Preagostic Rh-H interactions and C-H bond functionalization: A Combined Experimental and Theoretical Investigation of Rhodium (I) Phosphinite Complexes

Article abstract of DOI:10.1021/om050700i

Three Rh-phosphinite complexes with the general structural formula [RhCl(i-Pr₂POXy)(L)]₂ (Xy = 2,3-xylyl; L = PPh ₃, PMe₃, t-BuNC) were synthesized. Characterization of these complexes using crystallographic and spectroscopic techniques revealed rare examples of preagostic C-H...M interactions. ¹H NMR chemical shielding calculations on a geometry-optimized model complex were used to provide a connection between the solution and solid-state data, which additionally supported assignment of the preagostic interaction. One of these Rh-phosphinite complexes (L = PPh₃) was used to catalyze the ortho alkylation of phosphinites and phenols with an unactivated alkene. Finally, DFT calculations were used to provide evidence for the involvement of the observed preagostic interaction in the cyclometalation step of the catalytic cycle.

Full text of DOI:10.1021/om050700i

Organometallics 2005, 24, 5737-5746
5737
Preagostic Rh-H Interactions and C-H Bond
Functionalization: A Combined Experimental and
Theoretical Investigation of Rhodium(I) Phosphinite
Complexes
Jared C. Lewis,† Jessica Wu,† Robert G. Bergman,*,†,‡ and Jonathan A. Ellman*,†
Department of Chemistry, University of California, Berkeley, and the Division of Chemical
Sciences, Lawrence Berkeley National Laboratory, Berkeley, California 94720
Received August 12, 2005
Three Rh-phosphinite complexes with the general structural formula [RhCl(i-Pr2POXy)(L)]2
(Xy ) 2,3-xylyl; L ) PPh3, PMe3, t-BuNC) were synthesized. Characterization of these
complexes using crystallographic and spectroscopic techniques revealed rare examples of
1
preagostic C-H‚‚‚M interactions. H NMR chemical shielding calculations on a geometry-
optimized model complex were used to provide a connection between the solution and solid-
state data, which additionally supported assignment of the preagostic interaction. One of
these Rh-phosphinite complexes (L ) PPh3) was used to catalyze the ortho alkylation of
phosphinites and phenols with an unactivated alkene. Finally, DFT calculations were used
to provide evidence for the involvement of the observed preagostic interaction in the
cyclometalation step of the catalytic cycle.
Introduction
catalyst to the most accessible C-H bond or,3 more
commonly, by using heteroatom coordination to direct
the catalyst to a site proximal to the C-H bond to be
functionalized.4
Bedford and co-workers have applied the latter method
in their novel protocol for the ortho arylation of phenols.5
This process involves the use of a Rh catalyst and a
catalytic amount of phosphinite to generate a different,
but more reactive, phosphinite from a phenol. Once
formed, this reactive phosphinite presumably coordi-
nates to the metal catalyst and undergoes cyclometa-
lation, which then allows for selective ortho arylation
by a standard cross-coupling mechanism. Transfer of
the PR2 group to another molecule of phenol releases
Catalytic methods for selective activation and subse-
quent functionalization of carbon-hydrogen (C-H)
bonds provide an atom-economical alternative to con-
ventional procedures using halogenated or metalated
starting materials.1 However, exceptionally demanding
requirements are placed on such systems. A potential
catalyst must activate a relatively inert C-H bond in
the presence of other diverse functional groups, ef-
ficiently functionalize the metalated carbon, and dis-
criminate between the various C-H bonds in a molecule
to provide the desired product.
Despite these stringent requirements, a number of
transformations proceeding via C-H bond activation
have been developed.2 These methods generally rely on
the use of functional-group-tolerant late metals (groups
8-10 in particular) with ligands capable of stabilizing
a reactive metal fragment. The challenge of regioselec-
tivity has been met either by taking advantage of steric
interaction between substrate and catalyst to force the
(3) For examples of methods utilizing steric factors to affect regio-
selectivity of C-H bond functionalization: (a) Chen, H.; Schlecht, S.;
Semple, T. C.; Hartwig, J. F. Science 2000, 287, 1995. (b) Cho, J.-Y.;
Iverson, C. N.; Smith, M. R. J. Am. Chem. Soc. 2000, 122, 12868. (c)
Ishiyama, T.; Takagi, J.; Kousaku, I.; Miyaura, N.; Anastasi, N. R.;
Hartwig, J. F. J. Am. Chem. Soc. 2001, 124, 390.
(4) Two distinct mechanisms for heteroatom-directed C-H func-
tionalization have been noted. One mechanism involves selective
activation of
a C-H bond proximal to a directing group by a
* To whom correspondence should be addressed. E-mail: bergman@
cchem.berkeley.edu (R.G.B.); jellman@uclink.berkeley.edu (J.A.E.).
† Department of Chemistry, University of California.
coordinated metal fragment followed by functionalization of the
metalated carbon: (a) Kakiuchi, F.; Sekine, S.; Tanaka, Y.; Kamatini,
A.; Sonoda, M.; Chatani, N.; Murai, S. Bull. Chem. Soc. Jpn. 1995,
68, 62-83. The second involves reversible C-H activation of many
sites in a molecule by a metal fragment with functionalization
occurring only proximal to a directing group: (b) Lenges, C. P.;
Brookhart, M. J. Am. Chem. Soc. 1999, 121, 6616-6623. For recent
examples of methods utilizing heteroatom directed C-H functional-
ization see: (c) Kametani, Y.; Satoh, T.; Miura, M.; Nomura, M.
Tetrahedron Lett. 2000, 41, 2655-2658. (d) Oi, S.; Fukita, S.; Hirata,
N.; Watanuki, N.; Miyano, S.; Inoue, Y. Org. Lett. 2001, 3, 2579-2581.
(e) Oi, S.; Ogino, Y.; Fukita, S.; Inoue, Y. Org. Lett. 2002, 4, 1783-
1785. (f) Kakiuchi, F.; Murai, S. Acc. Chem. Res. 2002, 35, 826-834.
(g) Tan, K. L.; Park, S.; Ellman, J. A.; Bergman, R. G. J. Org. Chem.
2004, 69, 7329-7335. (h) Thalji, R. K.; Ellman, J. A.; Bergman, R. G.
J. Am. Chem. Soc. 2004, 126, 7192-7193. (i) Lewis, J. C.; Wiedemann,
S. H.; Bergman, R. G.; Ellman, J. A.; Org. Lett. 2004, 6, 35-38.
(5) (a) Bedford, R. B.; Coles, S. J.; Hursthouse, M. B.; Limmert, M.
E. Angew. Chem., Int. Ed. 2003, 42, 112-114. (b) Bedford, R. B.;
Limmert, M. E. J. Org. Chem. 2003, 68, 8669-8682.
‡ Division of Chemical Sciences, Lawrence Berkeley National Labo-
ratory, University of California.
(1) For reviews focusing on C-H bond activation with discussions
on the challenges of functionalization, see: (a) Crabtree, R. H. Chem.
Rev. 1985, 85, 245. (b) Arndsten, B. A.; Bergman, R. G.; Mobley, T. A.;
Peterson, T. H. Acc. Chem. Res. 1995, 28, 154-162. (c) Shilov, A. E.;
Shul’pin, G. B. Chem. Rev. 1997, 97, 2879. (d) Crabtree, R. H. Dalton
Trans. 2001, 2437-2450. (e) Bercaw, J. E.; Labinger, J. A. Nature 2002,
417, 507-514.
(2) For reviews of methods involving catalytic C-H functionalization
see: (a) Dyker, G. Chem. Ber. 1997, 130, 1567-1578. (b) Dyker, G.
Angew. Chem., Int. Ed. 1999, 38, 1698-1712. (c) Miura, M.; Nomura,
M. Top. Curr. Chem. 2002, 219, 212-237. (d) Kakiuchi, F.; Murai, S.
Acc. Chem. Res. 2002, 35, 826-834. (e) Jia, C.; Kitamura, T.; Fujiwara,
Y. Acc. Chem. Res. 2001, 34, 633-634. (f) Guari, Y.; Sabo-Etienne, S.;
Chaudret, B. Eur. J. Inorg. Chem. 1999, 1047-1055. (g) Kakiuchi, F.;
Chatani, N. Adv. Synth. Catal. 2003, 345, 1077-1101. (h) Kakiuchi,
F.; Murai, S. Top. Curr. Chem. 1999, 3, 47-77.
(6) Taken from ref 5b.
10.1021/om050700i CCC: $30.25 © 2005 American Chemical Society
Publication on Web 10/11/2005

5738 Organometallics, Vol. 24, No. 23, 2005
Lewis et al.
Scheme 16
Scheme 2
Table 1. Comparison of NMR Data for Ha in 1a
and 2a-c
1a
2a
2b
2c
δ (ppm)
∆δ (ppm)
splitting
J (Hz)
7.11
10.30
3.19
d
9.51
2.40
d
9.25
2.14
d
-
dd
8.4, 3.7
the desired product and regenerates another 1 equiv of
the reactive phosphinite to complete the catalytic cycle
(Scheme 1). We became interested in using the metalla-
cycle intermediate proposed in this mechanism for
different types of transformations such as hydroaryl-
ation.2d A detailed study of Rh(I) phosphinite complexes
was therefore undertaken in order to better understand
the behavior of these species.
We disclose here the synthesis of three non-cyclo-
metalated rhodium-phosphinite complexes, which were
found to exhibit rare examples of preagostic interactions
between the rhodium centers and the hydrogen ortho
to the oxygen of the phosphinite ligands. Such an
interaction is of interest due to both its possible implica-
tions for the mechanism of C-H activation and func-
tionalization reactions and the general lack of charac-
terization of such complexes involving d8 transition
metals in the literature. A number of experimental and
computational techniques were used to characterize this
interaction. One of the complexes underwent stoichio-
metric C-H bond activation and subsequent alkylation
ortho to the phosphinite directing group. The use of
these systems for catalytic alkylation of phenols in both
an inter- and intramolecular fashion was then investi-
gated. Finally, a mechanism for the C-H activation
reaction occurring during these transformations is
proposed on the basis of density functional theory (DFT)
calculations.
8.2
8.2
8.3
phosphinite ligands, and no Rh-H coupling was ob-
served for any of the complexes. Although these initial
data did not indicate the presence of a formal agostic
interaction, it still seemed likely that the anomalous
chemical shifts of the protons of the phosphinite ligands
were cause by some type of coordination to the nearby
Rh atom. We decided to further investigate this interac-
tion, due to its possible implications for subsequent
reactions of this C-H bond.
Crystals of complexes 2a-c were obtained by vapor
diffusion of pentane into toluene solutions at room
temperature. X-ray analysis provided the dimeric struc-
tures shown in Figures 1-3 for complexes 2a-c,
respectively. The coordination geometry around each
rhodium is essentially square planar with deviations in
the angles of no more than 10° and no unusual Rh-L
or Rh-Cl bond lengths. The fold angle (λ) between the
two square planes comprising the dimers varies signifi-
cantly in 2a-c, as has been noted in many other
[L2RhCl]2 structures.7
Further examination of these structures also revealed
a relatively short interatomic distance between Ha of
the phosphinite ligand and the rhodium center (H15 and
H53, H1 and H1_2, and H24 and H56 in Figures 1-3,
respectively). The Rh‚‚‚Ha distances and the Ca-Ha‚‚‚Rh
angles of the three complexes ranged from 2.77 to 3.01
Å and from 136 to 141°, respectively (Table 3). These
observations supported our original hypothesis regard-
ing the involvement of the Rh center in the aforemen-
tioned downfield chemical shift perturbations of the
ortho protons of the phosphinite ligands of 2a-c;
however, full characterization of this interaction still
needed to be completed.
Results and Discussion
Synthesis and Characterization of Rh-Phos-
phinite Complexes. Complexes 2a-c were prepared
by heating toluene solutions of [RhCl(coe)2]2 (coe )
cyclooctene), phosphinite 1a, and the appropriate ligand,
L (Scheme 2).
A number of metal-hydrogen interactions (X-H‚‚‚M,
High conversions of 1a to the corresponding rhodium
complexes were observed by 1H and 31P NMR spectros-
copy. Vapor diffusion of pentane into toluene solutions
of the reaction mixtures provided analytically pure
products in good yields after filtration (2a, 73%; 2b,
65%; 2c, 72%). Analysis of the 1H NMR spectra of
complexes 2a-c revealed significantly downfield shifted
resonances corresponding to the protons ortho (Ha) to
the phosphinite oxygens (Table 1). A nondecoupled 1-D
HMQC NMR spectroscopic analysis of 1a and 2c
X ) C, N), including agostic interactions,8 preagostic
(7) For discussions of fold angles in [L2RhCl]2 complexes see: (a)
Wang, K.; Goldman, M. E.; Emge, T. J.; Goldman, A. S. J. Organomet.
Chem. 1996, 518, 55. (b) Schnabel, R. C.; Roddick, D. M. Inorg. Chem.
1993, 32, 1513. (c) Curtin, M. D.; Butler, W. M.; Greene, J. Inorg. Chem.
1978, 17, 2928. (d) Ibers, J. A.; Snyder, R. G. Acta Crystallogr. 1962,
15, 923. (e) Dahl, L. F.; Martell, C.; Wampler, D. L. J. Am. Chem. Soc.
1961, 83, 1761.
(8) (a) Brookhart, M.; Green, M. L. H. J. Organomet. Chem. 1983,
250, 395. (b) Brookhart, M.; Green, M. L. H. Prog. Inorg. Chem. 1988,
36, 1. (c) Crabtree, R. H.; Hamilton, D. G. Adv. Organomet. Chem. 1988,
36, 1. (d) Crabtree, R. H. Angew. Chem., Int. Ed. Engl. 1993, 32, 789.
(e) Hall, C.; Perutz, R. N. Chem. Rev. 1996, 96, 3125.
1
revealed only a minor increase in JC-H for the ortho
carbon (Ca) of the free (159.5 Hz) and bound (162.3 Hz)

Preagostic Interactions and C-H Functionalization
Organometallics, Vol. 24, No. 23, 2005 5739
Figure 1. ORTEP drawings of (a) the complete dimer structure of 2a and (b) one of the two Rh centers. Hydrogen atoms
distant from the metal center and solvent atoms have been omitted for clarity. Thermal ellipsoids are shown at the 50%
probability level.
Figure 2. ORTEP drawings of (a) the complete dimer structure of 2b and (b) one of the two Rh centers. Hydrogen atoms
distant from the metal center have been omitted for clarity. Thermal ellipsoids are shown at the 50% probability level.
interactions,9 and hydrogen bonding,10 have been de-
scribed in the literature (vide infra). While the divisions
between these types of interactions are not absolute,
each has signature spectroscopic and geometric proper-
ties which allow them to be generally distinguished from
one another (Table 4) and from simple metal-hydrogen
close contacts.
Agostic interactions are three-center-two-electron
bonds in which the electron density of a C-H bond is
donated to a metal center.8a As a consequence of this
interaction, the C-H bond is typically elongated and
the C-H‚‚‚M angle is relatively acute. Agostic interac-
tions exhibit a number of specific NMR characteristics
such as upfield shifts in C-H bond proton resonances,
M-H coupling for NMR-active metals, and reduction
1
of JC-H values for the carbon of the coordinated C-H
bond; however, none of these were observed for 2a-c.
Linear metal-hydrogen bonds to hydrocarbons, which
involve a filled metal d orbital as a hydrogen bond
acceptor and a C-H bond as a hydrogen bond donor,
represent an extreme in X-H‚‚‚M geometries opposite
to that of agostic interactions.10 As for traditional
hydrogen bonds, a nearly linear arrangement of hydro-
gen bond donor and acceptor is preferred in order to
maximize electrostatic interactions between the hydro-
gen bond donor and acceptor. Metal-hydrogen bonds
have most often been observed with relatively polar
N-H bonds, but there has been some discussion of these
interactions in C-H‚‚‚M systems.10a,b The C-H‚‚‚M
angles present in 2a-c deviate significantly from 180°,
but weak hydrogen bonding cannot be discounted.9e
Interactions between C-H bonds and metal centers that
(9) (a) Roe, D. M.; Bailey, P. M.; Moseley, K.; Maitlis, P. M. J. Chem.
Soc. Chem. Commun. 1972, 1273-1274. (b) Albinati, A.; Anklin, C.
G.; Ganazzoli, F.; Ru¨egg, H.; Pregosin, P. S. Inorg. Chem. 1987, 26,
503-508. (c) Albinati, A.; Arz, C.; Pregosin, P. S. Inorg. Chem. 1987,
26, 508-513. (d) Ablinati, A.; Pregosin, P. S.; Wombacher, F. Inorg.
Chem. 1990, 29, 1812-1817. (e) Yao, W.; Eisenstein, O.; Crabtree, R.
H. Inorg. Chim. Acta 1997, 254, 105.
(10) (a) Brammer, L.; Charnock, J. M.; Goggin, P. L.; Goodfellow,
R. J.; Koetzle, T. F.; Orpen, A. J. J. Chem. Soc., Chem. Commun. 1987,
443. (b) Brammer, L.; Charnock, J. M.; Goggin, P. L.; Goodfellow, R.
J.; Orpen, A. J.; Koetzle, T. F. J. Chem. Soc., Dalton Trans. 1991, 1789.
(c) Brammer, L. McCann, M. C.; Bullock, R. M.; McMullan, R. K.;
Sherwood, P. Organometallics 1992, 11, 2339. (d) Lee, J. C.; Rheingold,
A. L.; Muller, B.; Pregosin, P. S.; Crabtree, R. H. J. Chem. Soc., Chem.
Commun. 1994, 1021. (e) Braga, D.; Grepioni, F.; Tedesco, E. Organo-
metallics 1997, 16, 1846-1856.

5740 Organometallics, Vol. 24, No. 23, 2005
Lewis et al.
Figure 3. ORTEP drawings of (a) the complete dimer structure of 2c and (b) one of the two Rh centers. Hydrogen atoms
distant from the metal center and solvent atoms have been omitted for clarity. Thermal ellipsoids are shown at the 50%
probability level.
Table 2. Crystal Data and Structure Refinement for 2a-c
2a
2b
2c
empirical formula
formula wt
cryst color, habit
cryst dimens (mm)
cryst syst
lattice type
lattice params
a, Å
Rh2Cl2P4O2C64H76‚C3.5H4
1323.99
Rh2Cl2P4O2C34H64
905.49
Rh2Cl2P2O2N2C38H64‚C3.5H4
965.67
yellow, platelike
0.10 × 0.07 × 0.27
monoclinic
red, tabular
0.06 × 0.10 × 0.16
monoclinic
golden, polyhedral
0.14 × 0.25 × 0.29
triclinic
primitive
C centered
primitive
12.404(2)
30.956(4)
16.505(2)
31.419(3)
7.5160(6)
17.707(2)
12.854(1)
13.311(2)
15.473(2)
105.193(2)
95.559(2)
111.160(2)
2328.1(4)
P1h (No. 2)
2
1.377
1002.00
9.24
-144 ( 1
5.00-49.00
49.4
b, Å
c, Å
R, deg
â, deg
90.904(2)
93.679(1)
γ, deg
V (Å3)
6336(1)
P21/c (No. 14)
4
1.388
2740.00
7.48
-133 ( 1
5.00-47.00
49.4
4172.9(6)
C2/c (No. 15)
4
1.441
1872.00
10.98
-113 ( 1
5.00-49.00
49.5
space group
Z value
Dcalcd (g/cm3)
F000
µ(Mo KR) (cm-1
temp (°C)
)
2θ range (deg)
2θmax (deg)
no. of rflns measd
total
27 984
10 738
0.071
0.0300
6428
689
9.33
8998
11 748
7464
unique
3713
Rint
p factor
no. of observns (I > 3.00σ(I))
no. of variables
rfln/param ratio
residuals:a R, Rw, Rall
goodness of fit indicator
0.041
0.045
0.0300
0.0300
5208
464
11.22
0.044, 0.054, 0.069
1.31
2761
199
13.87
0.044, 0.049, 0.091
1.34
0.031, 0.039, 0.043
1.34
2
a R ) ∑||Fo| - |Fc||/∑|Fo| and Rw ) [∑w(|Fo| - |Fc|)2/∑wFo
]
1/2, where w ) 1/σ2(Fo).
have characteristics between the defined extremes of
agostic and hydrogen bonding have been termed remote,
pregostic, or preagostic interchangeably.12 Originally
identified by Pregosin and co-workers, these are typified
by relatively long M‚‚‚H distances and large X-H‚‚‚M
angles.9a-d Their primary spectroscopic signature is a
downfield shift in resonances for protons involved in the
interaction.9e Coupling of this proton to NMR-active
metals as well as changes in 1JC-H are not observed. In
comparison to the defining characteristics of agostic,
hydrogen-bonding, and preagostic interactions, the geo-
metric parameters observed for 2a-c best fit the
definition of complexes containing a preagostic metal-
hydrogen interaction. To gain further insight into this
interaction, and to provide a connection between the
solution and solid-state data used to characterize 2a-
c, a computational study was undertaken.
(11) Data taken from ref 9e.
(12) (a) Crabtree, R. H. Angew. Chem., Int. Ed. Engl. 1993, 32, 789.
(b) Bortolin, M.; Bucher, U.; Reugger, H.; Venanzi, L. M.; Albinati, A.;
Lianza, F. Organometallics 1992, 11, 2514. (c) Cano, M.; Heras, J. V.;
Maeso, M.; Alvaro, M.; Ferna´ndez, R.; Pinilla, E.; Campo, J. A.; Monge,
A. J. Organomet. Chem. 1997, 534, 159. (d) Krumper, J. R.; Gerisch,
M.; Magistrato, A.; Rothlisberger, U.; Bergman, R. G.; Tilley, T. D. J.
Am. Chem. Soc. 2004, 126, 12492.
Computational Studies of the Rh‚‚‚H-C Interac-
tion. The atomic coordinates for 2a were used to build

Preagostic Interactions and C-H Functionalization
Organometallics, Vol. 24, No. 23, 2005 5741
Table 3. Summary of Interatomic Distances (Å)
and Bond Angles (deg) Relevant to Rh‚‚‚Ha
Interactions in 2a-ca
to generate cyclometalated phosphinite species from this
complex, since such intermediates were proposed by
Bedford and coworkers in their studies on the ortho
arylation of phenols using Rh-phosphinite mixtures.13
However, no significant change in the observed chemical
shift for the ortho proton of the phosphinite ligand of
2a was observed by in situ NMR spectroscopy experi-
ments in d8-toluene or d8-THF at room temperature.
Thermolysis of 2a in d8-toluene did not result in reaction
until decomposition occurred at temperatures ranging
from 45 to 135 °C, and attempts to intercept any
intermediate hydridic species as the corresponding
chloride by exchange with CCl4 met with no success.14
Although synthesis of cyclometalated products was
not successful, it was observed that addition of 2 equiv
of PPh3 to a solution of 2a in toluene provided the
bis(phosphine)-phosphinite complex cis-RhCl(PPh3)2-
(i-Pr2POXy) (4a) (Scheme 3). In situ 1H NMR spectros-
copy revealed that the characteristic downfield doublet
for the ortho proton of the phosphinite ligand in 2a had
shifted significantly upfield to 9.07 ppm. The 31P{1H}
NMR spectrum of 4a showed three major signals at
2a
2b
2c
Rh‚‚‚Ha
2.77
3.01
2.98
Rh′‚‚‚Ha′b
Rh‚‚‚Ca
2.84
3.01
2.92
3.56
3.61
141.34
139.25
205.5
3.76
3.76
136.16
136.16
234.9
3.73
3.69
136.01
138.91
69.7
Rh‚‚‚Ca′b
Rh-Ha-Ca
Rh-Ha′-Ca′b
λc
a The non-hydrogen atoms were refined anisotropically; hydro-
gen atoms were included but not refined. Bond length or angle
derived from calculated location of Ha. b The prime symbol refers
to data for the second half of the dimer complex. c Fold angle
between two square planes of the dimeric structure.
Table 4. Comparison of X-H‚‚‚M Interactions11
1
2
153.3 ppm (ddd, JRh-P1a ) 163.4 Hz, JP1a-PPh ,trans
)
3
368.6 Hz, 2JP1a-PPh ,cis ) 36.7 Hz), 49.4 ppm (dt, 1JRh-P,P
3
2
) 195.8 Hz, JP-PPh /1a ) 36.5 Hz), and 27.5 ppm
3
1
2
(ddd, JRh-P1a ) 163.4 Hz, JP1a-PPh ,trans ) 368.6 Hz,
3
2
2
J
) 36.7 Hz). The relatively large JP-P
P1a-PPh3,cis
coupling constant indicated a trans relationship between
Rh-bound 1a and one of the PPh3 ligands, while the
2
smaller JP-P coupling constant was consistent with a
cis relationship between the remaining PPh3, Rh-bound
1a, and the PPh3 ligand trans to 1a.15 It was also found
that heating a d8-toluene solution of 2a and neohexene
provided the alkylated ligand in 17% yield (Scheme 3).16
Addition of 2 equiv of PPh3 or 1a to this reaction
mixture provided the same alkylated product in 35%
and 67% yields, respectively. These results indicated
that a bis(phosphine)-phosphinite or phosphine-
bis(phosphinite) complex might be involved in the
alkylation reaction.
Alkylation of 1a with neohexene was also attempted
in a catalytic fashion, in analogy to the work by Bedford
and coworkers (vide supra).5 It was found that heating
d8-toluene solutions of 2,3-dimethylphenol, neohexene,
and 2a provided the desired products in 31% yield
(Scheme 4; Table 5, entry 1). In analogy to the stoichio-
metric reaction, acceleration of the alkylation in the
presence of PPh3 implicated a bis(phosphine) complex
Figure 4. Superimposed structures of 2a (red), 2b (green),
and 3 (blue). Only the structural features common to all
structures are included, and portions of the aryl rings are
omitted for clarity.
the model compound [RhCl(PMe3)(Me2POXy)]2 (3), which
was geometry optimized in Gaussian03 using DFT
calculations at the mPW1PW91/SDD level of theory.
The calculated structure (dRh‚‚‚H ) 2.792, 2.805 Å;
C-H‚‚‚Rh ) 137.89, 137.02°; λ (fold angle, vide supra) )
231.3°) agreed well with that of 2a (dRh‚‚‚H ) 2.765, 2.842
Å; C-H‚‚‚Rh ) 141.34, 139.25°; λ ) 205.5°) and 2b (dRh‚
‚‚ ) 2.984, 2.919 Å;
H
) 136.01, 138.91°; λ )
C-H‚‚‚Rh
234.9°) (Figure 4). The calculated bond distances of 3
were within 5% of those in 2a,b, except for those of the
P-O bonds, which were slightly elongated (7-10%) in
the calculated structure.
(13) (a) Parshall, G. W. Acc. Chem. Res. 1970, 3, 139. (b) Dehand,
J.; Pfeffer, M. Coord. Chem. Rev. 1976, 18, 327. (b) Bruce, M. I. Angew.
Chem., Int. Ed. 1977, 16, 73. (c) Omae, I. Chem. Rev. 1979, 79, 287.
(d) Rothwell, I. P. Acc. Chem. Res. 1988, 21, 153. (e) Newkome, G. R.;
Puckett, W. E.; Gupta, V. K.; Kiefer, G. E.; Chem. Rev. 1986, 86, 451.
(14) Crabtree, R. H. Metal Alkyls, Aryls, and Hydrides and Related
σ-bonded Ligands. In The Organometallic Chemistry of the Transition
Metals, 3rd ed.; Wiley: New York, 2001; p 70.
Importantly, 1H NMR chemical shielding calculations
indicated that the preagostic, ortho proton of 3 should
have a chemical shift of 9.68 ppm. This value agrees
well with the chemical shifts observed experimentally
for 2a,b at 10.30 and 9.51 ppm, respectively. This result
provides further support for the presence of a preagostic
interaction in both the solid and solution state by
providing a link between the crystal structures and
spectroscopic data for 2a-c. Having extensively char-
acterized the preagostic interactions in 2a-c, we next
desired to study the reactivity of these complexes.
Chemistry of Rh-Phosphinite Complexes. Initial
investigations of the reactivity of 2a focused on attempts
(15) Similarly large 2JP-P coupling has been observed in rhodium(I)
complexes with trans phosphinites: (a) Feiken, N.; Pregosin, P.;
Trabesinger, B. Organometallics 1998, 17, 4510-4518. (b) Irvine, D.
J.; Cole-Hamilton, D. J.; Barnes, J.; Hodgson, P. K. G. Polyhedron 1989,
8, 1575-1577.
(16) Stoichiometric alkylation of cyclometalated Ru-phosphinite
complexes has been accomplished: Parshall, G. W.; Knoth, W. H.;
Schunn, R. A. J. Am. Chem. Soc. 1969, 91, 4990-4995. Catalytic
alkenylation of phenols has been accomplished: Nishinaka, Y.; Satoh,
T.; Miura, M.; Morisaka, H.; Nomura, M.; Matsui, H.; Yamaguchi, C.
Bull. Chem. Soc. Japn. 2001, 74, 1727-1735. Catalytic alkylation of
phenols has been accomplished using norbornene (Dorta, R.; Togni,
A. Chem. Commun. 2003, 760-761) and ethylene (Lewis, L. N.; Smith,
J. F. J. Am. Chem. Soc. 1986, 108, 2728-2735).

5742 Organometallics, Vol. 24, No. 23, 2005
Lewis et al.
Scheme 3
Scheme 4
Scheme 5
ation. Specifically, 2,3-dimethylphenol, neohexene,
Wilkinson’s catalyst, and 1b were refluxed at 135 °C
for 6 h to provide the alkylated phenol in 72% isolated
yield.
Table 5. Effects of Added Phosphine/Phosphinite
Additives on Alkylation of 1aa
An intramolecular variant of the olefin coupling was
also explored (Scheme 5). In this case, the substrate,
Wilkinson’s catalyst, and 1b were refluxed in toluene
to provide quantitative conversion to the desired product
(80% isolated yield). A number of substrates bearing
tethers of varying olefin substitution, length, and hetero-
atom constitution were submitted to these reaction
conditions, but either olefin isomerization or no reaction
occurred in each case. Despite the limited substrate
scope of these reactions, the aforementioned successes
represent a useful extension of Bedford’s arylation
method for alkylation of phenols.
Proposed Reaction Pathway. We finally wished to
establish a link between the preagostic interaction
observed in the catalysts (2a and 4a) and the oxidative
addition step of the catalytic cycle for the alkylation
reaction. The importance of inter- and intramolecular
C-H bond activation by metal complexes has led to a
number of investigations into the individual intermedi-
ates and dynamics of these transformations.17 In par-
ticular, Crabtree and co-workers studied a collection of
compounds exhibiting a range of C-H‚‚‚M interactions
in order to model the approach of C-H bonds to a metal
center using the static structure method of Bu¨rgi and
Dunitz.18 Their analysis focused on the effective covalent
radius for C-H bonding electrons (rbp) and provided
evidence that C-H oxidative addition begins with a
weak interaction between a C-H bond and a metal
center (rbp > 1.0 Å) and proceeds through geometries
typical of agostic intermediates (0.7 Å < rbp < 1.0 Å) to
provide the final product (rbp < 0.7 Å) (Figure 5).
alkylation (%)
amt of PR3
reacn
R′
PR3
(equiv)
2.8 h 5.5 h 24 h
1
2
3
4
5
H
H
H
0
22
59
88
95
92
26
69
31
83
98
PPh3
i-Pr2POXy
0.1
0.1
0
96
P-i-Pr2
P-i-Pr2 PPh3
100
92
0.1
a Yields determined by 1H NMR spectroscopy relative to
dimethoxytoluene internal standard.
as the active catalyst (entry 2). Further rate acceleration
was observed using a phosphinite cocatalyst, as shown
in Table 5 (entry 3), and nearly quantitive alkylation
was obtained under these conditions. No reaction was
observed in the absence of phosphinite ligand, and no
increases in conversion were observed upon addition of
triethylamine (to facilitate phosphinite catalyst turn-
over) to the reaction mixture. Alkylation occurred
exclusively at the ortho position of the phenol. It was
also found that heating a d8-toluene solution of 1a and
a catalytic amount of 2a under 1 atm of D2 led to 75%
incorporation of D at the ortho position and no measur-
able incorporation of D at the meta or para position of
2,3-dimethylphenol following acidic workup of the reac-
tion mixture. Presumably, the rate acceleration associ-
ated with the addition of phosphinite ligand in the
catalytic reaction, as well as the increased conversion
observed with the addition of phosphinite ligand in the
stoichiometric reaction, indicated that the phosphinite
served both as a ligand and as a substrate for alkylation.
While 2a is an efficient catalyst for the ortho alkyl-
ation of 2,3-dimethylphenol, this complex is not ideal
for the alkylation of other phenols, due to the generation
of alkylated 1a from the initial turnover of 2a. Further-
more, 2a is not commercially available, which limits
possible application of this chemistry. We therefore
established a simple alkylation protocol utilizing Wilkin-
son’s catalyst as an inexpensive, commercially available
catalyst precursor and (2,6-xylyl)diisopropylphosphinite
(1b) as a phosphinite ligand. Phosphinite 1b, which has
its ortho positions blocked with methyl groups, allowed
use of this species as a cocatalyst for the ortho alkyl-
Bedford and co-workers proposed that C-H bond
activation of a phosphinite generated in situ during
their method for ortho arylation of phenols occurs to give
a metallacycle intermediate. Cyclometalation of phos-
phinite ligands by a variety of metals has been ob-
Figure 5. General coordinate for oxidative addition of
C-H bonds.

Preagostic Interactions and C-H Functionalization
Organometallics, Vol. 24, No. 23, 2005 5743
Scheme 6. Proposed Reaction Pathway for Cyclometalation
Table 6. Selected Energies, Distances, and Angles
served,19 but it is well-known that cyclometalated
intermediates are not necessarily isolable, despite ap-
parently favorable geometric and electronic factors.13
Although no cyclometalated products were observed in
our alkylation studies or in the arylation chemistry
developed by Bedford, both investigations provided
strong evidence for the intermediacy of a cyclometalated
species produced by C-H bond activation of the phos-
phinite ligand. The correlations that we observed be-
tween preagostic interactions that were identified for
complexes 2a-c and the catalytic activity of 2a in the
ortho alkylation reaction of 2,3-dimethylphenol prompted
our investigation into the mechanism for the cyclo-
metalation of phosphinite ligands using Crabtree’s
analysis of C-H bond oxidative addition as a guide for
the geometries of intermediates along the proposed
pathway.
The observed formation of cis-RhCl(PPh3)2(i-Pr2POXy)
from both 2a/PPh3 and Wilkinson’s catalyst/1a, and the
increased catalytic activity of these complexes relative
to 2a alone, implicated this species as the resting state
for the alkylation catalyst.20 We therefore envisioned
formation of a bis(phosphine) intermediate, 4a, from
dimer fragmentation and addition of PPh3 to 2a (Scheme
6). In catalytic reactions involving Wilkinson’s catalysts,
this step corresponds to a ligand exchange of PPh3 for
1a. Displacement of phosphine by the incoming C-H
bond could give rise to the agostic intermediate 4c,
which following C-H activation would give the product
4e.
DFT calculations (B3LYP, LACVP**++) were used
to model the reaction coordinate for the C-H activation
pathway proposed in Scheme 6 using Me2POPh and
PMe3 in place of 1a and PPh3, respectively. All inter-
mediates were confirmed by frequency calculations (all
positive frequencies observed for intermediates and a
single imaginary frequency observed for transition
states), and the vibration associated with each imagi-
nary frequency was observed to be consistent with the
proposed transformation. The pathways for both cis- and
trans-bis(phosphine) complexes were modeled in order
to identify any major differences in the reactivity of the
two isomers.
for cis- and trans-5a-e
∆G
dC-H dRh-C dRh-H C-H‚‚‚Rh
rbp
(Å)
complex (kcal/mol)
(Å)
(Å)
(Å)
(deg)
cis-5a
0
1.085 3.859 3.027
1.153 2.393 1.847
1.174 2.274 1.765
1.532 2.102 1.595
2.318 2.054 1.517
1.087 3.641 2.770
1.106 2.709 2.077
1.114 2.502 2.000
1.664 2.103 1.558
2.508 2.082 1.513
133.87
103.4
99.4
1.9
cis-5b
34.1
13.2
19.7
10.7
-3.2
18.6
8.1
0.6
0.5
cis-5c
cis-5d
84.5
0.3
cis-5e
60.6
0.0
trans-5a
trans-5b
trans-5c
trans-5d
trans-5e
137.0
113.1
103.2
81.4
1.6
0.9
0.7
20.1
14.6
0.2
56.0
-0.1
analogous to those of complex 4a (Table 6). The Rh‚‚‚H
bond lengths, C-H‚‚‚M angles, and effective covalent
radius (rbp, vide supra) value (1.9) of complex 5a are
consistent with a preagostic interaction. While cis-4a
was observed by NMR spectroscopy, trans-5a was
calculated to be more stable than cis-5a by 3.2 kcal/mol.
This small discrepancy could be a result of the simpli-
fications made to 4a present in 5a. Transition state 5b
involves the loss of a phosphine ligand and was mini-
mized next. A number of attempts to minimize penta-
coordinate bis(phosphine) structures provided either 5a
or 5c, presumably due to the steric congestion about the
Rh center for these complexes. The calculated energetic
barrier leading from the trans isomer of 5a to transition
state 5b is 21.8 kcal/mol; it is 34.1 kcal/mol for the cis
isomer. These values are consistent with experimental
and calculated values for phosphine dissociation from
late-metal complexes.21 A relatively large difference in
energy (15.5 kcal/mol) between the cis and trans com-
plexes (5b) was calculated, indicating that the cyclo-
metalation may actually proceed through the trans
pathway, but additional experimental evidence is needed
to confirm this. The Rh-C/H bond distances, C-H‚‚‚Rh
angles, and rbp values (0.5 and 0.7 Å) calculated for cis-
and trans-5c agreed well with literature examples of
agostic interactions.
Next, the oxidative addition transition state 5d was
located. The barrier for C-H activation for the trans
complex was calculated to be 12.0 kcal/mol and only 6.5
kcal/mol in the cis complex, which is consistent with
literature reports of low energetic barriers for cyclo-
metalation of ligands on d8 metal-phosphine complexes.
Finally, the cyclometalated structures, 5e, were mini-
mized and found to be about 15 kcal/mol higher in
energy than the preagostic complexes 4a; this energy
difference may explain our inability to observe or isolate
any cyclometalated products. These calculations are
consistent with the involvement of the preagostic in-
teraction observed in 2a along the proposed path for
cyclometalation of this complex (Figure 6).
Calculations commenced with a geometry optimiza-
tion of complex 5a, which has structural parameters
(17) For computational studies see: (a) Saillard, J.-Y.; Hoffmann,
R. J. Am. Chem. Soc. 1984, 106, 2006. (b) Cundari, T. R. J. Am. Chem.
Soc. 1994, 116, 340. (c) Koga, N.; Morokuma, K. Chem. Rev. 1991, 91,
823. For experimental studies, see: (d) ref i. (e) Ryabov, A. D. Chem.
Rev. 1990, 90, 403.
(18) Crabtree, R. H.; Holt, E. M.; Lavin, M.; Morehouse, S. M. Inorg.
Chem. 1985, 24, 1986.
(19) Bedford, R. B.; Hazelwood, S. L.; Horton, P. N.; Hursthouse,
M. B. Dalton Trans. 2003, 4164.
(20) When PPh3 was replaced with 1a or 1b as both the ligand and
the substrate, a bis(phosphinite) species, RhCl(PPh3)(i-Pr2POXy)2,
could also serve as the active catalyst; however, PMe3 should still serve
as a useful model for computational studies.
(21) Schmid, R.; Herrmann, W. A.; Frenking, G. Organometallics
1997, 16, 701.

5744 Organometallics, Vol. 24, No. 23, 2005
Lewis et al.
Figure 6. Calculated reaction pathway for cyclometalation.
Materials. Unless otherwise noted, all reagents were
obtained from commercial suppliers and used without further
purification. Molecular sieves (3 Å) were heated at 300 °C
under 0.5 mmHg vacuum overnight to remove any traces of
water. Toluene, THF, and pentane used in the preparation of
1a,b and 2a-c and in the catalytic hydroarylation reactions
were passed through columns of activated alumina (type A2,
size 12 × 32, Purify Co.) under nitrogen pressure, sparged with
N2, and stored over molecular sieves prior to use. THF-d8 was
stirred under nitrogen over sodium/benzophenone ketyl,
vacuum-transferred into a sealable storage vessel, degassed
using three consecutive freeze-pump-thaw cycles, and stored
over activated 3 Å molecular sieves. Toluene-d8 was trans-
ferred into a sealable storage vessel, degassed using three
consecutive freeze-pump-thaw cycles, and stored over acti-
vated 3 Å molecular sieves.
Phosphinite 1a. In an inert-atmosphere box, 2,3-dimethyl-
phenol (1.80 mmol, 0.259 g) and 4 mL of THF were added to
a glass Kontes vial. NaH (1.98 mmol, 0.0475 g) was added in
small portions to the solution. The reaction mixture was stirred
for 5 min, and chlorodiisopropylphosphine (1.80 mmol, 0.290
mL) was slowly added. The vial was capped, removed from
the inert-atmosphere box, and heated to 100 °C for 14 h. The
reaction vessel was returned to the inert-atmosphere box, and
the reaction mixture was filtered through basic alumina, with
hexanes as eluent. Solvents were evaporated to give 0.35 g of
1a as a colorless oil (82%). 1H NMR (400 MHz, CDCl3): δ 7.11
(dd, J ) 3.7, 8.4 Hz, 1H), 6.92 (t, J ) 7.9 Hz, 1H), 6.70 (d, J )
7.3 Hz, 1H), 2.22 (s, 3H), 2.15 (s, 3H), 1.92 (m, 2H), 1.11 (m,
12H). 31P NMR (162 MHz, CDCl3): δ 141.45. Anal. Calcd for
C14H23OP: C, 70.56; H, 9.73. Found: C, 70.35; H, 9.96. GC-
MS (m/z (% relative intensity, ion)): 238.30 (100, M+), 223.20
(19, M - Me), 195.20 (42, M - i-Pr), 151.10 (42, M - i-Pr2).
Conclusion
The results presented in this work have supplied
evidence for the involvement of a preagostic interac-
tion in a C-H activation reaction. We have synthesized
three Rh-phosphinite complexes which exhibit a rare
preagostic interaction. We characterized these com-
plexes using crystallographic and spectroscopic tech-
niques as described for previous examples of preagostic
interactions. In addition, we established a connection
between solution and solid-state data obtained for these
structures using DFT calculations to correlate molecular
structure with observed spectroscopic data. We next
showed that complex 2a catalyzes the ortho alkylation
of phosphinites and developed a catalytic method for
ortho alkylation of phenols with an unactivated alkene.
Finally, we used DFT calculations to provide evidence
for the involvement of complexes containing a preagostic
interaction in the reaction coordinate for cyclometala-
tion of phosphinite ligands.
Experimental Section
General Procedures. All reagents were degassed and
handled under an inert nitrogen atmosphere using syringe and
cannula techniques. Unless otherwise noted, all organic prepa-
rations were carried out in flame- or oven-dried glassware and
all organometallic procedures were conducted in a nitrogen-
filled Vacuum Atmospheres inert-atmosphere box. Flash col-
umn chromatography was carried out using Merck 60 230-
400 mesh silica gel. NMR spectra (1H, 13C, and 31P) were
obtained on a Bruker AMX-400 spectrometer at room temper-
ature. Chemical shifts are reported in ppm, and coupling
constants are reported in Hz. 1H resonances are referenced to
residual protonated solvent, while 31P resonances are refer-
enced to a trimethyl phosphate external standard. Mass
spectrometry was performed by the University of California,
Berkeley, mass spectrometry facility or by the aforementioned
GC/MS system. X-ray crystal structures were obtained by Drs.
Frederick J. Hollander and Allan G. Oliver at the University
of California, Berkeley (UCB), X-ray facility (CHEXRAY).
Elemental analyses were performed at the UCB Micro-
analytical facility on a Perkin-Elmer 2400 Series II CHNO/S
analyzer.
Phosphinite 1b. In an inert-atmosphere box, 2,6-dimethyl-
phenol (10.0 mmol, 1.28 g), triethylamine (15.0 mmol, 2.10
mL), and toluene (10 mL) were added to an oven-dried bomb.
Chlorodiisopropylphosphine (10.0 mmol, 1.50 g) was added to
this solution. The bomb was removed from the inert-atmo-
sphere box and heated at 100 °C for 12 h. In the inert-
atmosphere box, the reaction mixture was filtered through
basic alumina, with hexanes as eluent, and solvents were
removed in vacuo to give 2.20 g of 1b as a colorless oil (92%).
1H NMR (400 MHz, CDCl3): δ 6.97 (d, J ) 7.3, 2H), 6.84 (t, J
) 7.5, 1H), 2.35 (s, 6H), 2.00 (m, 2H), 1.14 (m, 12H). 31P NMR
(162 MHz, CDCl3): δ 155.34. Anal. Calcd for C14H23OP: C,

Preagostic Interactions and C-H Functionalization
Organometallics, Vol. 24, No. 23, 2005 5745
70.56; H, 9.73. Found: C, 70.80; H, 9.95. GC-MS (m/z (%
relative intensity, ion)): 238.30 (100, M+), 195.20 (53, M -
i-Pr), 151.10 (56, M - i-Pr2).
EtOAc/hexanes) to give 19.7 g of the desired phosphate as a
colorless oil (92%). 1H NMR (400 MHz, CDCl3): δ 7.33 (t, J )
7.9, 4H), 7.27 (dt, J ) 2.4, 6.8, 1H), 7.20 (m, 6H), 6.94 (d, J )
7.9, 1H), 6.89 (m, 2H), 5.25 (d, J ) 8.6, 2H), 3.76 (s, 3H). 13C
NMR (101 MHz, CDCl3): δ 159.75, 156.21, 153.84, 150.43,
136.64, 129.75, 125.34, 120.09, 114.63, 113.07, 70.49, 55.20.
31P NMR (162 MHz, CDCl3): δ -11.82. HRMS-EI (m/z): [M]+
calcd for C20H19O5P, 370.0970; found, 370.0972.
General Procedure for the Preparation of Phosphin-
ite Complexes 2a-c. In an inert-atmosphere box, the ap-
propriate ligand, L (0.200 mmol), was added to an orange
solution of 1a (0.200 mmol, 0.0477 g) and [RhCl(coe)2]2 (0.100
mmol, 0.0716 g) in 4 mL of toluene in a Kontes vial. The vial
was sealed, removed from the inert-atmosphere box, and
heated at 45 °C for 6 h. The vial was returned to the box, and
the red (2a and 2b) or yellow (2c) solution was concentrated
to give an orange oil. This oil was dissolved in a minimal
volume (ca. 0.5 mL) of toluene, placed in a vial containing ca.
5 mL of pentane, sealed, and allowed to sit at room temper-
ature to give analytically pure solid samples of 2a-c after
filtration and rinsing with pentane.
1-(2,2-Dimethylbut-3-enyl)-3-methoxybenzene. A stir
bar, 100 mL of THF (distilled from Na/benzophenone ketyl
under N2 immediately prior to use), and magnesium (1.17 mol,
28.3 g, ground in a mortar and pestle immediately prior to
use) were placed in an oven-dried, three-necked round-bottom
flask purged with N2. 1-Chloro-3-methyl-2-butene (0.146 mol,
16.4 mL) and 100 mL of THF were added to an addition funnel
attached to the center neck of the reaction flask. The Mg
suspension was cooled to -30 °C, and the alkene solution was
slowly added dropwise. The reaction mixture was warmed to
room temperature and was stirred for 3 h. Diphenyl (3-
methoxybenzyl)phosphonate (0.0600 mol, 22.2 g) and THF (100
mL) were placed in another oven-dried round-bottom flask
equipped with a stir bar under nitrogen. This solution was
cooled to -40 °C, and the Grignard solution (titrated to 0.6 M
immediately before use) was transferred dropwise via cannula.
The reaction mixture was warmed to room temperature and
stirred for 16 h. Saturated NH4Cl was added slowly to the
reaction mixture, and the aqueous layer was extracted with
EtOAc. The combined organic layers were washed with brine
and dried with MgSO4, and the solvents were removed in
vacuo. The residue was purified by column chromatography
(SiO2, 100% hexanes) to give 11.4 g of the desired product as
[RhCl(PPh3)(i-Pr2POXy)]2 (2a). Yield: 0.134 g (73%) of
light red powder. Anal. Calcd for C64H76Cl2O2P4Rh2: C, 60.15;
1
H, 6.01. Found: C, 60.42; H, 6.09. H NMR (d8-toluene, 400
MHz): δ 10.30 (d, J ) 8.2, 2H), 8.01 (m, 10H), 7.1-6.9 (m,
24H), 2.11 (m, 4H), 2.01 (s, 6H), 1.79 (s, 6H), 1.15 (m, 12H),
1.05 (m, 12H). 31P NMR (d8-toluene, 162 MHz): δ 156.34 (dd,
2
1
2
1JRh-P ) 210.2, JP-P ) 44.1), 49.76 (dd, JRh-P ) 203.4, JP-P
) 44.4). A small portion of this material was dissolved in
toluene with gentle heating (ca. 45 °C) and recrystallized by
vapor diffusion of pentane into the toluene solution of the red
powder to give dark orange crystals suitable for X-ray analysis.
[RhCl(PMe3)(i-Pr2POXy)]2 (2b). Yield: 0.119 g (65%) of
red crystals suitable for X-ray analysis. Anal. Calcd for
C34H64Cl2O2P4Rh2: C, 45.09; H, 7.14. Found: C, 44.85; H, 7.25.
Solution NMR spectra showed a mixture (1.7:1) of two isomers.
1H NMR (d8-toluene, 400 MHz): δ 9.51 (m, J ) 8.2 Hz, 2 H),
6.9 (m, 2H), 6.76 (m, 2 H), 2.50 (m, 4 H), 2.13 (m, 12H), 1.81
1
a colorless oil (82.0%). H NMR (400 MHz, CDCl3): δ 7.20 (t,
J ) 7.8 Hz, 1H), 6.75 (m, 3H), 5.90 (dd, J ) 10.8, 17.7 Hz,
1H), 4.93 (m, 2H), 3.82 (s, 3H), 2.59 (s, 2H), 1.04 (s, 6H). 13C
NMR (101 MHz, CDCl3): δ 158.93, 148.11, 140.42, 128.39,
123.16, 116.50, 111.02, 110.50, 55.08, 49.11, 37.63, 26.52.
HRMS-EI (m/z): [M]+ calcd for C13H18O, 190.1358; found,
190.1358.
(m, 7.7H), 1.71 (m, 4.3H), 1.30 (m, 12H), 1.28 (s, 11H), 2.10 (s,
1
7H). 31P NMR (d8-toluene, 162 MHz): δ 171.0 (dd, JRh-P
)
2
1
2
226.1, JP-P ) 50.4), 170.0 (dd, JRh-P ) 225.2, JP-P ) 49.4),
6.73 (dd, 1JRh-P ) 187.6, 2JP-P ) 50.4), 6.36 (dd, 1JRh-P ) 188.6,
2JP-P ) 50.4).
[RhCl(t-BuNC)(i-Pr2POXy)]2 (2c). Yield: 0.067 g (72%)
of yellow crystals suitable for X-ray analysis. 1H NMR (d8-
toluene, 400 MHz): δ 9.25 (d, J ) 8.3, 2H), 7.10 (t, J ) 7.8,
2H), 6.71 (d, J ) 7.5, 2H), 2.52 (m, 2H), 2.15 (s, 6H), 2.10 (s,
6H), 1.66 (d, J ) 7.1, 6H), 1.63 (d, J ) 7.1, 6H), 1.54 (d, J )
7.1, 6H), 1.51 (d, J ) 7.1, 6H), 0.75 (s, 18H). 31P NMR (d8-
toluene, 162 MHz): δ 171.66 ppm (d, 1JRh-P ) 215.1). HRMS-
EI (m/z): [M]+ calcd for C38H64O2N2P2Cl2Rh2, 918.193; found,
918.195.
3-(2,2-Dimethylbut-3-enyl)phenol. A 60% dispersion of
sodium hydride in mineral oil (65.0 mmol, 2.62 g) was placed
in an oven-dried round-bottom flask. The flask was fitted with
a condenser, the apparatus was purged with N2, and 60 mL
of DMF (anhydrous, Acros) was added. The suspension was
cooled to 0 °C with stirring. Ethanethiol (62.5 mmol, 4.6 mL)
was slowly added to the suspension through the condenser to
control the vigorous reaction that occurs upon addition. Once
the reaction subsided, the reaction mixture was heated to 110
°C to give a brown solution. A solution of 1-(2,2-dimethylbut-
3-enyl)-3-methoxybenzene (25.0 mmol, 4.78 g) in 30 mL of
DMF (combined directly into a syringe) was slowly added to
the reaction mixture (through the condenser). An additional
30 mL of DMF was added through the condenser, and the
mixture was heated at 110 °C for 16 h. The mixture was cooled
to room temperature, and water (250 mL) was added. The
layers were separated, and the aqueous layer was extracted
with EtOAc (3 × 50 mL). The combined organic layers were
washed with brine and dried with MgSO4, and the solvents
were removed in vacuo. The resulting oil was purified by
column chromatography (SiO2, 5% EtOAc/hexanes) to give 3.60
g of the desired product as a colorless oil, which solidified after
In Situ Analysis of RhCl(PPh3)2(i-Pr2POXy) (3a). In an
inert-atmosphere box, 2a (0.006 mmol, 0.0071 g), PPh3 (0.012
mmol, 0.0031 g), and 0.6 mL of d8-toluene were added to a J.
Young tube. The tube was sealed, heated to 45 °C, and
monitored by NMR spectroscopy. Three major signals were
observed by 31P NMR spectroscopy. 31P NMR (d8-toluene, 162
1
2
MHz): δ 153.3 (ddd, JRh-P1a ) 163.4, JP1a-PPh ,trans ) 368.6,
3
2JP1a-PPh ,cis ) 36.7), 49.4 (dt, 1JRh-PP ) 195.8, 2JP-PPh /1a ) 36.5),
3
3
27.5 (ddd, 1JRh-P1a ) 163.4, 2JP1a-PPh ,trans ) 368.6, 2JP1a-PPh ,cis
3
3
) 36.7).
Diphenyl (3-Methoxybenzyl)phosphonate. (3-methoxy-
phenyl)methanol (0.0580 mol, 7.22 mL), diisopropylethylamine
(0.0880 mol, 15.0 mL), DMAP (0.00580 mol, 0.709 g), and 160
mL of CH2Cl2 (distilled from CaH2 under N2 immediately prior
to use) were added to an oven-dried round-bottom flask purged
with N2. The solution was cooled to -50 °C, and diphenyl
chlorophosphate (0.0880 mol, 18.2 mL) was added dropwise.
The reaction mixture was warmed to room temperature and
stirred for 22 h. A saturated solution of NH4Cl was added to
the reaction mixture, the layers were separated, and the
aqueous layer was extracted with three portions of CH2Cl2.
The combined organic layers were washed with brine and dried
with Na2SO4, and solvents were evaporated in vacuo. The
residue was purified by column chromatography (SiO2, 1/4
1
refrigeration (81%). H NMR (400 MHz, CDCl3): δ 7.15 (t, J
) 7.9, 1H), 6.72 (m, 2H), 6.67 (m, 1H), 5.88 (dd, J ) 10.7, 15.5,
1H), 5.36 (s, 1H), 4.92 (m, 2H), 2.56 (s, 2H), 1.03 (s, 6H). 13C
NMR (101 MHz, CDCl3): δ 154.63, 148.04, 140.81, 128.69,
123.37, 117.57, 112.90, 110.58, 48.91, 37.61, 26.51. HRMS-EI
(m/z): [M]+ calcd for C12H16O, 176.1201; found, 176.1202.
6,6-Dimethyl-5,6,7,8-tetrahydronaphthalen-1-ol. In an
oven-dried bomb, 3-(2,2-dimethyl-but-3-enyl)phenol (1.08 mmol,
0.190 g), Wilkinson’s catalyst (0.0108 mmol, 0.0102 g), iPr2POXy
(0.108 mmol, 0.0331 g), and toluene (10.8 mL) were added.
The vessel was sealed, removed from the inert-atmosphere box,

5746 Organometallics, Vol. 24, No. 23, 2005
Lewis et al.
and heated at 130 °C for 17 h. The solution was diluted with
toluene (10 mL) and transferred to a round-bottom flask. To
this was added 1 M HCl (10-15 mL), and the mixture was
stirred vigorously for 50 min. The layers were separated, and
the aqueous layer was extracted with EtOAc. The combined
organic layers were washed with brine and dried with MgSO4,
and the solvents were evaporated in vacuo. The residue was
purified by column chromatography (SiO2, 5% EtOAc/hexanes)
standard. 1H NMR (400 MHz, CDCl3): δ 6.96 (m, 1H), 6.60
(d, J ) 7.57, 1H), 6.35 (d, J ) 8.04, 0.25H), 2.28 (s, 3H), 2.17
(s, 3H).
Computational Details. DFT geometry optimization and
1H NMR chemical shielding calculations of model complex 4
were performed using the program Gaussian03 at the
mPW1PW91/SDD level of theory, as described elsewhere.22
DFT geometry optimization and single point calculations for
cis-5a, cis-5c, cis-5e, and trans-5a-e were performed using
the program Jaguar at the B3LYP,LACVP**++ level of theory
as described elsewhere.23 Geometry optimization of cis-5b and
cis-5d was performed at the B3LYP, LACVP++ level of theory,
while the single-point calculations for these structures were
carried out at the B3LYP, LACVP**++ level of theory.
X-ray Structure Determinations. General Consider-
ations. X-ray crystal structures were obtained by Drs.
Fredrick Hollander and Allan Oliver at the UCB X-ray facility
(CHEXRAY). Crystals were mounted on glass fibers using
Paratone N hydrocarbon oil. All measurements were made on
a SMART24 CCD area detector with graphite-monochromated
Mo KR radiation. Data were integrated by the program
SAINT25 and were corrected for Lorentz and polarization
effects. Data were analyzed for agreement and possible
absorption using XPREP.26 The structures were solved by
direct methods27 and expanded using Fourier techniques.27
Crystal data and structure refinement details for 2a-c are
given in Table 2.
1
to give 0.15 g of the desired product as a yellow oil (80%). H
NMR (400 MHz, CDCl3): δ 7.05 (t, J ) 8.1, 1H), 6.71 (d, J )
8.1, 1H), 6.66 (d, J ) 8.1, 1H), 4.75 (s, 1H), 2.70 (t, J ) 6.8,
2H), 2.58 (s, 2H), 1.65 (t, J ) 6.8, 2H), 1.03 (s, 6H). 13C NMR
(101 MHz, CDCl3): δ 153.35, 138.35, 126.27, 122.00, 122.07,
111.82, 43.48, 35.34, 29.18, 27.90, 20.51. HRMS-EI (m/z): [M]+
calcd for C12H16O, 176.1201; found, 176.1201.
6-(3,3-Dimethylbutyl)-2,3-dimethylphenol. In an inert-
atmosphere box, 2,3-dimethylphenol (1.39 mmol, 0.170 g),
i-Pr2POXy (0.139 mmol, 0.0355 g), 3,3-dimethyl-1-butene (6.95
mmol, 0.896 mL), Wilkinson’s catalyst (0.0695 mmol, 0.0643
g), and toluene (13.9 mL) were placed in an oven-dried bomb
equipped with a stir bar. The bomb was sealed, brought out
of the inert-atmosphere box, and heated to 135 °C, and the
reaction mixture was stirred for 17 h. The reaction mixture
was transferred to a 50 mL round-bottom flask with toluene.
An aqueous solution of HCl (1 M, 15-25 mL) was added, and
the solution was stirred vigorously for 1 h. The layers were
separated, and the aqueous layer was extracted with EtOAc.
The organic layers were combined, washed with brine, dried
with MgSO4, and concentrated in vacuo. The residue was
purified by column chromatography (SiO2, 5% EtOAc/hexanes)
to give 0.21 g of the desired product (72%). 1H NMR (400 MHz,
CDCl3): δ 6.88 (d, J ) 7.7, 1H), 6.71 (d, J ) 7.6, 1H), 2.53 (m,
2H), 2.26 (s, 3H), 2.18 (s, 3H), 1.48 (m, 2H), 0.99 (s, 9H).
HRMS-EI (m/z): [M]+ calcd for C14H22O, 206.1671; found,
206.1674.
Typical Procedure for NMR-Tube Reactions. In an
inert-atmosphere box, 2,3-dimethylphenol, complex 2a, and
phosphinite 1b were rinsed into an oven-dried, medium-walled
NMR tube using 2 × 0.2 mL of d8-toluene. Neohexene was
added to the tube via syringe, and an additional 0.2 mL of
d8-toluene was used to rinse the residue from the reagent vials
and neohexene down the NMR tube. The tube was fitted with
a Cajon adapter, removed from the inert-atmosphere box,
attached to a vacuum line, and flame-sealed. The tube was
heated at 135 °C, and the reaction was monitored by NMR
spectroscopy, which showed 88% conversion to 6-(3,3-dimethyl-
butyl)-2,3-dimethylphenol after 2.8 h. This result matched
those obtained using the aforementioned screening conditions
(Table 4, entry 3). Essentially the same procedure was used
for all reactions described in the text, with substitutions for
alternative reagents.
Procedure for Incorporation of D into 1a. In an inert-
atmosphere box, 1a (0.28 mmol, 0.67 g), 2a (0.0070 mmol,
0.0090 g), and toluene (2.8 mL) were added to an oven-dried
Kontes reaction vessel equipped with a Teflon stopper. The
tube was sealed, removed from the box, degassed using three
freeze-pump-thaw cycles, and charged with ca. 0.8 atm of
D2 gas. The tube was again sealed and heated in an oil bath
at 105 °C for 10 h. The reaction mixture was then transferred
to a vial and stirred vigorously with 5 mL of 1 M HCl for 1 h.
The organic layer was separated, and the aqueous layer was
extracted with 2 × 10 mL of Et2O. The combined organics were
washed with 10 mL of brine, dried over Na2SO4, and concen-
trated to give a brown oil. The oil was purified by column
chromatography (SiO2, 10% EtOAc/hexanes) to give 0.022 g
of 2,3-dimethylphenol (63%) as a white solid. 1H NMR showed
75% incorporation of deuterium at the ortho position and no
incorporation of D at any other site in the molecule relative
to the meta methyl group, which was used as an internal
Acknowledgment. This work was supported by
NIH Grant No. GM069559 (to J.A.E.) and by the
Director, Office of Energy Research, Office of Basic
Energy Sciences, Chemical Sciences Division, U. S.
Department of Energy, under Contract No. DE-AC02-
05CH11231 (to R.G.B.). J.W. is funded by a Jean
Dreyfus Boussevain Award. All calculations were con-
ducted on the UCB Molecular Graphics and Computa-
tion Facility computer cluster supported by NSF Grant
No. CHE-0233882. This work could not have been
completed without many helpful discussions with Dr.
Jennifer R. Krumper. Thanks also to Dr. Fredrick J.
Hollander and Dr. Allen G. Oliver at the UCB X-ray
diffraction facility (CHEXRAY) for solving the struc-
tures of 2a-c, Dr. Yong Zhang and Prof. Eric Oldfield
for geometry optimization and calculated NMR chemical
shifts of 4, and Kathy Durkin and Dr. Nate Crawford
for help with DFT calculations for 5a-e.
Supporting Information Available: CIF files giving
X-ray structure data for 2a-c, tables giving Cartesian coor-
dinates for 3, cis-5a-e, and trans-5a-e, and text giving the
complete Gaussian reference. This material is available free
of charge via the Internet at http://pubs.acs.org.
OM050700I
(22) (a) Pople, J. A. et al. Gaussian 03; Gaussian, Inc., Pittsburgh,
PA, 2003. (b) Godbout, N.; Havlin, R.; Salzmann, R.; Debrunner, P.
G.; Oldfield, E. J. Phys. Chem. A 1998, 102, 2342-2350. (c) Zhang, Y.;
Guo, Z.; You, X.-Z. J. Am. Chem. Soc. 2001, 123, 9378-9387.
(23) Tan, K. L.; Bergman, R. G.; Ellman, J. A. J. Am. Chem. Soc.
2002, 124, 3202-3203.
(24) Area-Detector Software Package; Siemens Industrial Automa-
tion, Inc., Madison, WI, 1995.
(25) SAX Area-Detector Integration Program, version 4.024;
Siemens Industrial Automation, Inc., Madison, WI, 1995.
(26) XPREP, version 5.03, part of the SHELXTL Crystal Structure
Determination Package; Siemens Industrial Automation, Inc., Madi-
son, WI, 1995.
(27) Beurskens, P. T.; Admiraal, G.; Beurskens, G.; Bosman, W. P.;
de Gelder, R.; Israel, R.; Smits, J. M. M. The DIRDIF-94 Program
System; Technical Report of the Crystallography Laboratory; Univer-
sity of Nijmegen, Nijmegen, The Netherlands, 1994.