Alkyl- and aryl-oxygen bond activation in solution by rhodium(I), palladium(II), and nickel(II). Transition-metal-based selectivity

Article abstract of DOI:10.1021/ja9738889

Reaction of [RhCl(C₈H₁₄)₂]₂ (C₈H₁₄ = cyclooctene) with 2 equiv of the aryl methyl ether phosphine 1 in C₆D₆ results in an unprecedented metal insertion into the strong sp²-sp³ aryl-O bond. This remarkable reaction proceeds even at room temperature and occurs directly, with no intermediacy of C-H activation or insertion into the adjacent weaker ArO-CH₃ bond. Two new phenoxy complexes (8 and 9), which are analogous to the product of insertion into the ArO-CH₃ bond (had it taken place) were prepared and shown not to be intermediates in the Ar-OCH₃ bond cleavage process. Thus, aryl-O bond activation by the nucleophilic Rh(I) is kinetically preferred over activation of the alkyl-O bond. The phenoxy Rh(I)-η¹-N₂ complex (8) is in equilibrium with the crystallographically characterized Rh(I)-μ-N₂-Rh(I) dimer(12). Reaction of [RhClC₈H₁₄)₂]₂ with 2 equiv of the aryl methyl ether phosphine 2, PPh₃, and excess HSiR₃ (R = OCH₂CH₃, CH₂CH₃) results also in selective metal insertion into the aryl-O bond and formation of (CH₃O)SiR₃. Thus, transfer of a OCH₃ group from carbon to silicon was accomplished, showing that hydrosilation of an unstrained aryl-O single bond by a primary silane is possible. The selectivity of C-O bond activation is markedly dependent on the transition-metal complex and the alkyl group involved, allowing direction of the C-O bond activation process at either the aryl-O or alkyl-O bond. Thus, contrary to the reactivity of the rhodium complex, reaction of NiI₂ or Pd(CF₃CO₂)2 with 1 equiv of 1 in ethanol or C₆D₆ at elevated temperatures results in exclusive activation of the sp³-sp³ ArO-CH₃ bond, while reaction of the analogous aryl ethyl ether 4 and Pd(CF₃CO₂)2 results in both sp³-sp³ and sp²-sp³ C-O bond activation. The resulting phenoxy Pd(II) complex (18) is fully characterized by X-ray analysis. Heating the latter under mild dihydrogen pressure results in hydrodeoxygenation to afford an aryl-Pd(II) complex (19).

Full text of DOI:10.1021/ja9738889

J. Am. Chem. Soc. 1998, 120, 6531-6541
6531
Alkyl- and Aryl-Oxygen Bond Activation in Solution by
Rhodium(I), Palladium(II), and Nickel(II). Transition-Metal-Based
Selectivity
Milko E. van der Boom, Shyh-Yeon Liou, Yehoshoa Ben-David, Linda J. W. Shimon, and
David Milstein*
Contribution from the Department of Organic Chemistry, The Weizmann Institute of Science,
RehoVot 76100 Israel
ReceiVed NoVember 12, 1997
Abstract: Reaction of [RhCl(C8H14)2]2 (C8H14 ) cyclooctene) with 2 equiv of the aryl methyl ether phosphine
1 in C6D6 results in an unprecedented metal insertion into the strong sp2-sp3 aryl-O bond. This remarkable
reaction proceeds even at room temperature and occurs directly, with no intermediacy of C-H activation or
insertion into the adjacent weaker ArO-CH3 bond. Two new phenoxy complexes (8 and 9), which are analogous
to the product of insertion into the ArO-CH3 bond (had it taken place) were prepared and shown not to be
intermediates in the Ar-OCH3 bond cleavage process. Thus, aryl-O bond activation by the nucleophilic
Rh(I) is kinetically preferred over activation of the alkyl-O bond. The phenoxy Rh(I)-η1-N2 complex (8) is
in equilibrium with the crystallographically characterized Rh(I)-µ-N2-Rh(I) dimer (12). Reaction of [RhCl-
(C8H14)2]2 with 2 equiv of the aryl methyl ether phosphine 2, PPh3, and excess HSiR3 (R ) OCH2CH3, CH2-
CH3) results also in selective metal insertion into the aryl-O bond and formation of (CH3O)SiR3. Thus,
transfer of a OCH3 group from carbon to silicon was accomplished, showing that hydrosilation of an unstrained
aryl-O single bond by a primary silane is possible. The selectivity of C-O bond activation is markedly
dependent on the transition-metal complex and the alkyl group involved, allowing direction of the C-O bond
activation process at either the aryl-O or alkyl-O bond. Thus, contrary to the reactivity of the rhodium
complex, reaction of NiI2 or Pd(CF3CO2)2 with 1 equiv of 1 in ethanol or C6D6 at elevated temperatures
results in exclusive activation of the sp3-sp3 ArO-CH3 bond, while reaction of the analogous aryl ethyl ether
4 and Pd(CF3CO2)2 results in both sp3-sp3 and sp2-sp3 C-O bond activation. The resulting phenoxy Pd(II)
complex (18) is fully characterized by X-ray analysis. Heating the latter under mild dihydrogen pressure
results in hydrodeoxygenation to afford an aryl-Pd(II) complex (19).
Introduction
systems or relatively weak C-O bonds, or systems driven by
aromatization, are well-known.16-21 For example, C-O bond
activation of strained cyclic ethers by transition metals has been
applied to catalysis of isomerization to carbonyl compounds,
coupling to form esters, and carbonylation to lactones.16,19,22
Only a few examples of CH3-O bond cleavage in aryl alkyl
ethers have been reported with soluble metal complexes.7-9,23
Activation of the sp3-sp3 C-O bond results in the formation
of phenoxy metal complexes (Scheme 1). For instance, reaction
of anisole with Fe(0) complexes results in activation of the
CH3-O bond,7,8 while the stronger aryl-O bond remains intact.
Activation processes of unstrained C-O single bonds by
metal complexes are proposed as key steps in the hydrodeoxy-
genation (HDO) of crude oil and may lead to the design of novel
catalytic reactions.1-4 Cleavage of C-O bonds is applied in
the manufacture of drugs, pharmaceuticals, and other fine
chemicals. Direct activation of unstrained C-O bonds by metal
complexes in solution is rare, and mechanistic information
regarding these processes is scarce.5-15 C-O bond cleavage
by transition-metal complexes which involve either strained
(1) Caldwell, T. E.; Abdelrehim, I. M.; Land, D. P. J. Am. Chem. Soc.
1996, 118, 907-908.
(2) Furimsky, E. Catal. ReV. Sci. Eng. 1983, 25, 421-458.
(3) Gary, J. H. H., G. E. Petroleum Refining: Technology and Economics,
3rd ed.; Marcel Dekker: New York, 1993.
(4) Speight, J. G. The Chemistry and Technology of Petroleum; Marcel:
New York, 1983.
(13) Van der Boom, M. E.; Liou, S.-Y.; Ben-David, Y.; Vigalok, A.;
Milstein, D. Angew. Chem., Int. Ed. Engl. 1997, 36, 625-626.
(14) Grotjahn, D. B.; Joubran, C. Organometallics 1995, 14, 5171.
(15) Jang, S.; Atagi, L. M.; Mayer, J. M. J. Am. Chem. Soc. 1990, 112,
6413-6414.
(16) Milstein, D. Acc. Chem. Res. 1984, 17, 221-226.
(17) Zlota, A. A.; Frolow, F.; Milstein, D. J. Am. Chem. Soc. 1990, 112,
6411-6413.
(5) Bonanno, J. B.; Henry, T. P.; Neithamer, D. R.; Wolczanski, P. T.;
Lobkovsky, E. B. J. Am. Chem. Soc. 1996, 118, 5132-5133.
(6) Moulton, C. J.; Shaw, B. L. J. Chem. Soc. 1976, 1020-1024.
(7) Tolman, C. A.; Ittel, S. D.; English, A. D.; Jesson, J. P. J. Am. Chem.
Soc. 1979, 101, 1742-1751.
(18) Aye, K.-T.; Colpitts, D.; Ferguson, G.; Puddephatt, R. J. Organo-
metallics 1988, 7, 1454-1456.
(19) Khumtaveeporn, K.; Alper, H. Acc. Chem. Res. 1995, 28, 414-
422.
(8) Ittel, S. D.; Tolman, C. A.; English, A. D.; Jesson, J. P. J. Am. Chem.
Soc. 1978, 100, 7577-7586.
(20) Chaudret, B. Bull. Soc. Chim. Fr. 1995, 132, 268-279.
(21) Rondon, D.; Chaudret, B.; He, X. D.; Labroue, D. J. Am. Chem.
Soc. 1991, 113, 5671-5676.
(22) Milstein, D. J. Am. Chem. Soc. 1982, 104, 5227-5228.
(23) Poignant, G.; Nlate, S.; Guerchais, V.; Edwards, A. J.; Raithby, P.
R. Organometallics 1997, 16, 124-132.
(9) Dunbar, K. R.; Haefner, S. C.; Uzelmeier, C. E.; Howard, A. Inorg.
Chim. Acta 1995, 240, 524-527.
(10) Bryndza, H. E.; Tam, W. Chem. ReV. 1988, 88, 1163-1188.
(11) Yamamoto, A. AdV. Organomet. Chem. 1992, 34, 111-147.
(12) Grotjahn, D. B.; Lo, H. C. Organometallics 1996, 15, 2860-2862.
S0002-7863(97)03888-2 CCC: $15.00 © 1998 American Chemical Society
Published on Web 06/19/1998

6532 J. Am. Chem. Soc., Vol. 120, No. 26, 1998
Van der Boom et al.
Scheme 1
Scheme 2
Metal insertion into aryl-O bonds is unprecedented. Such a
process is expected to be followed by facile â-hydrogen
elimination of the generated alkoxy metal complexes.24,25 In
recent years, we reported the activation of unstrained aryl-C
bonds by Rh(I), Ir(I), and Pt(II) and studied the mechanism
involved.26-32 Cleavage of an aryl-Si bond by Pt(II)33,34 and
insertion of a Ta(III) complex into an aryl-N bond were
published recently.5
We report here the first examples of metal insertion into the
strong aryl-O bond of an aryl alkyl ether under mild conditions
in solution. These unique processes occur directly, with no
intermediacy of C-H activation or insertion into an adjacent
weaker alkyl-O bond. The alkoxy group can be transferred
to primary silanes, providing the first example of hydrosilation
of an unstrained C-O single bond. Moreover, we demonstrate
that the C-O bond activation process can be directed at either
an aryl-O or alkyl-O bond depending on the metal complex
employed and the alkyl group. Thus, reaction of the methyl
phenyl ether 1 with Rh(I) results in direct Ar-O activation,
while with Pd(II) or Ni(II) exclusive activation of the sp3-sp3
ArO-CH3 bond takes place. On the other hand, reaction of
the analogous ethyl phenyl ether 4 and Pd(II) results in both
sp3-sp3 and sp2-sp3 C-O bond cleavage. We report also an
example of hydrodeoxygenation (HDO) of an alkoxy metal
complex. Part of this study has been communicated previ-
ously.13
2,6-dimethylphenol via the Williamson reaction,36 followed by
bromination and phosphine exchange.6 Compounds 1-4 were
obtained in good yields as white powders and were fully
characterized by 1H, 1H{31P}, 31P{1H}, 13C{1H}, and 13C-DEPT-
135 NMR, MS, and elemental analysis.
Ar-O Activation by Rhodium(I). Reaction of the complex
[RhCl(C8H14)2]2 (C8H14 ) cyclooctene) with 2 equiv of 1 in
C6D6 at 85 °C for 3 h (in a sealed vessel) resulted in quantitative
formation of the known Rh(III) hydride complex 5, which was
unambiguously identified by various NMR techniques and by
comparison to an authentic sample (Scheme 2).6,37
This unprecedented Ar-O bond activation, proceeds even
at room temperature (20% conversion into 5 was observed by
31P{1H} NMR after 24 h). Former studies in our group have
shown that coordination of the di-tert-butyl phosphine to Rh
and Ir alkene dimers controls the overall rate in the activation
of an Ar-C bond.30 Likewise, the substitution of the olefin by
1 is probably slow relative to Ar-O bond activation. Using
deuterated solvents, no Rh(III)-D formation was observed by
2H NMR, indicating that the solvents do not contribute to the
Rh-H formed. The presumably formed (unobserved) Rh(III)-
OCH3 intermediate A undergoes readily â-hydrogen elimination
to afford complex 5 and formaldehyde.24 Very recently, we
prepared the analogous Rh(III)-CH2CH3 complex 6, which upon
heating gives complex 5 and ethylene.38
Although we have not directly detected formation of form-
aldehyde in the Ar-O bond activation reaction (1 f 5; Scheme
2), heating the orange product solution at 140 °C or performing
the reaction at this temperature leads also to formation of
complex 7,6 presumably by decarbonylation of the initially
formed formaldehyde. In support of this, heating of complex
5 at 140 °C in a sealed vessel with paraformaldehyde (10 equiv)
results in decarbonylation to yield complex 7 quantitatively. The
Results and Discussion
Preparation of Substrates. To probe the possibility of
activation of unstrained C-O single bonds, we prepared the
new phosphine aromatic ethers 1, 2, 4, and the phenol 3.
Phosphine coordination is expected to direct a metal to the
vicinity of both aryl and alkyl ether bonds. The preparation of
substrates 1-4 is similar to the synthesis of PCP-based ligands.6
Our synthetic route for 1 and 2 consists of reaction of HP(t-
Bu)2 or LiPPh2 with R,R′-dibromo-2-methoxy-m-xylene, which
is readily obtained by bromination of 2-methoxy-m-xylene.
Alkylation of p-cresol resulted in the formation of oxyuvitin
alcohol;35 halide and phosphine exchange afforded the new
diphosphine 3.6 The ethyl phenyl ether 4 was prepared from
(24) Milstein, D. J. Am. Chem. Soc. 1986, 108, 3525-3526.
(25) Blum, O.; Milstein, D. J. Am. Chem. Soc. 1995, 117, 4582.
(26) Gozin, M.; Weisman, A.; Ben-David, Y.; Milstein, D. Nature 1993,
364, 699-701.
(27) Gozin, M.; Aizenberg, M.; Liou, S. Y.; Weisman, A.; Ben-David,
Y.; Milstein, D. Nature 1994, 370, 42-44.
(28) Liou, S.-Y.; Gozin, M.; Milstein, D. J. Am. Chem. Soc. 1995, 117,
9774-9775.
(29) Van der Boom, M. E.; Kraatz, H.-B.; Ben-David, Y.; Milstein, D.
Chem. Commun. 1996, 2167-2168.
(30) Rybtchinski, B.; Vigalok, A.; Ben-David, Y.; Milstein, D. J. Am.
Chem. Soc. 1996, 118, 12406-12415.
(31) Gandelman, M.; Vigalok, A.; Shimon, L. J. W.; Milstein, D.
Organometallics 1997, 16, 3981-3986.
(32) Liou, S.-Y.; Gozin, M.; Milstein, D. J. Chem. Soc., Chem. Commun.
1995, 1965-1966.
(33) Valk, J.-M.; van Belzen, R.; Boersma, J.; Spek, A.; van Koten, G.
J. Chem. Soc., Dalton Trans. 1994, 2293-2302.
(36) Stoochnoff, B. A.; Benoiton, L. Tetrahedron Lett. 1973, 21-24.
(37) Nemeh, S.; Jensen, C.; Binamira-Sorigia, E.; Kaska, W. C. Orga-
nometallics 1983, 2, 1442-1447.
(38) Van der Boom, M. E.; Ben-David, Y.; Milstein, D. Manuscript in
preparation.
(34) Steenwinkel, P.; James, S. L.; Grove, D. M.; Kooijman, H.; Spek,
A. L.; van Koten, G. Organometallics 1997, 5133-515.
(35) Ullmann, F.; Brittner, K. Chem. Ber. 1909, 42, 2539-2548.

Transition-Metal-Based SelectiVity
J. Am. Chem. Soc., Vol. 120, No. 26, 1998 6533
Scheme 3
1
1
yellow complex 7 exhibits identical H, H{31P}, and 31P{1H}
NMR and IR spectra to those reported in the literature.6
Aldehyde decarbonylation by rhodium is well-known.39 Prod-
ucts resulting from metal insertion into the adjacent ArO-CH3
bond, if it occurs at all, are not observed.
The phenoxy complexes 8 and 9 (Scheme 3) were prepared
in order to evaluate whether activation of the weaker CH3-O
bond (compare bond dissociation energy (BDE) values of Ph-
OCH3 ) 91 kcal/mol, PhO-CH3 ) 80 kcal/mol)40,41 precedes
the observed aryl-O bond activation step (1 f 5; Scheme 2),
or perhaps it is a reversible parallel process. Complex 9 is the
iodide analogue of the expected product of Rh(I) insertion into
the ArO-CH3 bond (10; Scheme 4). Deprotonation of the
aromatic phosphine alcohol 3 with KH or NaH in THF, followed
by reaction with [RhCl(C8H14)2]2 (0.5 equiv) in THF under a
nitrogen atmosphere at room temperature leads to quantitave
formation of complex 8 as proven spectroscopically by 1H, 1H-
{31P}, and 31P{1H} NMR, IR, and FD-MS. The latter exhibits
the molecular ion (M+ 554) having the expected isotope pattern.
The 31P{1H} NMR shows a characteristic resonance at δ 91.4
Scheme 4
1
(d, JRhP ) 163.1 Hz), indicating that both phosphorus atoms
are magnetically equivalent and coordinated to the Rh(I) center.
The frequency of the NtN stretch in the IR (ν ) 2095 cm-1
)
is very close to those observed for RhCl(PCy3)2-η1-N2 (Cy )
cyclohexyl)42 and (PCP)RhN2 complexes,43,44 and it is indicative
for an “end-on” coordinated dinitrogen molecule.
Interestingly, no THF or cyclooctene coordination is observed
1
in the H NMR. N2 competes favorably with the solvent and
the olefin, probably as a result of the bulky tert-butyl substituents
on the phosphorus atoms. η1-N2 binding to a sterically crowded
T-shaped Rh(I) center competes favorably with coordination
of CO2 and even with ethylene.44
Scheme 5
Oxidative addition of 13CH3I (1 equiv) to complex 8 in C6D6
at 80 °C in a sealed tube leads within 10 min to quantitative
formation of complex 9, which was characterized by various
NMR techniques and FD-MS (Scheme 3). The 13C{1H} NMR
shows clearly the presence of the Rh(III)-13CH3 moiety, which
appears as a doublet of triplets at δ 10.0 (1JRhC ) 27.4 Hz and
2JPC ) 2.7 Hz), and in the 13C-DEPT-135 NMR, a positive
signal is observed indicative of an odd number of protons. The
FD-MS shows the molecular ion (M+ 668) and a correct isotope
pattern. Complex 9 exhibits similar spectroscopic features to
analogous phenoxy Rh(III)(Me)(Cl), Rh(III)(H)(Cl), and Ir(III)-
(H)(Cl) complexes.45 No intermediate products were detected
1
1
by monitoring this reaction at room temperature by H, H-
{31P}, 13C{1H}, and 31P{1H} NMR. Continued heating of 8
and 9 at 100 °C results eventually in the formation of
unidentified products, while 11 or other products indicative of
aryl-O bond cleavage are not observed. Importantly, com-
plexes 8 and 9 are stable under conditions in which 1 readily
undergoes aryl-O bond cleavage with [RhCl(C8H14)2]2. This
constitutes strong evidence that the observed sp2-sp3 C-O bond
activation by Rh(I) is a direct process (1 f 5; Scheme 4).
Thus, cleavage of the ArO-CH3 bond is not involved either
on the reaction coordinate or as a side equilibrium, even though
this bond is substantially weaker than the adjacent aryl-O bond.
Direct sp2-sp3 aryl-C oxidative addition at room temperature
to Rh(I) and Ir(I) was reported recently.30,31
(39) See, for example: Doughty, D. A.; Pignolet, L. H. J. Am. Chem.
Soc. 1978, 100, 7083-7085.
(40) Handbook of Chemistry Physics, 57th ed.; CRC Press: Boca Raton,
FL, 1976-1977.
(41) Gordon, A. J.; Ford, A. The Chemist’s Companion; Wiley: New
York, 1972.
(42) Van Gaal, H. L. M.; Van den Bekerom, F. L. A. J. Organomet.
Chem. 1977, 134, 237-248.
(43) Vigalok, A.; Kraatz, H.-B.; Konstantinovsky, L.; Milstein, D. Chem.
Eur. J. 1997, 3, 253-260.
(44) Vigalok, A.; Ben-David, Y.; Milstein, D. Organometallics 1996,
15, 1839-1844.
(45) Milstein, D. Unpublished results.

6534 J. Am. Chem. Soc., Vol. 120, No. 26, 1998
Van der Boom et al.
Table 1. Crystal Data for Complexes 12 and 18
compd
formula
fw
12
18
1/2 (C50H90N2P4O2Rh2)
540.7
C26H43P2F3O3Pd
628.9
space group
cryst syst
a, Å
b, Å
c, Å
R, deg
â, deg
γ, deg
V, Å3
Z
C2/c (No. 15)
monoclinic
25.575(5)
18.219(4)
11.656(2)
90
103.99(3)
90
5270(2)
P1h (No. 2)
triclinic
10.714(2)
15.988(3)
19.420(4)
68.96(3)
74.09(3)
70.47(3)
2880.9(10)
4
1.450
0.799
110
0.1 × 0.1 × 0.05
8
D
calcd, g cm-3
1.362
0.786
110
0.25 × 0.25 × 0.3
µ(Mo KR), mm-1
T (K)
Figure 1. ORTEP view of complex 12. Selected bond lengths (Å):
Rh(1)-N(2) ) 1.912(2); Rh(1)-O(1) ) 2.041(2); Rh(1)-P(2) )
2.3847(8); Rh(1)-P(3) ) 2.3512(7); N(2)-N(2a) ) 1.129(4); C(1)-
O(1) ) 1.343(3); Rh(1)‚‚‚C(1) ) 2.610(2). Selected bond angles
(deg): N(2)-Rh(1)-O(1) ) 175.88(8); P(2)-Rh(1)-P(3) ) 155.64(2);
N(2)-Rh(1)-P(2) ) 103.12(7); N(2)-Rh(1)-P(3) ) 98.52(7); O(1)-
Rh(1)-P(2) ) 79.41(6); O(1)-Rh(1)-P(3) ) 79.68(6); N(2)-Rh(1)-
C(1) ) 152.85(8); Rh(1)-N(2)-N(2a) ) 175.25(7); C(1)-O(1)-Rh(1)
) 98.85(14); O(1)-Rh(1)‚‚‚C(1) ) 30.56(7); O(1)-C(1)-C(2) )
119.9(2); O(1)-C(1)-C(6) ) 119.7(2).
cryst size, mm3
and comparable with those found in Rh(I) complexes with
terminally bound dinitrogen ligands.43,44,48-50 It is close to that
of molecular nitrogen (1.0968 Å) and almost identical to N2
+
(1.116 Å; bond order 21/2) suggesting small π-delocalization in
the collinear Rh(I)sNtNsRh(I) moeity compared to other
µ-N2-bimetallic complexes.51 Complexes 12 and 14 have an
almost linear XsRhsNtNsRhsX linkage (X ) O or H) and
two trans bulky phosphorus groups with comparable cone angles
(e.g., compare cone angles θ, EtP(tBu)2 ) 162° vs PiPr3 )
160°).52,53 The Rh(1)-O(1) and Rh(1)-N(2) distances of
2.041(2) and 1.912(2) Å, respectively, are normal. The nitrogen
and oxygen atoms are mutually trans with an almost linear
N(2)-Rh(1)-O(1) angle of 175.88(8)°. Least-squares plane
analysis through the atoms Rh(1), P(2), P(3), O(1), and N(2)
shows that the mean deviation from planarity is 0.1322 Å. The
C(1)-O(1) distance (1.343(3) Å) is comparable with the Ph-
OH distance of phenol (1.362 Å), and thus, it is hardly affected
by binding of the oxygen atom to the metal center.54 The C(1)-
O(1)-Rh(1) angle of 98.85(14)° is indicative of a sp3-hybridized
oxygen atom, which fits well with the observed C(1)-O(1) and
Rh(1)-O(1) distances. The P(2)-Rh(1)-P(3) angle of 155.64-
(2)° is similar to those recently observed in the structures of
two monomeric benzylic PCP-based Rh(I) and Rh(III) com-
plexes.26,28 The expected steric hindrance between the two
rhodium moieties is significantly reduced by bending of the
bulky phosphorus groups, as is observed for 14, allowing the
formation of complex 12. However, there is still considerable
steric hindrance between the tert-butyl groups as indicated by
the relatively short intermolecular distances of C(26)‚‚‚C(24a)
) 3.691 Å and C(36)‚‚‚C(36a) ) 3.308 Å. The rhodium-
phosphorus distances (2.3512(7) and 2.3847(8) Å) are similar
to those observed in other rhodium complexes.26,28,30,37,55 The
ipso carbon of the aromatic ring is close to the metal center
(C(1)‚‚‚Rh(1) ) 2.610(2) Å), but no additional evidence for
any bonding interaction is observed. The aromaticity is not
distorted as indicated by the normal sp2-sp2 C-C distances
(range 1.391(4)-1.401(3) Å)54 and the ring planarity (mean
deviation from plane ) 0.0196 Å).
Orange prismatic crystals of complex 12 were obtained upon
slow concentration of a THF solution of 8 under a nitrogen
atmosphere at room temperature (Scheme 5). 31P{1H} NMR
1
of the THF solution after precipitation of 12 (δ 88.2, JRhP
)
168.0 Hz and 4JRhP ) 24.5 Hz) shows the presence of complexes
8 and 12 in an approximately 4:1 ratio. Thus, 12 crystallizes
more readily than the monomeric analogue (8). Performing the
synthesis of complex 8 in a concentrated solution or bubbling
argon for 3 h through a benzene solution of 8 at room
temperature results also in the formation of complex 12 (30%
yield by 31P{1H} NMR). This indicates that the monomeric
dinitrogen precursor (8) is in equilibrium with a 14-electron
species (13), which might be stabilized by the bulky, basic
phosphine ligands. Subsequentially recombination of 8 and 13
affords 12 (Scheme 5). Reversible dissociation of N2 to form
unsaturated, three-coordinated Rh(I) complexes has been ob-
served.43,44,46 Stabilization of reaction intermediates with N2
in PCP-based Rh(I) complexes was recently reported.43
X-ray Crystal Structure Analysis of Complex 12.
A
single-crystal analysis was performed, clearly showing that one
dinitrogen molecule linearly bridges two Rh(I) atoms and
indirectly supporting the structures of precursors 8 and 13. An
ORTEP view of the molecular structure of 12 and the adopted
numbering scheme is shown in Figure 1. Table 1 gives details
of the crystal structure determination. Complex 12 is a
crystallographic dimer which packs with 1/2 a molecule per
asymmetric unit. An X-ray diffraction study of a dinuclear Rh-
(I)-µ-N2 complex 14 was reported.47
Methoxy Group Transfer to Silanes Promoted by Rh(I).
The new ligand 2 was prepared in order to evaluate the role of
(48) Allen, A. D.; Bottomley, F. Acc. Chem. Res. 1968, 1, 360-365.
(49) Chatt, J.; Leigh, G. J. Q. ReV. Chem. Soc. 1972, 1, 121-144.
(50) Leigh, G. Acc. Chem. Res. 1992, 25, 177-181.
The N(2)-N(2a) distance (1.129(4) Å) is only slightly shorter
than that reported for [RhH(PiPr3)2]2(µ-N2) 14 (1.134(5) Å)47
(51) Hidai, M.; Mizobe, Y. Chem. ReV. 1995, 95, 1115-1133.
(52) Brown, T. L. Inorg. Chem. 1992, 31, 1286-1294.
(53) Tolman, C. A. Chem. ReV. 1977, 77, 313-348.
(46) Van Gaal, H. L. M.; Moers, F. G.; Steggerda, J. J. J. Organomet.
Chem. 1974, 65, C43-C45.
(47) Yoshida, T.; Okana, T.; Thorn, D. L.; Tulip, T. H.; Otsuka, S.; Ibers,
J. A. J. Organomet. Chem. 1979, 181, 183.
(54) Mathematical, Physical and Chemical Tables; Kluwer: Dordrecht,
The Netherlands, 1992; Vol. C.

Transition-Metal-Based SelectiVity
J. Am. Chem. Soc., Vol. 120, No. 26, 1998 6535
Scheme 6
characterized spectroscopically and by comparison to an au-
thentic sample. PPh3 was used to stabilize the organometallic
product 16. The demethoxylated arene 17 can be readily
released from complex 16 by mild heating under H2 (20 psi) in
the presence of excess PMe3.27,60
Mechanistically, the first step is probably coordination of 2
by both phosphine arms to afford 15, followed by oxidative
addition of the H-Si bond to afford intermediate B. Formation
of the strong Cl-Si bond would yield a Rh(I)-H species. A
second oxidative addition of Si-H followed by H2 elimination
and by aryl-O bond cleavage may afford the unobserved Rh-
(III)(OMe)(SiR3) complex (C; the silane analogue of A).
Formation of ClSiEt3 was detected by GC-MS analysis of the
product solution, showing the molecular ion and the expected
fragmentation pattern.58 Intermediate C may undergo facile
irreversible O-Si reductive elimination to afford the observed
CH3OSiR3 and complex 16 from which the organic product (17)
can be liberated.27,60
the subsituents on the phosphorus atoms on the C-O bond
activation process. Reaction of RhCl(PPh3)3 with 1 equiv of 2
in C6D6 at room temperature for 0.5 h suggests the formation
1
of 15 as indicated by H and 31P{1H} NMR, although charac-
1
terization is not conclusive (eq 1). The H NMR shows free
Regardless of the exact mechanism inVolVed, unprecedented
transfer of an alkoxy group from carbon to silicon was obserVed,
showing that functionalization of an unstrained aryl-O single
bond by a silane is possible. We are unaware of other examples
of hydrosilation of an unstrained ether. Moreover, the reaction
is selective toward the stronger Ar-O bond. Formation of the
very strong O-Si single bond provides a large contribution to
the driving force of this reaction. No â-hydrogen elimination
products resulting from the presumed Rh(III)-OCH3 intermedi-
ate (C) were observed. The selective transfer of the OCH3 group
to the silane fits well with our observations that aryl-O bond
activation with 1 and Rh(I) is a direct process (vide supra;
Schemes 3-5). Products indicative of sp3-sp3 C-O bond
activation were not formed. MeSiR3 was not detected either
cyclooctene and broad signals for coordinated 2; the singlet
resonance of the OCH3 group (δ 3.61) is broadened but appears
in almost the same chemical shift as in the free ligand (δ 3.58),
showing that the C-O bonds are not preactivated by the metal
center in the coordinated ligand. The 31P{1H} NMR shows two
broad signals at δ 49.0 and 40.5 (ratio 2:1), consistent with the
proposed structure 15. Similar features have been observed
previously for related eight-membered PCP-based Pt(II) com-
plexes.56
1
by H and 29Si NMR or by GC-MS analysis of the reaction
Performing the same reaction with 2 and various Rh(I)
precursors such as HRh(PPh3)4 or PhRh(PPh3)3 at elevated
temperatures, or thermolysis of the postulated 15 in benzene at
120 °C in a sealed vessel, afforded mixtures of unknown
products. No products indicative of aryl-O or alkyl-O bond
mixture, and only traces of methane (<3%) were detected by
GC analysis of the gas phase, ruling out a consecutive C-O
bond activation process or a side equilibrium involving sp3-
sp3 C-O bond cleavage. In both systems 1 and 2, exclusive
aryl-O bond cleavage with Rh(I) was observed, showing that
this unusual bond activation process can take place with
significantly different electron density and bulk at the metal
center.
Mechanistically, following coordination of the chelating
diphosphines (1 and 2) coordination of the OCH3 group to the
metal center might take place,61-64 although we did not see any
evidence for it. The Rh(I) insertion into the aryl-O bond may
proceed via an η2-arene complex E (Scheme 7), followed by
1
activation were observed either by H or 31P{1H} NMR.
Significantly, reacting [RhCl(C8H14)2]2 with 2 equiv (each)
of 2 and PPh3 in the presence of a 7-fold excess of HSiR3 (R3
) (OEt)3 or Et3) for 0.5 h at 130 °C in benzene or dioxane (in
a pressure bottle) results in the formation of the previously
reported aryl-Rh(I) complex 16 as a major organometallic
product (∼95%) and (MeO)SiR3 (Scheme 6).32 RhCl(PPh3)3
can be used as well. Formation of the methoxysilanes is
unambiguous based on 29Si NMR and GC-MS (CI and EI)
analysis of the product solution, showing the molecular peaks
and expected fragmentation patterns.57-59 Complex 16 was
(59) Scholl, R. L.; Maciel, G. E.; Musker, W. K. J. Am. Chem. Soc.
1972, 94, 6376-6385.
(60) Liou, S.-Y. Ph.D. Thesis, Weizmann Institute of Science, Rehovot,
Israel, 1995.
(55) Weisman, A.; Gozin, M.; Kraatz, H.-B.; Milstein, D. Inorg. Chem.
1996, 35, 1792-1797.
(56) Van der Boom, M. E.; Gozin, M. E.; Ben-David, Y.; Shimon, L. J.
W.; Frolow, F.; Kraatz, H.-B.; Milstein, D. Inorg. Chem. 1996, 35, 7068-
7073.
(61) Linder, E.; Wang, Q.; Mayer, H. A.; Fawzi, R.; Steimann, M.
Organometallics 1993, 12, 1865-1870.
(62) Miller, E. M.; Shaw, B. L. J. Chem. Soc., Dalton Trans. 1974, 480-
485.
(57) Kintzinger, J.-P.; Marsmann, H. Oxygen-17 and Silicon-29;
Springer-Verlag: New York, 1981; Vol. 17.
(58) The Wiley/NBS Registry of Mass Spectra Data; McLafferty, F. W.,
Stauffer, D. B., Eds.; John Wiley & Sons: New York, 1989; Vol. 1.
(63) Empsall, H. D.; Hyde, E. M.; Jones, C. E.; Shaw, B. L. J. Chem.
Soc., Dalton Trans. 1974, 1980-1985.
(64) Linder, E.; Gierling, K.; Keppeler, B.; Mayer, H. A. Organometallics
1997, 16, 3531-3535.

6536 J. Am. Chem. Soc., Vol. 120, No. 26, 1998
Van der Boom et al.
Scheme 7
Scheme 9
19, which is the expected product of aryl-O bond cleavage, is
neither observed nor formed upon treatment of 18 with a 10-
fold excess of CF3CO2CH3 at 85 °C, suggesting that the sp3-
sp3 C-O bond cleavage is thermodynamically preferred in this
system. The thermally stable complex 19 is readily obtained
by heating the disphosphine 20 with Pd(CF3CO2)2 in toluene
at 130 °C for 1 h in a sealed vessel and exhibits similar
spectroscopic features in 1H, 1H{31P}, 31P{1H}, and 13C-DEPT-
135 NMR as its chloride, iodide, and hydride analogues.6,65
Likewise, the reaction of 1 with 1 equiv of NiI2 in ethanol
for 12 h at 120 °C in a closed vessel affords complex 21 (R )
H) and presumably CH3I (Scheme 9). No products of aryl-O
Scheme 8
1
bond activation were observed by H and 31P{1H} NMR. An
analogous phenoxy complex (22; R ) CH3) was obtained by
reacting 3 with NiI2 under the same reaction conditions. Both
Ni(II) complexes, which were characterized by various NMR
techniques and by FD-MS, exhibit very similar spectroscopic
properties. The 1H NMR spectra of complexes 21 and 22 show
that the resulting geometry render the four hydrogen atoms of
the CH2P groups (AB quartet) and the tert-butyl substituents
magnetically nonequivalent. The 31P{1H} NMR exhibits a
singlet resonance, indicating that the two phosphorus atoms are
identical.
Thus, in striking contrast to the reactivity of 1 with Rh(I),
alkyl-O bond activation takes place with Pd(CF3CO2)2 or NiI2.
Few other examples of sp3-sp3 C-O bond cleavage by
transition metals were reported, including metal insertion into
unstrained ArO-CH3 bonds.7-9,23,63,66,67 It was suggested that
CH3-O bond cleavage with Pt(II) might occur via a polar four-
centered transition state.66,67
1,2-migration of the methoxy group from carbon to the metal
center, or via a concerted, tri-centered transition state F, as
observed for Rh(I) and Ir(I) insertion into an Ar-C bond in an
analogous system.30 The kinetic preference leading exclusively
to the aryl-O bond activation might indeed be a result of E. In
the C-C activation process, competitive C-H activation is
observed.
ArO-CH3 Activation by Palladium(II) and Nickel(II).
Interestingly, it is possible to direct the bond activation process
toward the CH3-O bond by employing Pd(II) or Ni(II). Thus,
upon reaction of Pd(CF3CO2)2 with a stoichiometric amount of
1 in C6D6 at 85 °C for 3 h (in a sealed vessel), exclusive
activation of the sp3-sp3 C-O bond took place, leading to
quantitative formation of the phenoxy Pd(II) complex 18
(Scheme 8). The reaction proceeds even at room temperature,
resulting in 40% conversion to 18 after 24 h. No intermediates
were observed, compatible with rate-determining phosphine
coordination. No other products were formed even at higher
temperatures (130 °C). Complex 18, which was isolated and
Aromatic ethers, such as anisole (Scheme 1), undergo
exclusive sp3-sp3 C-O cleavage upon reaction with HI or HBr
at elevated temperatures to give an alkyl halide and an aromatic
alcohol (eq 2).68 Lewis acids such as AlCl3 or BF3 are also
effective. The reaction proceeds by protonation of (or Lewis
acid binding to) the oxygen followed by an external nucleophilic
attack of the anion on the alkyl group (SN2 mechanism). It is
likely that the electrophilic Pd(II) and Ni(II) centers react as
Lewis acids and coordinate to the ether oxygen, forming an
intermediate such as G.69-71 Such an interaction could promote
the loss of the alkyl group by an internal interaction with a
1
1
fully characterized by H, H{31P}, and 31P{1H} NMR, FD-
MS, and X-ray analysis, exhibits similar spectroscopic features
to two analogous benzylic and phenoxy Pd(II) complexes (23;
vide infra).65 CF3CO2CH3 was also formed, as detected by 1H
NMR by comparison with an added authentic sample. Complex
(66) Jones, C. E.; Shaw, B. L.; Turtle, B. L. J. Chem. Soc., Dalton Trans
1974, 992-999.
(67) VanCheesan, S.; Kuriacose, J. C. J. Sci. Ind. Res. 1983, 42, 132.
(68) Bhatt, M. V.; Kulkarni, S. U. Synthesis 1983, 249-282.
(69) Kataoka, Y.; Tsuji, Y.; Matsumoto, O.; Ohashi, M.; Yamagata, T.;
Tani, K. J. Chem. Soc., Chem. Commun. 1995, 2099-2100.
(65) Ohff, M.; Ohff, A.; Van der Boom, M. E.; Milstein, D. J. Am. Chem.
Soc. 1997, 119, 11687-11688.

Transition-Metal-Based SelectiVity
J. Am. Chem. Soc., Vol. 120, No. 26, 1998 6537
coordinated anion, as suggested by Shaw,66 or by an external
nucleophilic attack affording 18 or 21 and an alkyl halide.
X-ray Crystal Structure of 18. The structure of complex
18 was determined by single-crystal X-ray diffraction. Orange
crystals (plates) were obtained upon slow concentration of a
benzene solution of 18 under nitrogen at room temperature. An
ORTEP view of 18 and the adopted numbering scheme is shown
in Figure 2. Table 1 gives details of the crystal structure
determination. The overall features are similar to the phenoxy
Rh(I) complex 12 and will be discussed only briefly. The Pd-
(II) center has a distorted square planar coordination geometry.
Least-squares planar analysis through the atoms Pd(1), P(3),
P(4), O(1), and O(5) shows a mean deviation from planarity of
0.1573 Å. The bulky phosphines P(2) and P(3) as well as the
phenoxy group O(11) and the trifluoracetate anion O(4) are in
a trans arrangement with a rather acute angle of P(2)-Pd(1)-
P(3) (157.06(10)°) and an almost linear angle of O(11)-Pd-
O(10) (176.6(3)°). The Pd(1)-P(2) (2.354 Å), Pd(1)-P(3)
(2.360 Å), Pd(1)-O(1) (1.981 Å), and Pd(1)-O(4) (2.057 Å)
distances are in the range normally found for such bonds.54 The
second oxygen of the CF3CO2 anion is relatively close to the
metal center (Pd(1)‚‚‚O(52) ) 3.014 Å). The sp2-sp3 C-O
bond length (1.344 Å) is not weakened by bonding of O(1) to
the metal center. The intramolecular Pd(1)‚‚‚C(1) distance is
relatively short (2.629 Å) as observed for 12 (Rh(1)‚‚‚C(1) )
2.610 Å), but as in the latter case, there is no spectroscopic
evidence for any metal-carbon interaction. The C(11)-O(1)-
Pd(1) angle (102.8(5)°) is a little larger compared to the one
observed in 12 (98.85°) and indicates an approximate tetrahedral
geometry around O(1).
Figure 2. ORTEP view of complex 18, showing that the Pd(II) atom
has selectively inserted into the sp3-sp3 C-O bond, with overall
retention of the metal oxidation state. Selected bond lengths (Å):
Pd(1)-O(1) ) 2.041(2); Pd(1)-O(1) ) 2.041(2); Pd(1)-P(2) )
2.3847(8); Pd(1)-P(3) ) 2.3512(7); Pd(1)‚‚‚C(1) ) 2.610(2); C(1)-
O(1) ) 1.343(3). Selected bond angles (deg): O(1)-Pd(1)-O(5) )
175.88(8); P(2)-Pd(1)-P(3) ) 155.64(2); O(2)-Pd(1)-P(2) )
103.12(7); O(2)-Pd(1)-P(3) ) 98.52(7); O(1)-Pd(1)-P(2) ) 79.41(6);
O(1)-Pd(1)-P(3) ) 79.68(6); C(1)-O(1)-Pd(1) ) 98.85(14); O(1)-
Pd(1)‚‚‚C(1) ) 30.56(7).
Scheme 10
Competitive sp3-sp3 and sp2-sp3 C-O Bond Activation
with Palladium(II). Reaction of the aryl ethyl ether 4 with a
stoichiometric amount of Pd(CF3CO2)2 in toluene for 3 h at
130 °C (in a closed vessel) led to the formation of complexes
18 and 19 in a ratio of 9:1, respectively (Scheme 10).
Quantitave formation of CF3CO2Et (based on 18) was observed
by 1H, 13C{1H}, and 13C-DEPT-135 NMR and GC-MS analysis
of the reaction mixture.58,72 Heating the yellow reaction solution
overnight at 130 °C did not change the product distribution,
suggesting that complexes 18 and 19 are formed independently
and do not intercovert. Thus, both C-O bonds are readily
accessible to the metal center, and the oVerall C-O bond
actiVation processes are probably kinetically controlled. A
rough estimation of the oVerall relative rates is kalkyl-O/karyl-O
≈ 9.
competitive sp2-sp3 C-O bond cleavage, while with Rh(I) the
opposite is true. Interestingly, the balance between aryl-O and
alkyl-O is not only readily influenced by the metal but also
by the nature of the alkyl group. This is in line with our
postulated mechanism for the sp3-sp3 CH3-O bond activation
with compound 1 and Pd(CF3CO2)2 or NiI2, which probably
involves a SN2 attack on the methyl group by the anion (G;
-
vide supra). Apparently a nucleophilic attack by a CF3CO2
anion on the R carbon of the ethyl moiety of 4 is less favorable
than on the methyl group, mainly due to steric hindrance,
rendering the Ar-O activation to afford 19 more competitive.
This latter process can proceed by a concerted oxidative addition
process or indirectly by formation of an arenium intermediate.73
Ar-O Hydrodeoxygenation (HDO) by Palladium(II).
Reaction of the diphosphine 3 with a stoichiometric amount of
Pd(CF3CO2)2 in toluene at room temperature led to the quantita-
tive formation of the phenoxy palladium(II) complex 23 and
CF3CO2H by selective cleavage of the O-H bond (eq 3).
Prolonged stirring for 5 days or performing this reaction at
various temperatures resulted in the same product. The
thermally stable complex 23 exhibits similar spectroscopic
These observations indicate that there is only a small
difference in the kinetic barriers of sp3-sp3 and sp2-sp3 C-O
bond cleavage. The sp3-sp3 C-O bond activation process with
M(II) (M ) Pd, Ni) is kinetically more favorable than the
(70) Beer, P. D.; Drew, M. G. B.; Leeson, P. B.; Ogden, M. I. J. Chem.
Soc., Dalton Trans. 1995, 1273.
(71) Steffey, B. D.; Miedaner, A.; Maciejewski-Farmer, M. L.; Bernatis,
P. R.; Herring, A. M.; Allured, V. S.; Carperos, V.; DuBois, D. L.
Organometallics 1994, 13, 4844-4855.
1
(72) The Aldrich Library of 13C and H FT NMR Spetra; Pouchert, C.
J., Behnke, J., Eds.; Aldrich Chemical Co., Inc.: Milwaukee, WI, 1993,
1993; Vol. 1.
(73) Terheijden, J.; Van Koten, G.; Vinke, I. C.; Spek, A. L. J. Am.
Chem. Soc. 1985, 107, 2891-2898.

6538 J. Am. Chem. Soc., Vol. 120, No. 26, 1998
Van der Boom et al.
Moreover, the alkoxy group can be transferred to silanes (eq
5), providing a rare example of hydrosilation of an unstrained
C-O bond.
features as the isostructural Pd(II) and Ni(II) complexes 18, 21,
and 22 and was also characterized by preliminary X-ray analysis.
Complex 24, which is the expected product of aryl-O bond
cleavage, is neither observed by 1H or 31P{1H} NMR nor formed
upon heating the product solution up to 180 °C for 24 h in a
sealed vessel. This suggests that the ArO-H bond activation
generating CF3CO2H is thermodynamically more favorable than
the competing Ar-OH bond activation process.
Remarkably, heating the orange product solution or the
isolated thermally stable complexes 18 or 23 (lacking the
p-methyl group on the aromatic ring) at 180 °C in toluene under
H2 (30-35 psi) resulted, within 8 h, in the formation of the
colorless aryl-Pd(II) complexes 19 or 24 as the only organo-
metallic products in approximately 55% yield (by 31P{1H}
NMR; eq 4).
It is possible to influence the selectivity of the kinetically
controlled sp2-sp3 vs sp3-sp3 C-O bond activation by proper
choice of metal precursor or the alkyl group. Our observations
clearly show that a nucleophilic metal favors direct aryl-O bond
activation, while electrophilic metals are likely to promote
alkyl-O bond cleavage. Although in all cases studied a d8
metal is used, the nucleophilic Rh(I) complex selectively
activates the very strong aryl-O bond, whereas the more
electrophilic Pd(II) or Ni(II) complexes preferentially cleave the
alkyl-O single bond.
Significantly, the phenoxy complexes are clearly neither
intermediates in the sp2-sp3 C-O bond activation processes
nor involved in side equilibria. Unprecedented hydrodeoxy-
genation of an aryl-O bond by Pd(II) has been observed. Our
results provide evidence that metal complexes might be designed
to selectively activate and functionalize unstrained C-O single
bonds under mild homogeneous conditions. The bond activation
processes proceed even at room temperature. We believe that
the observed C-O bond cleavage, which is based on model
systems, indicates that the development of a general strategy
for activation of unstrained C-O single bonds under mild
conditions in solution may be possible. Noteworthy, methylene
transfer from benzylic metal complexes to H-C, H-Cl, H-Si,
and even to Si-Si bonds was communicated by us.27,29 The
formation of isostructral phenoxy complexes might lead to
selective oxygen transfer to various substrates, which is a subject
of further studies.
The reaction proceeded also at lower temperatures at longer
reaction times (e.g., 16 h at 140 °C). Formation of 19 and 24
was clearly observed by 1H, 1H{31P}, and 31P{1H}, 13C-DEPT-
135 NMR and FD-MS. The spectroscopic properties are fully
in accord with a C2 symmetry. The structure of 19 was
confirmed by comparison with an authentic sample (vide supra).
Complex 24 shows very similar spectroscopic properties. For
Experimental Section
General Procedures. All reactions were carried out under nitrogen
in a Vacuum Atmospheres glovebox (DC-882) equipped with a
recirculation (MO-40) “Dri Train” or under argon using standard
Schlenk techniques. Oxygen levels (e2 ppm) were monitored with
Et2Zn (1 M solution in hexane, Aldrich), water levels (e2 ppm) were
monitored with TiCl4 (neat, BDH chemicals). Solvents were reagent
grade or better, dried, distilled, and degassed before introduction into
the glovebox, where it was stored over activated 4 Å molecular sieves.
Deuterated solvents were purchased from Aldrich and were degassed
and stored over 4 Å activated molecular sieves in the glovebox. [RhCl-
(C8H14)2]2 was prepared by a published procedure.75 NiI2 and Pd(CF3-
CO2)2 were obtained from Aldrich. Reaction flasks were washed with
dionized water, followed by acetone, and then oven dried prior to use.
GC analyses were performed on a Varian 3300 gas chromatograph
equipped with a molecular sieve column. Field desorption (FD) mass
spectra were measured at the Institute of Mass Spectrometry, University
of Amsterdam, The Netherlands. Elemental analyses were carried out
at the Hebrew University, Jerusalem, Israel.
1
instance, the well-resolved H NMR spectrum of 24 clearly
shows the characteristic tert-butyl and CH2P 1:2:1 triplets at δ
1.19 (JPH ) 6.9 Hz) and δ 2.81 (JPH ) 4.0 Hz), which collapse
1
into singlets in the H{31P} NMR.
Hydrodeoxygenation of a strong aryl-O bond by a transition
metal in solution is unprecedented. Kubiak reported deoxy-
genation of phenols by Pt(II) under mild CO pressure.74 Further
studies are underway to explore the mechanism of this unique
transformation. We believe that the ArO-M σ bond is cleaved
upon treating 18 or 23 with an excess of H2 to afford an Ar-
OH moeity. Competitive, reversible O-H activation and
irreversible activation of the Ar-OH bond yields 19 or 24.
Water was presumably formed, although we could not detect
it.
1
Spectroscopic Analysis. The H, 31P, 13C, and 29Si NMR spectra
Summary and Conclusions
were recorded at 400.19, 161.9, 100.6, and 79.5 MHz, respectively,
on a Bruker AMX 400 NMR spectrometer. 1H, 31P{1H}, and 13C{1H}
NMR spectra were also recorded at 250.17, 101.3, and 62.9 MHz,
respectively, on a Bruker DPX 250 NMR spectrometer. All chemical
shifts (δ) are reported in ppm, and coupling constants (J) are in hertz
We have demonstrated, for the first time, that metal insertion
(Rh(I), Pd(II)) into the strong aryl-oxygen single bond of an
aromatic ether and a phenol under mild homogeneous conditions
is possible. These fascinating processes occur directly without
prior activation of the substantially weaker alkyl-O bond.
1
(Hz). The H and 13C NMR chemical shifts are relative to tetrameth-
ylsilane; the resonance of the residual protons of the solvent was used
as an internal standard h1 (7.15 ppm benzene; 7.26 chloroform) and
(74) Ni, J.; Kubiak, C. P. In AdVances in Chemistry Series; Moser, W.
R., Slocum, D. W., Eds.; American Chemical Society: Washington, DC,
1992; Vol. 230, pp 515-528.
(75) Herde, J. L.; Senoff, C. V. Inorg. Nucl. Chem. Lett. 1971, 1029-
1031.

Transition-Metal-Based SelectiVity
J. Am. Chem. Soc., Vol. 120, No. 26, 1998 6539
all-d solvent peaks (128.0 ppm benzene; 77.0 chloroform), respectively.
31P NMR chemical shifts are relative to 85% H3PO4 in D2O at δ 0.0
(external reference), and 29Si NMR chemical shifts are relative to HSiEt3
(internal reference) with shifts downfield of the reference considered
positive. Assignments in the 1H and 13C{1H} NMR were made using
1H{31P}, 1H-1H COSY, and 13C-DEPT-135 NMR. All measurements
were carried out at 298 K unless otherwise specified. Ph3PO was used
as an internal standard for integration. IR spectra were recorded as
films between NaCl plates on a Nicolet 510 FT spectrometer. Field
desorption (FD) mass spectrometry was carried out using a JEOL JMS
SX/SX 102A four-sector mass spectrometer, coupled to a JEOL MS/
MP7000 data system; 10 µm tungsten wire FD emitters containing
carbon microneedles with an average length of 30 mm were used. The
samples were dissolved in methanol/water and then loaded onto the
emitter with the dipping technique. An emitter current of 0-15 mA
was used to desorb the sample. The ion source temperature was
generally 90 °C.
16.7 Hz, PPh2), 133.3 (d, JPC ) 18.8 Hz, PPh2), 131.3 (d, JPC ) 8.9
Hz, Cmeta), 129.3 (dd, JPC ) 9.8 Hz, JPC ) 2.4 Hz, Cortho,), 128.7 (s,
PPh2), 128.6 (d, JPC ) 6.4 Hz, PPh2), 123.8 (s, Cpara), 61.4 (t, JPC
)
4.5 Hz, OCH3), 30.0 (d, JPC ) 16.5, CH2P). 31P{1H} NMR (C6D6): δ
-11.4 (s). Anal. Calcd for C33H30P2: C, 78.56; H, 5.46. Found: C,
78.20; H 5.76.
Synthesis of 3. (a) Synthesis of Oxyuvitin Alcohol. An aqueous
formaldehyde solution (215 g, 37%) was added to a solution of p-cresol
(108 g, 1 mol) and NaOH (50 g) in H2O (200 mL). After 24 h of
stirring at room temperature, it was filtered and washed with an aqueous
saturated NaCl solution (1 L). The resulting salt was dissolved in water
(4 L) and neutralized with acetic acid to afford light-pink crystals, which
were filtered and washed with H2O. The product was dissolved in hot
ethyl acetate (400 mL), dried over Na2SO4, filtered hot, concentrated
to 270 mL, and recrystallized overnight. The formed crystals of
oxyuvitin alcohol were filtered and washed with cold ethyl acetate. 1H
NMR (CD3COCD3): δ 6.92 (s, 2H, m-C6H2), 4.70 (s, 4H, CH2OH),
3.19 (br, 3H, OH), 2.19 (s, 3H, CH3). (b) Synthesis of r,r′-Dibromo-
2-hydroxymesitylene. Oxyuvitin alcohol (7.8 g) was dissolved in a
HBr/acetic acid solution (33% HBr, 41 mL) and stirred overnight under
argon. The reaction mixture was diluted with excess H2O, and the
formed solid was filtered and dried in high vacuum to afford the
dibromide (10.2 g, 75%). 1H NMR (CDCl3): δ 7.06 (s, 2H, m-C6H2),
5.5 (br, 1H, OH), 4.50 (s, 4H, CH2Br), 2.24 (s, 3H, ArCH3). 13C{1H}
NMR (CDCl3): δ 150.94, 131.82, 130.50, 124.98 (s, Ar), 29.52 (CH2-
Br) 20.23 (s, ArCH3). (c) Phosphination. The procedure is analogous
to the one used for 1. White crystals were obtained from cold pentane
(-30 °C), and the product was further purified by column chromatog-
raphy (hexane/THF ) 95:5) to afford pure 3 (5.7 g, 75%). 1H NMR
(CDCl3): δ 7.58 (br, 1H, OH), 6.97 (s, 2H, m-C6H2), 2.95 (d, 4H, 2JPH
Synthesis of r,r′-Dibromo-2-methoxy-m-xylene. 2-Methoxy-m-
xylene (13.6 g, 100 mmol), N-bromosuccinimide (35.5 g; 200.0 mmol),
AIBN (∼0.1 g) and CCl4 (250 mL) were placed in a 500 mL three-
necked round-bottom flask equipped with an argon inlet, condenser,
and a stirring bar. The reaction mixture was heated to 80 °C and
refluxed overnight. The mixture was slowly cooled and filtered, and
the resulting solution was washed with H2O (3 × 25 mL), dried over
Na2SO4, and concentrated by rotary evaporation. The residue was kept
overnight at -20 °C to recrystallize and filtered, and the resulting
lachrimating solid was washed with cyclohexane (3 × 25 mL) to give
a white solid (22.3 g; 76%). 1H NMR (CDCl3): δ 7.31 (d, 2H, JHH
)
7.6 Hz, m-C6H3), 7.06 (t, 1H, JHH ) 7.6 Hz, p-C6H3), 4.50 (s, 4H,
CH2Br), 3.97 (s, 3H, OCH3). 13C{1H} NMR (CDCl3): δ 157.2 (Cipso),
132.8 (Cmeta), 132.5 (Cortho), 125.7 (Cpara), 62.9 (s, OCH3), 28.1 (s, CH2-
Br). Anal. Calcd for C9H10O1Br2: C, 36.77; H, 3.43; Br, 54.36.
Found: C, 36.85; H, 3.33; Br 54.66.
3
) 2.8 Hz, CH2P), 2.26 (s, 3H, CH3), 1.19 (d, 36H, JPH ) 11.2 Hz,
C(CH3)3). 13C{1H} NMR (CDCl3): δ 150.22 (t, JPC ) 2.0 Hz, Cipso),
129.15 (dd, JPC ) 1.8 Hz, JPC ) 8.5 Hz, Cortho), ∼129 (s, partly
overlapped, Cpara), 127.20 (d, JPC ) 7.5 Hz, Cmeta), 31.72 (d, JPC
19.3 Hz, C(CH3)3), 29.60 (d, JPC ) 12.1 Hz, C(CH3)3), 23.24 (d, JPC
)
)
Synthesis of 1. R,R′-Dibromo-2-methoxy-m-xylene (1.0 g, 3.4
mmol) and di-tert-butylphosphine (1.0 g, 6.8 mmol) in 5 mL of acetone
were refluxed with stirring for 40 min under argon to afford a white
precipitate. The solid was filtered and washed with ether, and the
resulting diphosphonium salt was dissolved in distilled degassed water
(10 mL) and treated with a solution of sodium acetate (4 g, 48 mmol)
in water (10 mL). The precipitated diphosphine 1 was extracted with
ether (3 × 50 mL) and dried over Na2SO4, and the ether solution was
filtered via a sinter tube under argon pressure. The solvent was
evaporated under vacuum, and the solid was extracted with ether. The
ether extract was filtered, and the solvent was evaporated, giving 0.86
19.3 Hz, CH2P), 20.56 (s, CH3). MS (M + H) ) 425. Anal. Calcd
for C25H45O1P2: C, 70.7; H, 10.9. Found: C, 70.3; H 10.7.
Synthesis of 4. (a) Synthesis of 2-Ethoxy-m-xylene. Ethyl iodide
(18.3 mL, 0.2 mol), NaH (6.0 g), and DMF (50 mL) were added to a
THF solution (500 mL) of 2,6-dimethylphenol (12.2 g, 0.1 mol) and
refluxed for 36 h at 85 °C. Upon cooling, a white percipitate was
formed. The reaction mixture was poured over H2O (1 L), saturated
with NaCl, dried over Na2SO4, filtered and concentrated in vacuo, and
dried under high vacuum to give a white powder (12.6 g, 85%).36 (b)
Synthesis of r,r′-Dibromo-2-ethoxy-m-xylene. The bromination was
performed analogously to the synthesis of R,R′-dibromo-2-methoxy-
m-xylene. 1H NMR (CDCl3): δ 7.35 (d, 2H, JHH ) 7.6 Hz, m-C6H3),
7.08 (t, 1H, JHH ) 7.6 Hz, p-C6H3), 4.54 (s, 4H, CH2Br), 4.16 (q, 3H,
JHH ) 7.0 Hz, OCH2CH3), 1.50 (t, 3H, JHH ) 7.0 Hz, CH2CH3). 13C-
{1H} NMR (CDCl3): δ 155.68, 132.14, 131.93, 124.84 (s, Ar), 70.49
(s, OCH2CH3), 27.72 (s, CH2Br), 15.17 (s, OCH2CH3). (c) Phosphi-
nation. The procedure is analogous to the one used for the preparation
of 1. White crystals of pure 4 (75%) were obtained from cold pentane
(-30 °C). 1H NMR (CDCl3): δ 7.28 (d, 2H, 3JHH ) 7.6 Hz, m-C6H3),
6.91 (t, 1H, 3JHH ) 7.6 Hz, p-C6H3), 3.84 (q, 2H, 3JHH ) 7.0 Hz, OCH2-
g (60%) of 1 as a white solid. 1H NMR (C6D6): δ 7.62 (d, 2H, JHH
)
7.7 Hz, m-C6H3), 7.01 (t, 1H, JHH ) 7.6 Hz, p-C6H3), 3.58 (s, 3H,
OCH3), 2.92 (d, 4H, JPH ) 9.4 Hz, CH2P), 1.13 (d, JPH ) 10.5 Hz,
C(CH3)3). 13C{1H} NMR (C6D6): δ 155.6 (t, JPC ) 4.2 Hz, Cipso)
,
134.4 (d, JPC ) 11.9 Hz, Cmeta) 129.1 (dd, JPC ) 15.2 Hz, JPC ) 1.8
Hz, Cortho), 124.0 (s, Cpara), 61.1 (s, OCH3), 31.9 (d, JPC ) 11.9 Hz,
C(CH3)3), 29.8 (d, JPC ) 13.2 Hz, C(CH3)3), 21.1 (d, JPC ) 22.4 Hz,
CH2P). 31P{1H} NMR (C6D6): δ 34.3 (s). Anal. Calcd for
C25H46P2O1: C, 70.72; H, 10.92. Found: C, 70.43; H 10.73.
Synthesis of 2. In a 250 mL three-necked round-bottom flask
equipped with an argon inlet, dropping funnel, and stirring bar was
added 1.6 M n-BuLi in n-hexane (8.81 mL; 14.1 mmol) under argon,
and the flask was then cooled to -40 °C. A solution of HPPh2 (2.62;
14.1 mmol) in 30 mL of THF was added dropwise under vigorous
stirring. Subsequently the reaction mixture was cooled to -78 °C and
a THF (120 mL) solution of R,R′-dibromo-2-methoxy-m-xylene (2.07
g, 7.05 mmol) was added dropwise under vigorous stirring. The
reaction mixture was allowed to warm to room temperature and was
stirred overnight and concentrated on a rotary evaporator. Toluene
(100 mL) was added, and the mixture was filtered and concentrated
again. After addition of CHCl3 (100 mL), the mixture was filtered
again, concentrated, and recrystallized from toluene at -30 °C to give
2 as a white solid (3.23 g, 91%). 1H NMR (C6D6): δ 7.41 (td, 8H,
JHH ) 7.1 Hz, JPH ) 1.9 Hz, PPh2), 7.03 (m, 12H, JHH ) 7.0 Hz,
PPh2), 6.83 (dt, 2H, JHH ) 7.6 Hz, JPH ) 1.4 Hz, m-C6H3), 6.58 (t, 1H,
JHH ) 7.6 Hz, p-C6H3), 3.58 (s, 3H, OCH3), 3.46 (s, 4H, CH2P). 13C-
2
3
CH3), 2.77 (d, 4H, JPH ) 3.2 Hz, CH2P), 1.40 (t, 3H, JHH ) 7.0 Hz,
OCH2CH3), 1.05 (d, 36H, JPH ) 10.8 Hz, C(CH3)3). 13C{1H} NMR
(CDCl3): δ 154.15 (t, JPC ) 3.6 Hz, Cipso), 134.15 (d, JPC ) 25.8 Hz,
meta), 128.50 (d, JPC ) 30.6 Hz, Cortho), 123.56 (s, Cpara), 69.04 (s,
3
C
OCH2CH3), 31.56 (d, JPC ) 18.7 Hz, C(CH3)3), 29.47 (d, JPC ) 13.2
Hz, C(CH3)3), 20.77 (d, JPC ) 19.2 Hz, CH2P), 15.71 (s, OCH2CH3).
31P{1H} (CDCl3): δ 35.95 (s). Anal. Calcd for C26H48O1P2: C, 71.20;
H, 11.03. Found: C, 71.31; H 11.06.
Ar-O Activation by Rh(I). Formation of Complexes 5 and 7.
[(RhCl(C8H14)2)2] (19 mg; 0.027 mmol) and 2 equiv of 1 (23 mg; 0.054
mmol) were dissolved in benzene (2.5 mL). The resulting red solution
1
was stirred for 24 h at room temperature. 1H, H{31P}, and 31P{1H}
NMR analysis of the reaction mixure showed the formation of complex
5 (20% conversion by 31P{1H}).6 The reaction was completed after
approximately 3 h at 85 °C in a sealed pressure vessel (quantitative
conversion). No Rh(III)-D formation was observed by 2H NMR when
{1H} NMR (C6D6): δ 157.1 (t, JPC ) 4.2 Hz, Cipso), 139.4 (d, JPC
)

6540 J. Am. Chem. Soc., Vol. 120, No. 26, 1998
Van der Boom et al.
C6D6 was used. Addition of an authentic sample of 5 to the reaction
crystals. 1H NMR (C6D6): δ 6.90 (d, 2H, JHH ) 7.2 Hz, m-C6H3),
6.69 (t, 1H, JHH ) 7.5 Hz, p-C6H3), 2.96 (d, 2H, JHH ) 13.5 Hz, CH2P,
left part of ABq), 2.39 (dvt, 2H, JHH ) 13.5 Hz, JPH ) 4.0 Hz, CH2P,
right part of ABq), 1.35 (vt, 18H, JPH ) 6.8 Hz, C(CH3)3), 1.04 (vt,
18H, JPH ) 6.6 Hz, C(CH3)3). 31P{1H} NMR (C6D6): δ 101.1 (s).
FD-MS M+ 628 (correct isotope pattern). Anal. Calcd for C26H43O3F3P2-
Pd1: C, 49.65; H, 6.89. Found: C, 49.74; H 6.82.
1
mixture resulted in overlap of signals in H and 31P{1H} NMR. Free
1
cyclooctene was observed by H NMR and GC-MS analysis of the
reaction mixture. Performing the reaction at 140 °C in a sealed vessel
results also in the formation of complex 7 (14% conversion by 31P-
{1H} NMR after 4 h). Addition of paraformaldehyde (16 mg, 0.53
mmol) to the reaction mixture (or reacting the isolated 5) result, upon
heating to 140 °C, in quantitave formation of 7.6
Formation of Complex 19. A solution of 20 (23 mg, 0.058 mmol)
in 2 mL of toluene was added dropwise to a stirred suspension of Pd-
(CF3CO2)2 (20 mg, 0.057 mol) in 2 mL of toluene. The resulting
colorless solution was loaded into a high-pressure vessel equipped with
a stirring bar and heated to 130 °C for 1 h. The solution was filtered
over a cotton pad and pumped to dryness, resulting in quantitave
formation of complex 19 as a white powder. 1H NMR (C6D6): δ 6.98
(t, 1H, JHH ) 7.3 Hz, p-C6H3), 6.84 (d, 2H, JHH ) 7.3, Hz, m-C6H3),
2.83 (vt, 4H, JPH ) 4.1 Hz, CH2P), 1.18 (vt, 36H, JPH ) 6.9 Hz,
C(CH3)3). 13C-DEPT-135 NMR (C6D6): δ 125.40 (s, Cpara), 122.74
(vt, JPC ) 10.2 Hz, Cmeta), 33.09 (vt, JPC ) 10.7 Hz, CH2P), 29.00 (vt,
JPC ) 3.1 Hz, C(CH3)3). 31P{1H} (C6D6): δ 77.85 (s). FD-MS M+
612 (correct isotope pattern).
Formation of Complexes 8 and 12. A THF solution (5 mL) of 3
(21 mg; 0.050 mmol) was treated with an excess of KH or NaH (30
equiv) and stirred at room temperature for 45 min. The suspension
was filtered and added dropwise to a stirred solution of [RhCl(C8H14)2]2
(18 mg; 0.025 mmol) in THF (5 mL). The resulting yellowish solution
was stirred overnight and filtered, and the solvent was removed in vacuo
to afford 8 quantitatively as a yellow-green powder. A loosely sealed
vial containing a THF solution (5 mL) of 8 was left under a dinitrogen
atmosphere in a drybox for a week at room-temperature resulting in
the formation of 12 as orange prismatic X-ray quality crystals. Data
for 8. IR: 2095 cm-1 (film). 1H NMR (C6D6): δ 6.83 (s, 2H, m-C6H2),
3.20 (d, 2H, JHH ) 13.3 Hz, CH2P, left part of ABq), 2.58 (dvt, 2H,
JHH ) 13.4 Hz, JPH ) 4.0 Hz, CH2P, right part of ABq), 2.31 (s, 3H,
Me-O Activation by Ni(II). Formation of Complexes 21 and
22. A solution of 3 (59 mg, 0.14 mmol) in 5 mL of ethanol was added
dropwise to a stirred solution of NiI2 (44 mg, 0.14 mmol) in 5 mL of
ethanol. The resulting green solution was heated overnight to 130 °C
in a pressure flask and was pumped to dryness. After addition of
toluene (20 mL), the reaction mixture was filtered through a cotton
pad and was pumped to dryness, yielding 52 mg (61%) of 22 as a
green powder. Use of compound 1 (instead of 3) gave 21 in 67% yield.
Data for 22. 1H NMR (C6D6): δ 6.88 (s, 2H, m-C6H2), 2.60 (br dvt,
ArCH3), 1.16 (vt, 18H, JPH ) 12.3 Hz, C(CH3)3). (vt, 18H, JPH
)
11.6 Hz, C(CH3)3). 31P{1H} NMR (C6D6): δ 91.4 (d, JRhP ) 163.1
Hz). FD-MS M+ 554 (correct isotope pattern). Data for 12. 31P{1H}
NMR (C6D6; 161.9 MHz): δ 88.2 (dd, 1JRhP ) 168.0 Hz, 4JRhP ) 24.5
Hz).
Reaction of Complex 8 with Methyl Iodide. Formation of
Complex 9. Complex 8 (17 mg; 0.021 mmol) was dissolved in C6D6
(0.6 mL) and transferred to a 5 mm screwcap NMR tube, treated with
1 or 2 equiv of CH3I or 13CH3I (1-2 µL; 0.020-0.040 mmol,
respectively), and heated for 10 min at 80 °C. The progress of the
reaction was monitored at room temperature by NMR in short time
intervals. No intermediates were observed by 1H, 31P{1H}, or 13C-
{1H} NMR. After completion, all volatiles were removed in vacuo to
afford 9 quantitatively as a yellow-green powder. 1H NMR (C6D6;
400.19 MHz): δ 6.85 (s, 2H, m-C6H2), 3.65 (ddt, JCH ) 142.0 Hz,
JRhH ) 3.0 Hz, JPH ) 3.9 Hz, 3H, RhCH3), 3.52 (dvt, 2H, JHH ) 12.8
Hz, JPH ) 4.2 Hz, CH2P, left part of ABq), 2.57 (dvt, 2H, JHH ) 12.8
Hz, JPH ) 3.7 Hz, CH2P, right part of ABq), 2.20 (s, 3H, ArCH3), 1.26
(m, 36H, JPH ) 11.4 Hz, C(CH3)3). 31P{1H} NMR (C6D6): δ 58.0 (d,
2
2
2H, JHH ) 13.0 Hz, left part of ABq, CH2P), 2.56 (m, 2H, JHH
)
13.0 Hz, right part of ABq, CH2P), 2.30 (s, 3H, ArCH3), 1.60 (vt, 18H,
JPH ) 6.2 Hz, C(CH3)3), 1.37 (vt, 18H, JPH ) 6.0 Hz, C(CH3)3). 31P-
{1H} NMR (C6D6): δ 81.01 (s). FD-MS M+ 609 (correct isotope
pattern). Anal. Calcd for C25H45O1P2Ni1I1: C, 49.29; H, 7.45.
Found: C, 50.61; H 7.64. Data for 21. 1H NMR (C6D6): δ 7.04 (d,
2H, JHH ) 7.5 Hz, m-C6H3), 6.86 (t, 1H, JHH ) 7.6 Hz, p-C6H3), 2.58
(br dvt, 2H, JHH ) 13.2 Hz, left part of ABq, CH2P), 2.39 (dvt, 2H,
JHH ) 13.2 Hz, JPH ) 4.9 Hz, right part of ABq, CH2P), 1.58 (vt, 18H,
JPH ) 6.2 Hz, C(CH3)3), 1.34 (vt, 18H, JPH ) 6.0 Hz, C(CH3)3). 31P-
{1H} NMR (C6D6): δ 77.69 (s). FD-MS M+ 595 (correct isotope
pattern).
1
JRhP ) 118.7 Hz). 13C{1H} NMR (C6D6): δ 10.0 (dt, JRhC ) 27.4
Hz, 2JPC ) 2.7 Hz, RhCH3). FD-MS M+ 668 (correct isotope pattern).
Methoxy Transfer. Formation of Complex 16. [RhCl(C8H14)2]2
(15 mg; 0.021 mmol) and PPh3 (11 mg; 0.042 mmol) were dissolved
in C6D6 or dioxane (1 mL) and added to 2 (20 mg; 0.040 mmol). The
resulting dark red solution was stirred for 30 min, loaded into a 5 mm
screwcap NMR tube, treated with 50 µL of HSiR3 (R ) OCH2CH3,
Formation of Complex 23. A solution of 3 (40 mg, 0.094 mmol)
in 2 mL of toluene was added dropwise to a stirred suspension of Pd-
(CF3CO2)2 (31 mg, 0.093 mmol) in 2 mL of toluene. The resulting
red solution was stirred overnight at room temperature. 31P{1H}
analysis of an aliquid showed 23 as the only product. Consequently,
the reaction mixture was filtered through a cotton pad and pumped to
dryness, resulting in quantitave formation of 23 as an orange powder.
The complex can be recrystallized by slowly concentration a benzene
solution to give orange X-ray suitable crystals. 1H NMR (C6D6): δ
1
CH2CH3), and heated for approximately 30 min at 130 °C. 1H, H-
{31P}, 31P{1H}, and 13C{1H} NMR analysis of the reaction solution
showed the formation of 16 as major organometallic product (∼95%).
FD-MS showed the molecular ion (M+ 838) and a correct isotope
pattern. GC-MS (CI and EI) and 29Si NMR analysis of the product
solution shows the presence of (MeO)SiR3 (R ) OCH2CH3,
CH2CH3).57-59 Formation of ClSi(CH2CH3)3 was observed by GC-
MS. The gas phase was collected by a standard vacuum line technique
and analyzed by GC, showing only traces of methane (<3%).
Formation of (Me)SiR3 (R ) OCH2CH3, CH2CH3) is not observed either
2
6.73 (s, 2H, m-C6H2), 3.03 (d, 2H, left part of ABq, JHH ) 13.5 Hz,
2
CH2P), 2.41 (dvt, 2H, right part of ABq, JHH ) 13.5 Hz, JPH ) 4.0
Hz, CH2P), 2.17 (s, 3H, ArCH3), 1.35 (vt, 18H, JPH ) 6.7 Hz, C(CH3)3),
1.05 (vt, 18H, JPH ) 6.4 Hz, C(CH3)3). 13C-DEPT-135 NMR (C6D6):
δ 127.89 (vt, JPC ) 2.0 Hz Hz, Cmeta), 28.74 (vt, JPC ) 3.1 Hz, C(CH3)3),
28.40 (vt, JPC ) 2.4 Hz, C(CH3)3), 25.13 (vt, JPC ) 6.7 Hz, CH2P),
21.22 (s, ArCH3). 31P{1H} (C6D6): δ 100.14 (s). FD-MS M+ 643
(correct isotope pattern).
1
by H NMR, 29Si NMR, or GC-MS analysis of the product solution.
Me-O Activation by Pd(II). Formation of Complex 18. Pd-
(CF3CO2)2 (12 mg, 0.036 mmol) and 1 (16 mg; 0.033 mmol) were
dissolved in C6D6 (3 mL). The resulting yellow solution was transferred
to a pressure vessel and heated for 6 h at 110 °C resulting in quantitave
Competitive sp2-sp3 and sp3-sp3 C-O Bond Activation by Pd-
(II). Reaction of 4 with Pd(CF3CO2)2. A solution of 4 (20 mg, 0.046
mmol) in 2 mL of toluene was added dropwise to a stirred suspension
of Pd(CF3CO2)2 (15 mg, 0.045 mmol) in 2 mL of toluene. The resulting
yellow solution was loaded into a high-pressure vessel equipped with
a stirring bar and heated for 3 h at 130 °C. Analysis of the gas phase
by GC showed traces of methane (<3%). The red reaction mixture
was filtered through a cotton pad and pumped to dryness, yielding an
1
1
formation of 18 as jugded by H, H{31P}, and 31P{1H} NMR. The
reaction proceeds also at room temperature. NMR analysis of the
reaction mixture shows the presence of unreacted 1 (60%) and complex
18 (40%) after 24 h. Heating the reaction mixture for 3 h at 85 °C
drives the reaction to completion. CF3CO2CH3 formation was observed
1
1
orange powder. 1H, H-1H COSY, H{31P}, 31P{1H}, 13C{1H}, and
13C-DEPT-135 NMR analysis of the powder showed the presence of
18 and 19 (9:1). CF3CO2Et was observed by NMR and GC-MS
analysis.72 Performing the same reaction for 16 h does not change the
1
by H NMR analysis of the product solution and by comparison with
an authentic added sample. A loosely sealed vial containing a benzene
solution (5 mL) of 18 was left under a nitrogen atmosphere for 1 week
at room-temperature resulting in formation of orange X-ray suitable

Transition-Metal-Based SelectiVity
J. Am. Chem. Soc., Vol. 120, No. 26, 1998 6541
product distribution. Complex 18 can be crystallized from the product
solution at room temperature to give orange crystals.
Å. An ORTEP view of the molecular structure and the adopted
numbering scheme are shown in Figure 1. Table 1 gives details of the
crystal structure determination.
Ar-O Hydrodeoxygenation (HDO) by Pd(II). Formation of
Complexes 19 and 24. A red toluene solution (3 mL) of 18 or 23 (15
mg) was loaded into a Fischer-Porter pressure tube, charged with H2
(30 psi), and heated at 180 °C for 8 h. 31P{1H} analysis of the colorless
reaction mixture indicated the formation of complexes 19 or 24 in
∼55% yield, some decomposition products, and the consumption of
the starting material (18 or 23). Subsequently, the reaction mixture
was filtered through a cotton pad and pumped to dryness to yield a
white powder. Complexes 19 and 24 were not further purified and
have been identified in situ. Addition of an authentic sample of 19 to
the reaction mixture of 18 resulted in overlap of signals in 1H and 31P-
{1H} NMR. Data for 24. 1H NMR (C6D6): δ 6.63 (s, 2H, m-C6H2),
2.81 (vt, 4H, JPH ) 4.0 Hz, CH2P), 2.18 (s, 3H, ArCH3), 1.19 (vt, 36H,
JPH ) 6.9 Hz, C(CH3)3). 13C-DEPT-135 NMR (C6D6): δ 123.71 (vt,
JPC ) 10.4 Hz, Cmeta), 32.97 (vt, JPC ) 10.8 Hz, CH2P), 29.03 (vt, JPC
) 3.5 Hz, C(CH3)3), 26.69 (s, ArCH3). 31P{1H} (C6D6): δ 77.60 (s).
FD-MS M+ 627 (correct isotope pattern).
X-ray Crystal Structure Determination of Complex 12. An
orange prismatic crystal (0.25 × 0.25 × 0.3 mm3) was mounted on a
glass fiber and flash frozen in a cold nitrogen stream (at 110 K) on a
Rigaku AFC5R four-circle diffractometer mounted on a rotating anode
with Mo KR radiation and a graphite monochromator (λ ) 0.710 73
Å). Accurate unit cell dimensions were obtained from a least-squares
fit to setting angles of 25 reflections in the range 1.39° e θ e 27.52°,
-32 e h e 33, -1e k e 23, -15 e l e 3; ω scan method, scan
speed 8°/min, typical half-height peak width ) 0.45°, three standards
were collected 40 times each with an 3% change in intensity, 6056
independent reflections. Structure 12 was solved using direct methods
(SHELXS-92) and refined by full-matrix least-squares techniques based
on F2 (SHELXL-93). The final cycle of the least-squares refinement
gave an agreement factor R1 of 0.0330 for data I > 2σ and R1 ) 0.0450
for all data based on 6056 reflections, goodness-of-fit on F2 ) 1.054,
largest electron density ) 0.928 e/Å. Hydrogens were calculated in
idealized positions and refined in a riding mode. The molecule is a
dimer with one-half a molecule per asymmetric unit, i.e., the full dimer
is also the crystallographic dimer. The N(2)-N(2a) distance is 1.129
X-ray Crystal Structure Determination of Complex 18. An
orange plate crystal (0.1 × 0.1 × 0.05 mm3) was mounted on a glass
fiber and flash frozen in a cold nitrogen stream (at 110 K) on a Rigaku
AFC5R four-circle diffractometer mounted on a rotating anode with
Mo KR radiation and a graphite monochromator (λ ) 0.710 73 Å).
Accurate unit cell dimensions were obtained from a least-squares fit
to setting angles of 25 reflections in the range 1.14° e θ e 23.02°,
-11 e h e 11, -17e k e 17, -21 e l e 21; ω scan method, scan
width ) 1.2°, scan speed 8°/min, typical half-height peak width ) 0.40°,
three standards were collected 84 times each with an 4% change in
intensity, 16 284 reflections collected, 8020 independent reflections.
Structure 18 was solved using direct methods (SHELXS-92) and refined
by full-matrix least-squares techniques based on F2 (SHELXL-93). The
space group is P1h with two independent molecules per asymmetric
unit. The final cycle of the least-squares refinement gave an agreement
factor R1 of 0.0630 (based on F2) for data I > 2σI and R1 ) 0.1301 for
all data based on 6408 reflections, goodness-of-fit on F2 ) 1.159, largest
electron density ) 0.701 e/Å. Hydrogens were calculated in idealized
positions and refined in a riding mode.
Acknowledgment. This work was supported by the Israel
Science Foundation, Jerusalem, Israel, and by the MINERVA
Foundation, Mu¨nich, Germany. D.M. is the holder of the Israel
Matz Professorial Chair in Organic Chemistry. We thank Han
Peeters and Roel Fokkens (University of Amsterdam, The
Netherlands) for performing the FD-MS experiments.
Supporting Information Available: Tables of crystal data
and structure refinement, atomic coordinates, bond lengths and
angles, anisotropic displacement parameters, and hydrogen atom
coordinates for complexes 12 and 18 (16 pages, print/PDF).
See any current masthead page for ordering information and
Web access instructions.
JA9738889