such behavior in Ru chemistry include the
formation of RuH(π-ArO)(PR3)2 by reaction of phenols with RuH2-
(PPh3)4, RuH2(H2)2(PCy3)2, or RuH(C6H4PPh2)(PPh3)2(Et2O), and of
n+
Ru(π-ArO)L3 (n ) 1, L3 ) arene; n ) 0, L3 ) Cp*) by reaction of
phenols or phenoxides with [RuCl(π-arene)]2(µ-Cl)2, [Cp*Ru(OMe)]2,
or [Cp*Ru(µ-Cl)]4. See: (a) Cole-Hamilton, D. J.; Young, R. J.; Wilkin-
son, G. J. Chem. Soc., Dalton Trans. 1976, 1995. (b) Rosete, R. O.;
Cole-Hamilton, D. J.; Wilkinson, G. J. Chem. Soc., Dalton Trans. 1979,
1618. (c) Christ, M. L.; Sabo-Etienne, S.; Chung, G.; Chaudret, B. Inorg.
Chem. 1994, 33, 5316. (d) Cole-Hamilton, D. J.; Wilkinson, G. New J.
Chem. 1977, 1, 141. (e) Bennett, M. A.; Matheson, T. W. J. Organomet.
Chem. 1979, 175, 87. (f) Bu¨cken, K.; Koelle, U.; Pasch, R.; Ganter, B.
Organometallics 1996, 15, 3095. (g) Chaudret, B.; Xiaodong, H.; Huang,
Y. J. Chem. Soc., Chem. Commun. 1989, 1844. (h) Koelle, U.; Wang,
M. H.; Raabe, G. Organometallics 1991, 10, 2573. (i) Loren, S. D.;
Campion, B. K.; Heyn, R. H.; Tilley, T. D.; Bursten, B. E.; Luth, K. W.
J. Am. Chem. Soc. 1989, 111, 4712.
10.1021/om049394j CCC: $30.25 © 2005 American Chemical Society
Publication on Web 12/08/2004

104 Organometallics, Vol. 24, No. 1, 2005
Snelgrove et al.
Table 1. 19F{1H} NMR Values for Perfluorophenol
Scheme 1
and Its Metal Complexes
compd
C6F5OH
solvent
CDCl3
δF (ppm)
ref
this worka
-87.7 to -88.1
(m, 4F)
-92.1 to -92.6
(m, 1F)
-81.8 to -84.1
(m, 2F)
b
TlOC6F5
CDCl3
this worka
Scheme 2
-88.55 (m, 2F)
-99.65 (m, 1F)
-86.45 (m, 2F)
-86.7 (m, 2F)
-94.27 (m, 2F)
-94.50 (m, 2F)
-104.8 to -106.0
(m, 2F)
-93.6 (m, 2 F)
-94.7 to -94.9
(m, 2F)
-107.8 (m, 1F)
-108.4 (m, 2F)
-112.3 (m, 2F)
-119.0 (m, 1F)
-98.4 (m, 2F)
-100.6 (m, 2F)
-103.6 (m, 1F)
Ru(σ-OC6F5)2(IMes)-
(py)(CHPh) (3)
CDCl3
5
noxide ring,13 while the chelate effect might be respon-
sible for the stability of 2. With the intention of
clarifying these issues via direct comparison with the
systems already examined, we undertook reaction of the
labile precursors 4 with perfluorophenoxide anion and
with the heterobifunctional phosphine-binaphtholate
ligand NaO-MOP (sodium 2′-(diphenylphosphino)-2-oxy-
1,1′-binaphthyl).
RuH(σ-OC6F5)-
(PPh3)3 (5F)
C6D6
this worka
[Cp*Ru(η5-C6F5O)]
CDCl3
13a, 16
13a, 16
[Cp*Ru(η6-C6F5OH)]+ CDCl3
a 19F NMR spectra measured at 282 MHz, 298 K, reported
relative to trifluoroacetic acid at 0 ppm; the patterns show second-
order effects. b The values in C6D6 are rather similar (-84.41 (m,
2F), -90.16 (m, 2F), -101.83 (m, 1F)), with limited perturbation
by π-stacking17 interactions.
Results and Discussion
Treatment of a THF solution of RuHCl(PPh3)3 (4a)
with thallium pentafluorophenoxide effected complete
conversion to the σ-aryloxide complex 5F within 3 h at
22 °C (Scheme 2).14 In contrast with the corresponding
protio-phenoxide species RuH(OC6H5)(PPh3)3 (5H), com-
plex 5F shows no tendency to isomerize. The red product
was isolated in ca. 80% yield, and NMR and combustion
analysis are consistent with formulation of the complex
as RuH(OC6F5)(PPh3)3. Room-temperature 31P{1H} NMR
analysis reveals a broad singlet centered at 47 ppm,
which resolves into an A2B pattern at -50 °C (δP 87.1
(t, 1P), 40.9 (d, 2P); 2JPP ) 26 Hz), confirming retention
of all three triphenylphosphine ligands. The hydride
signal is observed as a quartet of triplets, split by long-
range coupling to two 19F nuclei, as well as the expected
3, consistent with σ-coordination of the aryloxide ring
(Table 1). A comparable range is found for the 19F nuclei
within free perfluorophenol, as well as thallium per-
fluorophenoxide. In comparison, signals for the π-bound
perfluoroaryloxy ring in [CpRu(η5-C6F5O)] or [CpRu(η6-
C6F5OH)]+ are found 20-30 ppm upfield.13a,16 The
coordination mode of the aryloxide ligand within 5F is
confirmed by X-ray analysis of crystals obtained by slow
evaporation of a concentrated THF solution. An ORTEP
diagram, with relevant bond lengths and angles, is
shown in Figure 1.
Having confirmed the efficacy of electronic param-
eters in determining the coordination mode of the
aryloxide entity, we turned our attention to the pos-
sibility that incorporating an aryloxy group within a
chelating framework can likewise prevent isomerization.
An attractive candidate for exploration is the chiral
phosphine auxiliary “HO-MOP” (2′-(diphenylphosphino)-
1,1′-binaphthyl-2-ol), a rare bidentate member of the
monophosphine or MOP family of binaphthol-derived
ligands developed by Hayashi et al.18,19 We prepared
racemic HO-MOP in 82% yield from binaphthol: in a
minor modification of the literature route (Scheme 3),18
we used trichlorosilane to reduce the phosphine oxide
and to deprotect the triflate in a single step.
2
two-bond coupling to phosphorus (δH -22.3 ppm; JPH
) 27.9 Hz, 5JFH ) 7.2 Hz).15 The coupling assignments
are supported by 1H{19F} and 1H{31P} NMR experi-
ments, which reveal the hydride signal as a quartet and
a triplet, respectively.
19F NMR analysis of 5F reveals signals at chemical
shifts similar to those found for the σ-OAr ligand within
(12) While of limited utility for olefin metathesis, such π-aryloxide
species are of interest as catalysts for asymmetric Diels-Alder
reactions (see: Faller, J. W.; D’Alliessi, D. G. Organometallics 2003,
22, 2749. Faller, J. W.; Lavoie, A. R.; Grimmond, B. J. Organometallics
2002, 21, 1662) and are of potential relevance for nucleophilic phenol
functionalization (see: Le Bras, J.; Amouri, H.; Vaissermann, J. Inorg.
Chem. 1998, 37, 5056 and references therein) or carbonylation (see:
Dockter, D. W.; Fanwick, P. E.; Kubiak, C. P. J. Am. Chem. Soc. 1996,
118, 4846. Rees, W. M.; Churchill, M. R.; Fettinger, J. C.; Atwood, J.
D. Organometallics 1985, 4, 2179).
(13) Examples of π-bound fluorophenoxide have, however, been
reported. See: (a) Koelle, U.; Ho¨rnig, A.; Englert, U. Organometallics
1994, 13, 4064. (b) Curnow, O. J.; Hughes, R. P. J. Am. Chem. Soc.
1992, 114, 5895.
The HO-MOP ligand is readily converted into its
sodium salt by reaction with Na2CO3. Subsequent
reaction with 4a or 4b in CH2Cl2 affords orange RuX-
(O-MOP)(PPh3) (X ) H, 7a; X ) Cl, 7b; Scheme 4)
within 24 h at 22 °C. Strong evidence for the proposed
formulation is provided by inert-atmosphere MALDI
(14) Similar behavior is observed with sodium pentafluorophenoxide,
but the reaction is only ca. 30% complete at 3 h (100% at 48 h). The
accelerating effect of Ag+ and Tl+ salts in halide substitution of metal
complexes has been attributed to electrophilic “pull” by the heavy-metal
cations, possibly via direct electrophilic attack on the halogen atom.
See: Danopoulos, A. A.; Wilkinson, G.; Sweet, T. K. N.; Hursthouse,
M. B. Polyhedron 1994, 13, 2899.
(16) Literature values, originally reported relative to CFCl3, are cited
relative to trifluoroacetic acid at 0 ppm.
(17) Williams, J. H. Acc. Chem. Res. 1993, 26, 593.
(18) Uozumi, Y.; Tanahashi, A.; Lee, S. Y.; Hayashi, T. J. Org. Chem.
1993, 58, 1945.
(15) The JPF coupling is too small to observe in the 31P{1H} NMR
5
spectrum.
(19) Hayashi, T. Acc. Chem. Res. 2000, 33, 354.

Inhibiting σ f π Isomerization of Aryloxide Ligands
Organometallics, Vol. 24, No. 1, 2005 105
Figure 1. ORTEP representation of RuH(σ-OC6F5)(PPh3)3
(5F) with thermal ellipsoids at the 30% probability level.
Phenyl groups are represented by their ipso carbon atoms,
and the THF solvate is omitted for clarity. Selected bond
distances (Å) and angles (deg): Ru-O1 ) 2.1350(17),
Ru-P1 ) 2.2242(6), Ru-P3 ) 2.3389(7), Ru-P2 ) 2.3911-
(7), Ru-H1 ) 1.49(3), O1-C6 ) 1.311(3); O1-Ru-P1
Figure 2. MALDI-MS spectra (anthracene matrix), show-
ing simulated (top) and observed (bottom) isotope patterns
for M+. within (a) 7a (observed m/z 818.3, calculated m/z
818.1); (b) 7b (observed m/z 852.0, calculated m/z 852.1).
Chart 2. Reported (A-F) and Proposed (7)
Coordination Modes for Binap and Related MOP
Ligands
)
135.23(6), O1-Ru-P3 ) 94.91(5), P1-Ru-P3 )
99.22(3), O1-Ru-P2 ) 85.40(5), P1-Ru-P2 ) 98.42(3),
P3-Ru-P2 ) 154.40(2), H1-Ru-P1 ) 86(1), H1-Ru-P2
) 82(1), H1-Ru-P3 ) 81(1), H1-Ru-O1 ) 138(1),
C6-O1-Ru ) 135.80(15).
Scheme 3
Scheme 4
analysis. Attempts to circumvent this behavior by
carrying out the reaction in THF resulted in complex
mixtures, as we previously noted in the corresponding
phenoxide chemistry,10 while reactions in benzene did
not proceed to completion. Subsequent experiments
therefore targeted isolation and characterization of 7b.
We obtained the latter light orange complex in ca. 90%
yield, following chromatography to remove traces of an
unknown byproduct (δP 52 ppm, s). The stability of 7b
to aerobic chromatography is noteworthy. NMR samples
in C6D6 or CDCl3 likewise show no evidence of decom-
position in air for up to 3 days at room temperature,
and the complex also fails to react with sterically
undemanding donors such as CO and MeCN. All of
these features point toward a nonlabile, coordinatively
saturated structure.
While binap is conventionally considered a simple
κ2-PP donor (see A, Chart 2), a wealth of alternative
coordination modes (B-F) has been established for
binap and related MOP ligands in a series of elegant
NMR and X-ray crystallographic studies.21-32 1H NMR
analysis of 7b indicates the presence of a π-bound ring:
while most of the naphthyl protons appear in the
mass spectrometric analysis. For both 7a and 7b, a
peak-for-peak correspondence is found between the
simulated and observed isotope patterns for the molec-
ular ions (Figure 2). The complexity of these patternss
a function of the presence of seven naturally occurring
isotopes of ruthenium and (in 7b) two isotopes of
chlorinesprovides powerful insight into the molecular
composition. Critical to their observation is use of a
matrix that ionizes by charge transfer, rather than
protonation.20
31P NMR spectra of 7a and 7b show well-resolved AB
systems (7a, δ 63.1, 56.3, 2JPP ) 25 Hz; 7b, δ 49.9, 46.9,
2JPP ) 58 Hz), accompanied, in the case of 7a, by a
2
1
hydride triplet (-11.84 ppm, JPH ) 34 Hz) in the H
NMR spectrum. Isolation of 7a was complicated by
solvent-induced chlorination of the hydride ligand to
generate 7b (Scheme 4), as indicated by 31P NMR
(20) Eelman, M. D.; Moriarty, M. M.; Fogg, D. E. Manuscript in
preparation.

106 Organometallics, Vol. 24, No. 1, 2005
Snelgrove et al.
Table 2. 13C NMR Chemical Shifts (ppm) for the
aromatic region, two sets of doublets of doublets, each
σ,π-Bound Ring in 7b and Related Complexesa
integrating to 1H, are found further upfield (δH 4.50,
3
3
3JHH ) 7.4 Hz, JHP ) 3.9 Hz; δH 6.01, JHH ) 7.4 Hz,
27
7b
B,21-26 B′
Cb,28 D28
E26
F22,24,25,29-32
3JHP ) 2.5 Hz). Both signals collapse to doublets on 31
P
C1
C2
C3
C4
83.0
155.6
97.3
87-101
55-73
70.5 87 140-141
92-120
73-101
100-108
89-109
107-121
110-118
decoupling. These protons couple (TOCSY, COSY) to
each other but to no other 1H nuclei, supporting assign-
ment to H3 and H4, respectively, of the π-bound
naphthol ring.33 As the multiplicity suggests, each
couples to a single 31P nucleus: 1H-31P HMQC analysis
confirms that the signal for H3 (4.50 ppm) correlates
strongly only with the 31P NMR signal for PPh3 (46.9
ppm) and that for H4 (6.01 ppm) only with the 31P NMR
signal for the O-MOP phosphorus (49.9 ppm).34 This
implies a dihedral angle of ca. 90° between H3 and the
O-MOP phosphorus and between H4 and PPh3. Con-
sistent with this is the geometry shown for 7 in Scheme
4, with an anti disposition of the PPh3 ligand and the
aryloxy oxygen.35 Each of the protons H3 and H4 also
correlates with an upfield 13C{1H} NMR doublet (C3,
193.7 166 171-174
127.6 113 136
142.8 147 123
146.1 140 133
128.8 129 131
125-127
138.3
94.7
C4a 112.6
C8a 103.5
142-145
130-133
a For numbering, see Chart 2. b Values tabulated for E ) O only.
2
2
δC 97.3, JCP ) 10.0 Hz; C4, δC 94.7, JCP ) 8.9 Hz;
1H-13C HMQC) that collapses to a singlet on simulta-
neous 31P and 1H decoupling. While the absence of two
2JCP couplings for C3 and C4 is unexpected, similar
multiplicities are reported for related piano-stool com-
plexes of type F, in which two phosphine donors are
present.22,26,29,30
13C NMR analysis is invaluable in establishing the
hapticity of π-bound aromatic rings. Table 2 summarizes
the 13C NMR chemical shifts for the σ,π-bound ring in
7b and the reported complexes of Chart 2. Of particular
interest among the latter is Pd(acac)(κ1: κ1-O-MOP) (8),
a complex of class C recently reported by Pregosin,
Albinati, and co-workers.28 NMR and X-ray analysis
revealed that the binaphthol ring in 8 is bound via C1
and that the C-O bond has significant ketonic charac-
Figure 3. 1H-13C HMBC NMR spectrum for 7b (500
(21) Feiken, N.; Pregosin, P. S.; Trabesinger, G.; Albinati, A.; Evoli,
G. L. Organometallics 1997, 16, 5756.
MHz, 298 K, CDCl3).
(22) den Reijer, C. J.; Woerle, M.; Pregosin, P. S. Organometallics
2000, 19, 309.
ter. Notably, the “phenolic” carbon C2 gives rise to a
13C NMR signal at 194 ppm, a location characteristic of
the carbonyl carbon in R,â-unsaturated ketones. In
contrast, the corresponding carbon signal for 7b appears
at 155.6 ppm, very near the value of 151 ppm found for
the free ligand. The remaining quaternary carbon
signals of the π-bound ring are readily assigned on the
basis of their upfield location, their disappearance on
DEPT-135 analysis, and the couplings revealed by a
1H-13C HMBC experiment36 (Figure 3). The HMBC
correlations, reinforced by 1H-1H TOCSY analysis
(Figure 4), as well as HMQC experiments, enable full
(23) den Reijer, C. J.; Dotta, P.; Pregosin, P. S.; Albinati, A. Can. J.
Chem. 2001, 79, 693.
(24) Geldbach, T. J.; Pregosin, P. S.; Albinati, A.; Rominger, F.
Organometallics 2001, 20, 1932.
(25) Geldbach, T. J.; Pregosin, P. S. Eur. J. Inorg. Chem. 2002, 1907.
(26) Geldbach, T. J.; Pregosin, P. S.; Albinati, A. Organometallics
2003, 22, 1443.
(27) Lloyd-Jones, G. C.; Stephen, S. C.; Murray, M.; Butts, C. P.;
Vyskocil, S.; Kocovsky, P. Chem. Eur. J. 2000, 6, 4348.
(28) Dotta, P.; Kumar, P. G. A.; Pregosin, P. S.; Albinati, A.; Rizzato,
S. Organometallics 2003, 22, 5345.
(29) Geldbach, T. J.; Drago, D.; Pregosin, P. S. Chem. Commun.
2000, 1629.
(30) Geldbach, T. J.; Pregosin, P. S.; Bassetti, M. Organometallics
2001, 20, 2990.
1
assignment of the remaining 13C and H nuclei in the
(31) Hermatschweiler, R.; Pregosin, P. S.; Albinati, A.; Rizzato, S.
Inorg. Chim. Acta 2003, 354, 90.
naphthol ring: all signals, with the exception of that
for C1, due to the “aryloxy” ring appear at considerably
lower chemical shifts than they do in 8 (Table 2). Indeed,
a closer resemblance exists between 7 and the κ1: η6
structure F, numerous examples of which have now
been reported.22,24,25,29-32
(32) Geldbach, T. J.; Breher, F.; Gramlich, V.; Kumar, P. G. A.;
Pregosin, P. S. Inorg. Chem. 2004, 43, 1920.
(33) While the corresponding protons (H3′, H4′) of the phosphorus-
functionalized ring will give rise to an identical set of COSY correla-
tions and would likewise be shifted upfield if this ring were π-bound,
the PPh2 group would then result in a 31P NMR signal near 0 ppm.25
(The signals for H3′ and H4′, located from the TOCSY and COSY NMR
3
3
spectra, in fact appear at 7.45 ppm (H3′: dd, JHH ) 8.5 Hz; JHP
)
3
4
6.9 Hz) and 7.91 ppm (H4′: dd, JHH ) 8.6 Hz; JHP ) 1.5 Hz); they
are distinguished on the basis of the magnitude of their respective JHP
values. We distinguish between the signals for H3 and H4, in turn, on
2
(36) It will be noted that three of the anticipated JHC correlations
(those between H3, C2, and C4 and between H6 and C7) are missing
from the HMBC spectrum. Owing to the small size of the 2JHC coupling
constants in aromatic rings (often on the order of 1 Hz), these
3
the basis of the number and identity of JHC correlations revealed in
HMBC experiments. For further details, see the Supporting Informa-
tion.
3
correlations are generally less reliable than the corresponding JHC
(34) Assignment of the 31P NMR signals is made on the basis of the
coupling of the O-MOP phosphorus to H3′ and H4′.
(35) This geometry permits an incipient interaction between the
oxygen and phosphorus O-MOP substituents, a feature that may
account for the diastereomeric bias in formation of 7b (vide infra).
couplings, for which values of 5-10 Hz are typical (see: Kalinowski,
H.-O.; Berger, S.; Braun, S. Carbon-13 NMR Spectroscopy; Wiley:
3
Toronto, 1988) The chemical shift and the TOCSY, HMQC, and JHC
HMBC correlations for the 1H NMR signal at 4.50 ppm are uniquely
consistent with assignment to H3.

Inhibiting σ f π Isomerization of Aryloxide Ligands
Organometallics, Vol. 24, No. 1, 2005 107
ligands in late-transition-metal catalysts. In context of
the findings above, the apparent stability of the Hov-
eyda complex3,4 2 to σ f π isomerization is of consider-
able interest. We speculate that this may in fact reflect
the rather low lability of the dative Ru-O(ether) bond
(as indeed implied by the chromatographic stability and
the slow initiation of 23). The active catalyst, generated
by metathetical exchange of the chelating alkylidene,
is coordinatively unsaturated and may thus be unstable
toward σ f π isomerization. Indeed, derivatives of 2
modified to destabilize the ether chelate appear to
decompose much more readily.4 Conversely, the electron
deficiency of the perfluorophenoxide ligand in complex
3, which disfavors isomerization, is presumably a key
factor in the remarkable lifetime of this catalyst.
Experimental Section
General Procedures. All reactions were carried out at
room temperature (22 °C) under N2 using standard Schlenk
or drybox techniques, unless stated otherwise. Dry, oxygen-
free solvents were obtained using an Anhydrous Engineering
solvent purification system and stored over Linde 4 Å molec-
ular sieves. CDCl3 and C6D6 were dried over activated sieves
(Linde 4 Å) and degassed by consecutive freeze/pump/thaw
cycles. RuHCl(PPh3)3 (4a) and RuCl2(PPh3)3 (4b) were pre-
pared as previously described.38-40 (()-2′-(Diphenylphosphino)-
2-(((trifluoromethyl)sulfonyl)oxy)-1,1′-binaphthyl was prepared
over two steps from racemic binaphthol according to the
literature procedure.18 1H NMR (300 or 500 MHz), 31P NMR
(121 or 202 MHz), 19F NMR (282 MHz), and 13C NMR (75 or
125 MHz) spectra were recorded on a Bruker Avance-300,
Avance-500, or AMX-500 spectrometer. NMR spectra are
reported relative to external 85% H3PO4 (31P), trifluoroacetic
acid (19F), or TMS (1H, 13C) at 0 ppm. Variable temperature,
2-D, and selective decoupling NMR experiments were carried
out on the Bruker AMX-500 and Bruker Avance-500 instru-
ments, the latter being equipped with a triple-resonance probe.
IR spectra were measured on a Bomem MB100 IR spectrom-
eter. Inert-atmosphere MALDI-MS analyses were performed
using a Bruker OmniFlex MALDI TOF mass spectrometer
equipped with a nitrogen laser (337 nm) and interfaced to an
MBraun Labmaster 130 glovebox. Data were collected in
positive reflection mode, with the accelerating voltage held at
20 kV for all experiments. Matrix (anthracene) and analyte
solutions were prepared in CH2Cl2 at concentrations of 20 and
1 mg/mL, respectively; samples were mixed in a matrix to
analyte ratio of 20:1. Microanalyses were carried out by Guelph
Chemical Laboratories Ltd., Guelph, Ontario, Canada.
Note! The toxicity of thallium (particularly in the +1
oxidation state) is well established, and the subject has been
recently reviewed.41 Care must be taken to prevent introduc-
tion into the body by inhalation, by ingestion via contaminated
hands or gloves, or through the skin. All thallium reagents
and wastes, including contaminated solvents, were handled
using double-glove and secondary containment procedures,
with separate disposal of all wastes in accordance with
government regulations.
Figure 4. 1H-1H TOCSY spectrum of 7b (500 MHz, 298
K, CDCl3).
Coordination of “tethered” binaphthyl and biphenyl
ligands within a piano-stool structure can give rise to
diastereomers, if a sterogenic metal center is present.
We observe only one AB pattern in the 31P NMR
spectrum of 7b, however, even at -90 °C, suggesting
the presence of a single diastereomer. (We exclude the
possibility of rapid epimerization of the metal center,
given the nonlability of the phosphine ligand in reac-
tions with donors such as carbon monoxide and aceto-
nitrile; vide supra.) While Faller has described biphenyl
complexes in which unfavorable steric interactions
between the bulky PPh3 ligand and a dimethylamino
ring substituent dictate formation of a single diastere-
omer,37 the limited bulk of the O-MOP oxygen in 7b
should minimize such effects. The observed diastereo-
meric bias may instead arise from a favorable interac-
tion between the PPh3 ligand and the O-MOP oxygen.
Conclusions
The retention in 5F of a stable, σ-bonded perfluo-
rophenoxide and three labile PPh3 ligands contrasts
with the instability of the corresponding phenoxide
derivatives, which undergo rapid isomerization, with
loss of PPh3, to yield π-OAr complexes of type 6.10 From
this difference in behavior, we suggest that the electron
deficiency of the aryloxide ring is a key factor contribut-
ing to the stability of the Ru(σ-OAr) entity within
complexes such as 5F and 3. In contrast, the binaphthol-
derived O-MOP ligand in 7 is bound in a σ fashion
through phosphorus and in a π fashion via the η6-
aryloxide ring. Chelation thus appears insufficient to
inhibit σ f π isomerization of the Ru-OAr unit, where
the lability of the ancillary ligands in the precursor
complex is sufficient to liberate the three coordination
sites required to bind the phenolic ring.
(()-2′-(Diphenylphosphino)-1,1′-binaphthyl-2-ol (HO-
MOP). In a modification of the literature route,18 a solution
of (()-2′-(diphenylphosphino)-2-(((trifluoromethyl)sulfonyl)-
oxy)-1,1′-binaphthyl (677.4 mg, 1.124 mmol) in 20 mL of
toluene was treated with trichlorosilane (761 mg, 5.62
(38) Amoroso, D.; Snelgrove, J. L.; Conrad, J. C.; Drouin, S. D.; Yap,
G. P. A.; Fogg, D. E. Adv. Synth. Catal. 2002, 344, 757.
(39) Stephenson, T. A.; Wilkinson, G. J. Inorg. Nucl. Chem. 1966,
28, 945.
(40) Hallman, P. S.; Stephenson, T. A.; Wilkinson, G. Inorg. Synth.
1970, 12, 237.
(41) Galva´n-Arzate, S.; Santamar´ıa, A. Toxicol. Lett. 1998, 99, 1.
The broader significance of these findings lies in their
implications for the deployment of chelating aryloxide
(37) Faller, J. W.; D’Alliessi, D. G. Organometallics 2003, 22, 2749.

108 Organometallics, Vol. 24, No. 1, 2005
Snelgrove et al.
Table 3. Crystal Data and Structure Refinement
mmol) and Et3N (7 equiv). The solution was stirred at 100 °C
for 16 h and then cooled, diluted with Et2O, and quenched
with saturated NaHCO3. The resulting white gel was fil-
tered through Celite, which was then washed with Et2O (3 ×
30 mL). The organic phase was dried over MgSO4, concen-
trated, and chromatographed on silica gel (20% EtOAc/
hexanes) to afford the product as a white solid. Yield: 409 mg
(80%). NMR data are in good agreement with the reported
values.18 1H NMR (C6D6, 300 MHz): δ 8.22-6.42 (m, 22H,
aromatic), 4.63 (br s, 1H, -OH). 31P{1H} NMR (C6D6, 121
MHz): δ -13.4 (s).
Sodium (()-2′-(Diphenylphosphino)-2-oxy-1,1′-binaph-
thyl (NaO-MOP). Solid Na2CO3 (732 mg, 6.91 mmol) was
added to a solution of HO-MOP (93.0 mg, 0.205 mmol) in 5
mL of CH2Cl2. The suspension was stirred at room tempera-
ture for 24 h and then filtered through Celite to remove
NaHCO3 and excess Na2CO3. The solvent was removed under
vacuum to give the desired salt as a white solid. Yield: 87 mg
(89%).
Details for RuH(OC6F5)(PPh3)3‚THF (5F‚THF)
empirical formula
formula wt
C64H56F5O2P3Ru
1146.07
temp
203(2) K
0.71073 Å
wavelength
cryst syst, space group
unit cell dimens
monoclinic, P21/n
a ) 12.773(2) Å
b ) 25.338(5) Å
c ) 17.282(3) Å
â ) 105.381(3)°
5392.9(17) Å3
V
Z, calcd density
abs coeff
4, 1.411 g/cm3
0.44 mm-1
2360
F(000)
cryst size
θ range for data collecn
limiting indices
0.20 × 0.10 × 0.05 mm
1.46-28.65°
-16 e h e 16, 0 e k e 32,
0 e l e 23
no. of rflns collected/unique
completeness to θ ) 28.65°
abs cor
max and min transmissn
refinement method
37 344/12 027 (R(int) ) 0.0441)
86.7%
semiempirical from equivalents
1.000000 and 0.626317
full-matrix least squares on F2
RuH(OC6F5)(PPh3)3 (5F). Thallium pentafluorophenoxide
(176 mg, 0.45 mmol) was added to a purple solution of RuHCl-
(PPh3)3 (4a; 420 mg, 0.45 mmol) in 20 mL of THF. The mixture
was stirred for 3 h, over which time it turned red. The solvent
was removed under reduced pressure, and the resulting oil
was dissolved in benzene and filtered through Celite to remove
TlCl. The filtrate was concentrated and treated with pentane
to precipitate a red solid, which was filtered off. Yield after
drying in vacuo: 392 mg (81%). Blocklike crystals suitable for
X-ray analysis were grown by slow evaporation of a saturated
THF solution. 1H NMR (C6D6): δ 7.40-7.31 (m, 17H, Ar),
6.94-6.90 (m, 10H, Ar), 6.85-6.80 (m, 18H, Ar), -22.29 (q of
t, 2JPH ) 27.9 Hz, 5JFH ) 7.2 Hz, 1H, RuH). 31P{1H} NMR (298
K, C6D6): δ 47 (br s). 31P{1H} NMR (248 K, C7D8): δ 87.1 (t,
no. of data/restraints/params 12 027/0/635
goodness of fit on F2
1.033
final R indices (I > 2σ(I))
R indices (all data)
R1 ) 0.0403,a wR2 ) 0.1003b
R1 ) 0.0561, wR2 ) 0.1075
0.584 and -0.963 e Å-3
largest diff peak and hole
2
1/2
a R1 ) ∑||Fo| - |Fc||/∑||Fo|, b wR2 ) [∑wδ2/∑wFo
] .
3JHH ) 7.4 Hz, 1H, H5), 7.51-7.49 (m, 1H, Ph; overlaps with
3
3
H5), 7.45 (dd, JHH ) 8.5 Hz; JHP ) 6.9 Hz, 1H, H3′), 7.42-
3
7.37 (m, 3H, Ph), 7.37 (t, JHH ) 7.6 Hz, 1H, H6), 7.35-6.74
(m, 17H, Ph), 6.38-6.34 (m, 2H, Ph; overlaps with H7), 6.34
2
2JPP ) 26 Hz, 1P), 40.9 (d, JPP ) 26 Hz, 2P). 13C{1H} NMR
(t, 3JHH ) 8.4 Hz, 1H, H7), 6.01 (dd, 3JHH ) 7.4 Hz; 3JHP ) 2.5
Hz, 1H, H4), 5.93 (d, 3JHH ) 8.5 Hz, 1H, H8), 4.50 (dd, 3JHH
)
(C6D6): δ 144.7-139.6 (m, OC6F5), 136.6-136.1 (m, Ph),
134.9-134.6 (m, Ph), 132.5 (d, J ) 9.6 Hz, Ph), 129.2 (s, Ph).
19F{1H} NMR (282.4 MHz, C6D6): δ -93.6 (m, 2 F), -94.6 to
-94.9 (m, 2F), -107.8 (m, 1F). 19F{1H} NMR (223 K, C7D8): δ
-88.8 to -88.9 (m, 1F, Fo), -92.01 (t, 2JFF ) 20.9 Hz, 1F, Fm),
7.4 Hz, JHP ) 3.9 Hz, 1H, H3). 13C{1H} NMR (CDCl3, 125
3
2
MHz): δ 155.6 (s, C2), 133.8 (d, JCP ) 10.0 Hz), 133.7 (s),
2
133.0 (d, JCP ) 11.3 Hz), 131.1 (s, C6 + C7), 128.9 (s, C8),
2
128.5 (s, C5), 112.6 (s, C8a), 103.5 (s, C4a), 97.3 (d, JCP
)
2
10.0 Hz, C3), 94.7 (d, 2JCP ) 8.9 Hz, C4), 83.0 (s, C1). MALDI-
MS: m/z found, 852.0; m/z calcd for [M]•+, 852.1. Anal. Calcd
for RuCl(O-MOP)(PPh3)‚(hexane) (C56H51ClOP2Ru): C, 71.67;
H, 5.48. Found: C, 71.45; H, 5.25.
-93.38 (t, JFF ) 22.0 Hz, 1F, Fm), -102.5 to -102.6 (m, 1F,
Fo), -106.9 to -107.0 (m, 1F, Fp). A low-temperature 19F-19F
COSY experiment shows the following correlations (ppm): -89
and -93; -92 and -102; -92 and -107; -93 and -107. IR
(Nujol): ν(Ru-H) 2089 cm-1. Anal. Calcd for RuC60H46P3OF5:
C, 67.22; H, 4.33. Found: C, 66.88; H, 4.05.
Attempted Reaction of RuCl(O-MOP)(PPh3) (7b) with
CO and MeCN. A solution of 7b (10.2 mg, 0.012 mmol) in 1
mL of C6D6 was freeze-thaw degassed (× 5) and stirred under
CO (1 atm) or with MeCN (1 mL) under Ar. No change was
evident by 31P{1H} NMR analysis after 24 h.
RuH(O-MOP)(PPh3) (7a). RuHCl(PPh3)3 (4a; 50.1 mg,
0.0541 mmol), HO-MOP (26.2 mg, 0.0571 mmol), and Na2CO3
(143 mg, 1.35 mmol) were added to 2 mL of CDCl3 and stirred
for 24 h at room temperature. In situ 31P NMR analysis showed
signals for 7a and 7b. The corresponding reaction using NaO-
MOP in C6D6 showed considerable residual starting material,
probably owing to the limited solubility of the NaO-MOP salt;
X-ray Crystallography. Details of the crystal data and
structure refinement for 5F‚THF are contained in Table 3.
Crystals of 5F were mounted on a thin glass fiber using
paraffin oil and cooled to the data collection temperature. Data
were collected on a Bruker AXS SMART 1k CCD diffractome-
ter using 0.3° ω scans at 0, 90, and 180° in φ. Initial unit-cell
parameters were determined from 60 data frames collected
at different sections of the Ewald sphere. Semiempirical
absorption corrections based on equivalent reflections were
applied.42 Systematic absences in the diffraction data and unit-
cell parameters were uniquely consistent with the space group
P21/n. The structure was solved by direct methods, completed
with difference Fourier syntheses, and refined with full-matrix
least-squares procedures based on F2. The hydride ligand, H(1),
was located from the difference map. A cocrystallized THF
molecule was found in the initial solution, severely disordered
at an inversion center. The data were modified with the
Squeeze bypass filter of PLATON43 with 629.8 Å3 solvent
accessible volume and 99 electron count/cell consistent with
the initial solution and noncrystallographic data. All nonhy-
1
in THF, multiple products were evident. H NMR (CDCl3): δ
2
-11.84 (t, JPH ) 34 Hz, 1H, RuH). 31P{1H} NMR (CDCl3;
peaks given for 7a only): δ 63.1 (d, 2JPP ) 25 Hz, 1P), 56.3 (d,
2JPP ) 25 Hz, 1P). MALDI-MS, m/z found, 818.3; m/z calcd for
[M]•+, 818.1.
RuCl(O-MOP)(PPh3) (7b). To a solution of RuCl2(PPh3)3
(4b; 106.0 mg, 0.111 mmol) in CH2Cl2 (4 mL) was added a
solution of NaO-MOP (52.7 mg, 0.111 mmol) in 3 mL of
CH2Cl2. The solution was stirred for 24 h, following which the
solvent was removed, the residue redissolved in C6H6 and
filtered (Celite), and the filtrate concentrated and treated with
pentane to precipitate the product. Chromatography on silica
gel (EtOAc eluant; crude product dissolved in 0.5 mL of
CH2Cl2) and reprecipitation from benzene/hexanes, followed
by washing with hexanes, afforded the orange product. Yield:
86 mg (91%). 31P{1H} NMR (CDCl3, 121 MHz): δ 49.9 (d, 2JPP
) 58 Hz, 1P), 46.9 (d, 2JPP ) 58 Hz, 1P). 1H NMR (CDCl3, 500
3
3
MHz): δ 7.97 (t, JHH ) 7.3 Hz, 2H, Ph), 7.91 (dd, JHH ) 8.6
(42) Blessing, R. Acta Crystallogr. 1995, A51, 33.
(43) Spek, A. L. Acta Crystallogr. 1990, A46, C34.
4
Hz; JHP ) 1.5 Hz, 1H, H4′), 7.62-7.52 (m, 4H, Ph), 7.51 (d,

Inhibiting σ f π Isomerization of Aryloxide Ligands
Organometallics, Vol. 24, No. 1, 2005 109
dride hydrogen atoms were treated as idealized contributions.
All scattering factors are contained in the SHELXTL 6.12
program library.44
Ontario Innovation Trust, and the Ontario Research
and Development Corporation.
Supporting Information Available: Figures and tables
giving crystallographic data for 5F‚THF and detailed 2-D NMR
spectra for 7b. This material is available free of charge via
the Internet at http://pubs.acs.org.
Acknowledgment. This work was supported by the
Natural Sciences and Engineering Research Council of
Canada, the Canada Foundation for Innovation, the
(44) Sheldrick, G. M. Bruker AXS, Madison, WI, 2001.
OM049394J