Bond angle effects on the migratory insertion of ethylene and carbon monoxide into palladium(II)-methyl bonds in complexes bearing bidentate phosphine ligands

Article abstract of DOI:10.1021/om010489k

Labile (P-P)Pd(CH₃)(OEt₂)⁺BAr′₄^- complexes (2) have been prepared via protonation of (P-P)PdMe₂ (1), where P-P = cis-1,2-bis(diphenylphosphino)ethylene (a, dppee), 1,2-bis(diphenylphosphino)benzene (b, dpbz), 1,2-bis(diphenylphosphino)ethane (c, dppe), 1,2-bis(dimethylphosphino)ethane (d, dmpe), 1,3-bis(diphenylphosphino)propane (e, dppp), 1,3-bis(diisopropylphosphino)propane (f, dippp), 1,4-bis(diphenylphosphino)butane (g, dppb) and Ar′ = 3,5-(CF₃)₂C₆H₃. Unstable complex 2d (P-P = dmpe) was generated in situ. X-ray structures are reported for 1e and 2e-g. Treatment of 2a-g with CO in CH₂Cl₂ at -90 °C yields the (P-P)Pd(CH₃)(CO)⁺ complexes 3a-g. Barriers to migratory insertion in 3a-g were determined with the ordering to be: 3f (dippp) ≈ 3g (dppb) < 3e (dppp) ? 3a (dppee) ≈ 3b (dpbz) ≈ 3c (dppe) < 3d (dmpe). Exposure of 2a-g to ethylene at -80 °C yields the ethylene complexes (P-P)Pd(CH₃)(C₂H₄)⁺ (5a-g). Barriers to migratory insertion in these complexes were determined by NMR spectroscopy to be: 5b (dpbz) ≈ 5e (dppp) ≈ 5f (dippp) ≈ 5g (dppb) < 5a (dppee) ≈ 5c (dppe) < 5d (dmpe). Complexes (P-P)Pd(CH₂CH₃)(C₂H₄)⁺ (6a-e,g) produced from 5a-e,g under C₂H₄ are catalyst resting states for the dimerization of C₂H₄ to butenes. In the case of 5f (P-P = dippp), the catalyst resting state produced is the β-agostic ethyl complex (dippp)Pd(CH₂CH₂-μ-H)⁺ (8f), which has been isolated. This complex exhibits two dynamic processes studied by VT-NMR: interchange of C_α and C_β (ΔG?=10.3(2) kcal/mol) and rotation of the agostic methyl group (ΔG? ca. 6.4 kcal/mol). The β-agostic propyl complex 7f has been generated and identified as the n-propyl isomer (dippp)-Pd(CH₂CH₂-μ-HCH₃)⁺.

Full text of DOI:10.1021/om010489k

5266
Organometallics 2001, 20, 5266-5276
Bon d An gle Effects on th e Migr a tor y In ser tion of
Eth ylen e a n d Ca r bon Mon oxid e in to
P a lla d iu m (II)-Meth yl Bon d s in Com p lexes Bea r in g
Bid en ta te P h osp h in e Liga n d s
J ohn Ledford, C. Scott Shultz, Derek P. Gates, Peter S. White,
J oseph M. DeSimone, and Maurice Brookhart*
Department of Chemistry, University of North Carolina at Chapel Hill,
Chapel Hill, North Carolina 27599-3290
Received J une 8, 2001
-
Labile (P-P)Pd(CH3)(OEt2)+BAr′4 complexes (2) have been prepared via protonation of
(P-P)PdMe2 (1), where P-P ) cis-1,2-bis(diphenylphosphino)ethylene (a , dppee), 1,2-bis-
(diphenylphosphino)benzene (b, dpbz), 1,2-bis(diphenylphosphino)ethane (c, dppe), 1,2-bis-
(dimethylphosphino)ethane (d , dmpe), 1,3-bis(diphenylphosphino)propane (e, dppp), 1,3-
bis(diisopropylphosphino)propane (f, dippp), 1,4-bis(diphenylphosphino)butane (g, dppb) and
Ar′ ) 3,5-(CF3)2C6H3. Unstable complex 2d (P-P ) dmpe) was generated in situ. X-ray
structures are reported for 1e and 2e-g. Treatment of 2a -g with CO in CH2Cl2 at -90 °C
yields the (P-P)Pd(CH3)(CO)+ complexes 3a -g. Barriers to migratory insertion in 3a -g
were determined with the ordering to be: 3f (dippp) ≈ 3g (dppb) < 3e (dppp) , 3a (dppee)
≈ 3b (dpbz) ≈ 3c (dppe) < 3d (dmpe). Exposure of 2a -g to ethylene at -80 °C yields the
ethylene complexes (P-P)Pd(CH3)(C2H4)+ (5a -g). Barriers to migratory insertion in these
complexes were determined by NMR spectroscopy to be: 5b (dpbz) ≈ 5e (dppp) ≈ 5f (dippp)
≈ 5g (dppb) < 5a (dppee) ≈ 5c (dppe) < 5d (dmpe). Complexes (P-P)Pd(CH2CH3)(C2H4)+
(6a -e,g) produced from 5a -e,g under C2H4 are catalyst resting states for the dimerization
of C2H4 to butenes. In the case of 5f (P-P ) dippp), the catalyst resting state produced is
the â-agostic ethyl complex (dippp)Pd(CH2CH2-µ-H)+ (8f), which has been isolated. This
complex exhibits two dynamic processes studied by VT-NMR: interchange of CR and Câ (∆Gq
) 10.3(2) kcal/mol) and rotation of the agostic methyl group (∆Gq ca. 6.4 kcal/mol). The
â-agostic propyl complex 7f has been generated and identified as the n-propyl isomer (dippp)-
Pd(CH2CH2-µ-HCH3)+.
In tr od u ction
Cationic palladium(II) complexes bearing neutral
bidentate ligands are excellent catalysts for the alter-
nating copolymerization of CO and olefins as well as
olefin dimerization, particularly dimerization of ethyl-
ene to butenes.1-5 The fundamental steps in these
transformations involve the migratory insertion reac-
tions of alkyl carbonyl, acyl olefin, and alkyl olefin
species as shown in eqs 1-3.
Such migratory insertion reactions are members of a
larger class of transition-metal-mediated C-C bond
forming reactions that are observed in processes such
as hydroformylation, hydroesterification, olefin oligo-
merization, and polymerization. Several groups, includ-
ing ours, have examined aspects of these insertion
reactions employing Pd(II) systems containing bidentate
nitrogen ligands.6-9 For example, we have measured the
free energies of activation for migratory insertion for
the series (phen)Pd(CH3)(CO)+BAr′4-, (phen)Pd(CH3)-
(C2H4)+BAr′4-, (phen)Pd(CH2CH3)(C2H4)+BAr′4-, and
(phen)Pd(COCH3)(C2H4)+BAr′4- (phen ) η2-1,10-phenan-
-
throline, BAr′4 ) B(3,5-C6H3(CF3)2)4-) and, in combi-
nation with other kinetic and thermodynamic studies,
have constructed a complete catalytic cycle for the
copolymerization of ethylene and CO using the phenan-
throline-based Pd(II) system.7,8
(1) Drent, E. Pure Appl. Chem. 1990, 62, 661.
(2) Drent, E.; van Broekhoven, J . A. M.; Doyle, J . J . J . Organomet.
Chem. 1991, 417, 235.
(3) Sen, A. Acc. Chem. Res. 1993, 26, 303.
(4) Sen, A.; J iang, Z. Macromolecules 1993, 26, 911.
(5) For reviews see: Drent, E.; Budzelaar, P. H. M. Chem. Rev. 1996,
96, 663. Sommazzi, A.; Garbassi, F. Prog. Polym. Sci. 1997, 22, 1547.
Sen, A. Adv. Polym. Sci. 1986, 73-74, 125.
Since commercial catalyst systems for the production
of ethylene/CO copolymers have used bidentate phos-
phine ligands,10,11 we have recently focused our atten-
10.1021/om010489k CCC: $20.00 © 2001 American Chemical Society
Publication on Web 11/07/2001

Migratory Insertion into Pd(II)-Methyl Bonds
Organometallics, Vol. 20, No. 25, 2001 5267
prior to insertion. Thus, again, the primary insertion
barrier is not known. We have previously measured the
barrier of migratory insertion in complex 3a , (dppp)Pd-
(CH3)(CO)+, to yield ∆Gq ) 14.8 kcal/mol,12 which is
consistent with that observed qualitatively by To`th and
Elsevier, who reported the in situ generation of [(2S,4S)-
2,4-bis(diphenylphosphino)pentane]Pd(CH3)(CO)+ and
observed the half-life for migratory insertion, which was
indicative of a ∆Gq value of ∼14-15 kcal/mol.15
There have been several theoretical treatments of the
migratory insertion processes noted in eqs 1-3. Thorn
and Hoffman’s extended Hu¨ckel analysis illustrated that
for the migratory insertion of (PH3)2Pt(H)(C2H4)+ the
P-Pt-P angle goes from 95° in the starting cis complex
through 110° in the transition state.16 This suggests that
an appropriate bidentate phosphine ligand with a
P-Pt-P bond angle close to 110° would help to desta-
bilize the ground state with respect to the transition
state, thereby reducing the activation barrier. Widening
of the F-Pt-PH3 angle was also predicted by Sakaki
et al. to occur in the transition state for migratory
insertion of (PH3)(F)Pt(CH3)(CO) on the basis of ab initio
Hartree-Fock calculations.17 Ziegler et al. used cis-H2-
PCHdCHPH2 as the model bidentate ligand and cal-
culated barriers for migratory insertion in (P-P)Pd-
(C2H5)(CO)+ and (P-P)Pd(C2H5)(C2H4)+ of 11.5 and 15.6
kcal/mol, respectively.18,19 Morokuma and co-workers
carried out calculations using the diimine ligand HNd
CHCHdNH as the model bidentate ligand and predicted
barriers of insertion for (HNdCHCHdNH)Pd(CH3)-
(CO)+ and (HNdCHCHdNH)Pd(CH3)(C2H4)+ of 14.9
and 16.2 kcal/mol, respectively.20,21
In this paper we report detailed investigations of the
generation of complexes of the general formula (P-P)-
Pd(CH3)(CO)+ and (P-P)Pd(R)(C2H4)+ (R ) -CH3,
-CH2CH3) and the measurement of the barriers to
migratory insertion employing a variety of bidentate
phosphine ligands, a -g (shown in Figure 2). These
results provide quantitative information regarding these
barriers, with specific attention being paid to the role
of the P-Pd-P bond angle and donor properties of the
bidentate phosphine ligand. The catalytic cycle for
ethylene dimerization has been investigated; the cata-
lyst resting state is shown to be a function of the
structure of the bidentate phosphorus ligand.
F igu r e 1. Chiral phosphine-phosphite used by Nozaki
et al.
tion on studying the fundamental migratory insertion
reactions (eqs 1-3) of palladium complexes containing
bidentate phosphine based ligands, specifically the 1,3-
bis(diphenylphosphino)propane (dppp) ligand.12 We have
now extended this study to examine a series of bidentate
phosphine ligands to determine the effects that chang-
ing the structural and/or electronic properties of the
ligand have on migratory insertion processes. An in-
triguing observation concerning the overall catalytic
activity of these systems in the copolymerization of
ethylene and carbon monoxide is that the size of the
chelate ring dramatically affects productivity, with the
most productive catalyst being the one with a ligand
containing a three-carbon backbone.5,13
A number of studies concerning the insertion of CO
into PdII-CH3 bonds in (P-P)PdII complexes have been
reported. Dekker et al. have examined the reactions of
the series of complexes [(C6H5)2P(CH2)nP(C6H5)2]Pd-
(CH3)(Cl) and [(C6H5)2P(CH2)nP(C6H5)2]Pd(CH3)(CH3-
CN)+ with CO to yield acyl complexes.13 Qualitative rate
differences were observed with (n ) 4) g (n ) 3) . (n )
2). In these studies, conversion of the starting complexes
to acyl derivatives involves first the displacement of Cl-
(or CH3CN) by CO to yield the unobserved (P-P)Pd-
(CH3)(CO)+ followed by methyl migration to form the
acyl product. Since the (P-P)Pd(CH3)(CO)+ complexes
are not directly observed, it is not clear how the half-
lives of the overall transformations relate to the primary
migratory insertion step (eq 1). A similar experiment
reported by Nozaki et al.14 revealed the ∆Gq value for
the conversion of the chelate complex ((R,S)-BINAPHOS)-
PdCH2CH(CH3)C(O)CH3+ to the acyl-acetonitrile com-
plex ((R,S)-BINAPHOS)Pd(NCCH3)(C(O)CH2CH(CH3)C-
(O)CH3)+ to be ca. 19 kcal/mol (Figure 1). There is most
likely a significant energy cost for initial chelate opening
in this case to form the intermediate (P-P)Pd(R)(CO)+
Resu lts a n d Discu ssion
1. Syn th esis of Dieth yl Eth er Ad d u cts, (P -P )P d -
+
(CH )(OEt2) BAr ′4- (Ar ′ ) 3,5-(CF 3)2C6H3; 2a -g). A
3
series of seven bidentate phosphine ligands, including
the previously studied 1,3-bis(diphenylphosphino)pro-
pane (dppp) ligand,12 was selected for the present study
(see Figure 2). Analogous to previous work employing
bidentate nitrogen ligands,7-9,22,23 ether adducts of the
(6) Ru¨lke, R. E.; Delis, J . G. P.; Groot, A. M.; Elsevier, C. J .; van
Leeuwen, P. W. N. M.; Vrieze, K.; Goubitz, K.; Schenk, H. J .
Organomet. Chem. 1996, 508, 109.
(7) Rix, F. C.; Brookhart, M. J . Am. Chem. Soc. 1995, 117, 1137.
(8) Rix, F. C.; Brookhart, M.; White, P. S. J . Am. Chem. Soc. 1996,
118, 4746.
(9) Brookhart, M.; Rix, F. C.; DeSimone, J . M. J . Am. Chem. Soc.
1992, 114, 5894.
(10) Mul, W. P.; Oosterbeek, H.; Beitel, G. A.; Kramer, G.-J .; Drent,
E. Angew. Chem., Int. Ed. 2000, 39, 1848.
(15) To´th, I.; Elsevier, C. J . J . Am. Chem. Soc. 1993, 115, 10388.
(16) Thorn, D. L.; Hoffmann, R. J . Am. Chem. Soc. 1978, 100, 2079.
(17) Sakaki, S.; Kitaura, K.; Morokuma, K.; Ohkubo, K. J . Am.
Chem. Soc. 1983, 105, 2280.
(11) Drent, E.; Budzelaar, P. H. M. Chem. Rev. 1996, 96, 663.
(12) Shultz, C. S.; Ledford, J .; DeSimone, J . M.; Brookhart, M. J .
Am. Chem. Soc. 2000, 122, 6351.
(13) Dekker, G. P. C. M.; Elsevier, C. J .; Vrieze, K.; van Leeuwen,
P. W. N. M. Organometallics 1992, 11, 1598.
(18) Margl, P.; Ziegler, T. Organometallics 1996, 15, 5519.
(19) Margl, P.; Ziegler, T. J . Am. Chem. Soc. 1996, 118, 7337.
(20) Koga, M.; Morokuma, K. J . Am. Chem. Soc. 1986, 108, 6136.
(21) Svensson, M.; Matsubara, T.; Morokuma, K. Organometallics
1996, 15, 5568.
(14) Nozaki, K.; Sato, N.; Tonomura, Y.; Yasutomi, M.; Takaya, H.;
Hiyama, T.; Matsubara, T.; Koga, N. J . Am. Chem. Soc. 1997, 119,
12779.
(22) J ohnson, L. K.; Killian, C. M.; Brookhart, M. J . Am. Chem. Soc.
1995, 117, 6414.
(23) Tempel, D. J .; Brookhart, M. Organometallics 1998, 17, 2290.

5268 Organometallics, Vol. 20, No. 25, 2001
Ledford et al.
F igu r e 2. Bidentate phosphine ligands studied.
Ta ble 1. Selected NMR Da ta a for (P -P )P d CH3(OEt2)+BAr ′4
-
2a
dppee
2b
dpbz
2c
2d
2e
dppp
2f
2g
P-P
dppe
dmpe
dippp
dppb
1H NMR
0.52 (d, 7)
13C NMR
12.2 (d, 86)
b
c
[Pd]-CH3
[Pd]-CH3
0.63 (d, 7)
0.62 (d, 7)
-
-
0.44 (d, 4)
0.59 (br)
7.7 (d, 79)
0.69 (d, 7)
12.3 (d, 88)
13.8 (d, 68)
15.0 (d, 86)
16.5 (d, 87)
31P NMR
60.1 (d, 25)
36.8 (d, 25)
(P-P)[Pd]d
61.2 (br)
49.2 (br)
57.1 (br)
41.2 (d, 26)
-
-
30.3 (d, 51)
-0.2 (d, 51)
55.3 (d, 38)
11.0 (d, 38)
36.6 (br)
18.8 (br)
a
Spectra were collected in CD2Cl2 at -80 °C. Chemical shift values (δ) are relative to CD2Cl2. All couplings (given in parentheses) are
b
3
2
d
2
in Hz. Couplings are J HP
.
c Couplings are J CP
.
Couplings are J PP.
Ta ble 2. Selected Bon d Dista n ces (Å) a n d An gles
palladium methyl cations, (P-P)Pd(CH3)(OEt2)+, proved
to be excellent precursors for generating a number of
reactive catalytic intermediates through displacement
of the labile ether ligand at low temperatures. The
dimethyl complexes 1a-g, some of which are known,24-28
were prepared by a modification of the method previ-
ously reported by van Koten et al.24 through displace-
ment of TMEDA from (TMEDA)Pd(Me)2 in acetone at
25 °C. Compounds 2a -g were prepared by protonation
of the dimethyl complexes with H(Et2O)2BAr′4 in a
mixture of diethyl ether and CH2Cl2 at -30 °C (eq 4).12
(d eg) of Com p lexes 1e a n d 2e-g
1e
2e
2f
2g
Distances
2.065(5)
Pd-CH3
Pd-CH3
Pd-O
Pd-P1
Pd-P2
2.099(9)
2.092(11)
2.082(6)
2.226(4)
2.062(9)
2.226(5)
2.204(3)
2.307(3)
2.311(3)
2.3615(16) 2.3865(23) 2.392(3)
2.1985(16) 2.2069(18) 2.2039(24)
Angles
CH3-Pd-CH3 85.9(4)
CH3-Pd-O1
88.74(16)
97.26(6)
86.13(22)
97.66(7)
85.7(3)
102.42
P1-Pd-P2
P1-Pd-CH3
P2-Pd-CH3
P1-Pd-O1
93.91(9)
90.1(3)
90.0(3)
83.73(15)
90.59(9)
83.53(20)
92.65(11)
81.20(24)
90.93(17)
temperatures up to 25 °C for hours. Complex 2d was
unstable even at low temperatures (ca. -80 °C) and
1
could not be characterized by H or 31P NMR spectros-
copy; however, CO and olefin derivatives of 2d could be
prepared in a satisfactory manner from in situ prepared
2d . In methylene chloride solution, complex 2g contain-
ing the dppb ligand decays to other Pd(II) methyl
complexes in the temperature range of -40 to 5 °C,
depending on concentration. Traces of excess ligand
appear to accelerate this process. Similar behavior was
observed by van Leeuwen13,29 and was attributed to
dechelation of the poorly chelating dppb ligand and
formation of complexes involving dppb as a bridging
ligand. The ether adduct 2g could be handled without
problems from decay in a CH2Cl2/Et2O mixture at or
below 5 °C and in dilute solutions of CD2Cl2 below -40
°C.
With the exception of the dmpe complex 2d , the ether
1
adducts were readily isolated and characterized by H,
13C, and 31P NMR spectroscopy (Table 1) and can be
stored for months as solids at -30 °C and withstand
(24) de Graff, W.; Boersma, J .; Smeets, W. J . J .; Spek, A. L.; van
Koten, G. Organometallics 1989, 8, 2907.
(25) Ito, T.; Tsuchiya, H.; Yamamoto, A. Bull. Chem. Soc. J pn. 1977,
50, 1319.
(26) Ozawa, F.; Yamamoto, A. Nippon Kagaku Kaishi 1987, 5, 773.
(27) Tooze, R.; Chiu, K. W.; Wilkinson, G. Polyhedron 1984, 3, 1025.
(28) Fryzuk, M. D.; Clentsmith, G. K. B.; Rettig, S. J . J . Chem. Soc.,
Dalton Trans. 1998, 2007.
2. Str u ctu r a l Ch a r a cter iza tion of Com p lexes 1e
a n d 2e-g by X-r a y Cr ysta llogr a p h y. The lack of
(29) Dekker, G. P. C. M.; Elsevier, C. J .; Vrieze, K.; van Leeuwen,
P. W. N. M.; Roobeek, C. F. J . Organomet. Chem. 1992, 430, 357.

Migratory Insertion into Pd(II)-Methyl Bonds
Organometallics, Vol. 20, No. 25, 2001 5269
F igu r e 3. Crystal structure of (dppp)Pd(CH3)2 (1e). All
atoms are drawn as 50% probability ellipsoids.
Figu r e 5. Crystal structure of (dippp)Pd(CH3)(OEt2)+BAr′4
-
(2f). The counterion is not shown for clarity. All atoms are
drawn as 50% probability ellipsoids.
-
Figu r e 4. Crystal structure of (dppp)Pd(CH3)(OEt2)+BAr′4
(2e). The counterion is not shown for clarity. All atoms are
drawn as 50% probability ellipsoids.
structural data for dialkyl complexes and methyl ether
cations led us to characterize complexes 1e and 2e-g
by X-ray diffraction. The molecular structures of the
neutral dimethyl complex 1e and the cationic methyl
diethyl ether complexes 2e-g are shown in Figures
3-6, respectively. Selected bond distances and angles
are summarized in Table 2.
These structures provide useful information concern-
ing the P-Pd-P bond angles in the monocationic
systems and how they compare to the neutral systems.
The structures of 2e,f containing -(CH2)3- bridges are
quite similar with P-Pd-P bond angles of ca. 97°
(Figures 4 and 5) which are significantly larger than
the ideal 90° for a square-planar complex. The ether
complex 2g with a -(CH2)4- bridge shows an even
larger P-Pd-P bond angle of 102°, as expected (Figure
6). Other neutral Pd(II) complexes containing the dppp
ligand show somewhat smaller P-Pd-P angles relative
to 2e: (dppp)PdCl2 (90.6°)30 and (dppp)Pd(NCS)2 (89.3°).31
-
Figu r e 6. Crystal structure of (dppb)Pd(CH3)(OEt2)+BAr′4
(1g). The counterion is not shown for clarity. All atoms are
drawn as 50% probability ellipsoids.
The differing trans effects of the diethyl ether and
methyl ligands are reflected in the Pd-P bond distances.
The Pd-P2 bond distances (P2 trans to Et2O) in 2e-g
are ca. 0.16-0.19 Å shorter than the Pd-P1 distance
-
(P1 trans to -CH3), consistent with the fact that CH3
is a much stronger σ-donor ligand than Et2O. As
expected on steric grounds, the ether ligands lie per-
pendicular to the square plane. However, an unusual
feature of ether complexes 2e-g is that the ether ligand
adopts a gauche conformation around one of the C-O
bonds (for example, see the C7-O6 axis in Figure 4).
This is in contrast to the ether ligand in (1,10-
(30) Paviglianiti, A. J .; Minn, D. J .; Fultz, W. C.; Burmeister, J . L.
Inorg. Chim. Acta 1989, 159, 65.
(31) Palenik, G. J .; Mathew, M.; Steffen, W. L.; Beran, G. J . Am.
Chem. Soc. 1975, 97, 1059.

5270 Organometallics, Vol. 20, No. 25, 2001
Ta ble 3. Selected NMR a n d IR Da ta a for (P -P )P d CH3(CO)+BAr ′4
Ledford et al.
-
3a
dppee
3b
3c
3d
3e
3f
3g
P-P
dpbz
dppe
dmpe
1H NMR
dppp
dippp
dppb
b
c
[Pd]-CH3
0.90 (dd, 4, 4)
0.91 (dd, 4, 6)
0.73 (dd, 6, 4)
0.49 (dd, 5, 6)
0.48 (dd, 5, 5)
0.65 (dd, 6, 4)
0.70 (dd, 5, 5)
13C NMR
[Pd]-CH3
[Pd]-COc
1.8 (d, 70)
181.6 (dd,
119, 13)
2.9 (d, 70)
181.7 (dd,
118, 12)
1.5 (d, 70)
182.6 (dd,
148, 11)
-4.8 (d, 74)
182.3 (dd,
122, 13)
4.8 (d, 70)
180.1 (dd,
118, 14)
-3.1 (d, 69)
182.6 (dd,
113, 14)
5.8 (d, 70)
181.1 (br)
31P NMR
(P-P)[Pd]d
67.3 (d, 18)
61.9 (d, 18)
2136
57.1 (d, 30)
48.0 (d, 30)
2134
61.5 (d, 32)
46.5 (d, 32)
2133
41.4 (d, 30)
30.1 (d, 30)
f
16.6 (d, 60)
-7.9 (d, 60)
2132
23.7 (d, 67)
14.4 (d, 67)
2122
25.8 (d, 70)
1.6 (d, 70)
2132
IRe (ν) [Pd]-CO
a
Spectra were collected in CD2Cl2 at -80 °C. Chemical shift values (δ) are relative to CD2Cl2. All couplings (given in parentheses) are
b
3
2
d
2
in Hz. Couplings are J HP
.
c Couplings are J CP
.
Couplings are J PP
.
e CD2Cl2 solutions, given in cm-1
.
f Decomposed.
-
Ta ble 4. Selected NMR a n d IR Da ta a for (P -P )P d COCH3(CO)+BAr ′4
4a 4b 4c 4d 4e
dppee dpbz dppe dppp
4f
dippp
4g
dppb
P-P
dmpe
1H NMR
2.54 (dd, 2, 1)
13C NMR
b
[Pd]-COCH3
1.6 (s)
2.06 (br)
1.97 (s)
1.85 (s)
2.59 (s)
1.84 (s)
c
[Pd]-COCH3
[Pd]-COCH3
[Pd]-COd
42.0 (br)
42.0 (dd,
11, 30)
232.5 (dd,
19, 87)
179.0 (br)
42.1 (dd,
38, 22)
233.3 (br d,
150)
177.2 (br d,
150)
45.1 (dd,
19, 38)
239.4 (dd,
5, 92)
181.0 (br)
42.1 (dd,
24, 25)
232.0 (dd,
9, 85)
175.7 (dd,
20, 83)
45.9 (dd,
15, 36)
235.1 (dd,
6, 74)
177.4 (dd,
20, 79)
42.0 (dd,
21, 38)
229.4 (dd,
6, 84)
175.3 (dd,
22, 80)
d
222.0 (br)
181.6 (dd,
12, 119)
31P NMR
(P-P)[Pd]e
50.8 (d, 34)
46.6 (d, 34)
39.2 (d, 50)
38.5 (d, 50)
41.2 (d, 50)
37.2 (d, 50)
26.8 (d, 47)
19.7 (d, 47)
2.2 (d, 84)
-8.0 (d, 84)
23.7 (d, 67)
14.4 (d, 67)
33.8 (d, 43)
18.7 (d, 43)
IRf (ν)
[Pd]-CO
[Pd]-acyl
2130
1716
2132
1716
2132
1717
g
g
2130
1715
2121
1701
2130
1719
a
Spectra were collected in CD2Cl2 at -80 °C. Chemical shift values (δ) are relative to CD2Cl2. All couplings (given in parentheses) are
in Hz. b Couplings are 4J HP
.
c Couplings are 3J CP
.
Couplings are 2J CP e Couplings are 2J PP f CD2Cl2 solutions, given in cm-1 g Decomposed.
. . .
d
phenanthroline)Pd(CH3)(OEt2)+, in which both C-O
bonds exhibit an anti conformation.32 Examining the
ORTEP diagrams reveals that this conformation likely
results from the pseudoaxial phenyl groups forcing the
methyl groups of the bound ether away from the metal
center and into a gauche position.
3. P r ep a r a tion of Meth yl Ca r bon yl Com p lexes
(P -P )P d (CH3)(CO)+(BAr ′4)- (3a -g) a n d Kin etics of
Th eir Migr a tor y In ser tion Rea ction s. Purging CH2-
Cl2 solutions of 2a -g with CO at -90 °C results in
immediate and quantitative conversion to 3a -g (eq 5).
1H NMR resonances of the Pd-CH3 groups appear
between 0.5 and 1.0 ppm with 13C signals in the -5 to
6 ppm range coupled to the trans 31P by ca. 70 Hz. The
13C{1H} NMR resonances for the bound carbonyl ligand
all appear around 180 ppm, with a large coupling to
trans 31P (113-148 Hz) and a small coupling to cis 31P
(11-14 Hz). In each case two 31P{1H} NMR signals
appear as doublets with the low-field 31P shift assigned
to the phosphorus trans to CO.
When they are treated with various trapping ligands,
L, the methyl carbonyl complexes undergo migratory
insertion reactions to form acyl complexes of the type
(P-P)Pd(L)(COCH3)+. For kinetic studies, CO has been
employed as the trapping ligand, which results in the
quantitative in situ formation of acyl carbonyl complexes
(P-P)Pd(CO)(COCH3)+ (4a -g) (eq 6).
To avoid further reaction to acyl carbonyl complexes,
excess CO was removed by purging each solution with
argon at -90 °C. In the case of 3b-d and 3e, prepara-
tive-scale reactions could be carried out and the methyl
carbonyl complexes isolated in good yields (77-95%) as
thermally unstable solids (see the Supporting Informa-
tion for synthetic details). 1H, 13C, and 31P NMR
spectroscopic analyses unambiguously established for-
mation of the methyl carbonyl complexes (Table 3). All
(32) Rix, F. C.; Brookhart, M.; White, P. S. J . Am. Chem. Soc. 1996,
118, 2436.

Migratory Insertion into Pd(II)-Methyl Bonds
Organometallics, Vol. 20, No. 25, 2001 5271
Ta ble 5. Ra te Da ta for th e Migr a tor y In ser tion
Rea ction s of th e Meth yl Ca r bon yl Com p lexes 3a -g
complex
(ligand)
temp
(K)
correlation
rate constant
∆Gq
(kcal/mol)
coeff (r2)
k1 (104 s-1
)
3a (dppee)
3b (dpbz)
3c (dppe)
3d (dmpe)
3e (dppp)a
3f (dippp)
3g (dppb)
238
232
226
248
191
191
194
0.996
0.998
0.999
0.996
0.998
0.994
0.994
18
17.1(2)
17.1(2)
17.1(1)
18.4(2)
14.8(1)
14.1(2)
14.2(2)
4.23
1.9
3.2
0.45
3.4
5.5
F igu r e 7. Possible transition state structures for reductive
elimination of an alkyl cyanide complex and migratory
insertion of an alkyl carbonyl complex.
Data previously reported.12
a
Table 4 summarizes selected H, 13C, and 31P NMR
1
The barriers measured for the methyl to carbonyl
migratory insertion process clearly indicate that as the
P-Pd-P bond angle of these complexes increases, the
rate of reaction increases. For example, in the series
[Ph2P(CH2)nPPh2]Pd(CH3)(CO)+ involving the dppe (n
) 2), dppp (n ) 3), and dppb (n ) 4) ligands, the barrier
decreases as n increases: 17.1 kcal/mol for n ) 2, 14.8
kcal/mol for n ) 3, and 14.2 kcal/mol for n ) 4. The bond
angles for [Ph2P(CH2)nPPh2]Pd(CH3)(OEt2)+ are 102.4°
for n ) 4, 97.3° for n ) 3, and a bond angle of ∼85° can
be estimated for n ) 2 on the basis of the literature
structures.31,35 The barriers for 3a ,b, containing two-
carbon backbones, are quite similar (17.1 kcal/mol) to
that of 3c, which also has a two-carbon backbone.
Complex 3d , (dmpe)Pd(CH3)(CO),+ exhibits a barrier ca.
1.3 kcal/mol greater than 3a -c. The Pd(II) center for
this electron-rich phosphine is less electrophilic than
that for 3a -c and, in addition, the methyl substituents
are less bulky than the aryl substituents; both factors
would be expected to raise the barrier to migratory
insertion. The dippp complex [(i-Pr2)P(CH2)3P(i-Pr)2]Pd-
(CH3)(CO)+ (3f) exhibits a somewhat lower barrier (14.1
kcal/mol) than 3e (14.8 kcal/mol). On the basis of the
electronic effects, the barrier of 3f would be expected to
be higher than for 3e. The decrease in the barrier is
best attributed to the bulkier nature of the isopropyl
groups, which sterically crowd and destabilize the
ground state more than do the aryl groups. This effect
is dramatically illustrated in the next section, where
such crowding by the isopropyl groups results in a
change in the nature of the catalyst resting state for
ethylene dimerization.
Moloy and co-workers have studied the reductive
elimination reaction of a series of complexes of general
formula (P-P)Pd(CH2TMS)(CN) with various bidentate
phosphine ligands.36 These alkyl cyanide complexes are
isoelectronic with the alkyl carbonyl complexes dis-
cussed here. In addition, the reductive elimination
process should have a transition state very similar to
that of the alkyl to carbonyl migratory insertion process
(Figure 7). Moloy found that an increase in the ligand
bond angle increased the rate of reductive elimination.
Progressing from a small bond angle (∼85° for dppe) to
a large bond angle (∼100° for DIOP) results in a nearly
104-fold increase in the rate of reductive elimination at
80 °C. In both types of processes an increase in ligand
bond angle accompanies an increase in the steric influ-
ence of the ligand. Such increased steric demands
compress the C-Pd-C angle, destabilizing the ground-
data for these acyl carbonyl complexes (CD2Cl2, -30 °C).
1
The H NMR shifts of the methyl groups move signifi-
cantly downfield from the methyl carbonyl complexes
and appear in the range of 1.8-2.6 ppm; 13C NMR shifts
also move downfield to 42-46 ppm. The 13C{1H} NMR
resonances for the acyl carbon appear in the range of
ca. 230-240 ppm, while the carbonyl ligands appear in
the 175-180 ppm range, which is consistent with
similar complexes (see the Supporting Information for
complete spectroscopic details).7,8,13,29,33 The acyl car-
bonyl complex (dppp)Pd(CO)(CH3CO)+BAr′4- (4e) could
be isolated in moderate yields (35%), as we have
previously reported.12 Attempts to isolate other carbonyl
acyl complexes led to decarbonylated products.
Kinetics of conversion of the methyl carbonyl com-
plexes to the acyl carbonyl complexes were conveniently
followed by low-temperature 1H{31P} NMR spectroscopy
in CD2Cl2 using the sharp, well-defined methyl reso-
nances. Rates of conversion are independent of CO
concentration and follow excellent first-order kinetics
(see the Supporting Information for sample plots). Since
CO solubility in CD2Cl2 is low, the concentrations of
3a -g (ca. 4 µmol in 0.7 mL of CD2Cl2) must be kept
low to ensure that conversions to the acyl carbonyl
complexes are complete. First-order rate constants
together with ∆Gq values for reactions are listed in
Table 5. As with the (phen)Pd(CH3)(CO)+ system, the
lack of dependence of the rate on CO concentration
indicates that the migratory insertion step is rate-
determining (k2[CO] > k-1, eq 6) and thus the rate
constants in Table 5 represent the true rate constants
(k1) for migratory insertion of the methyl carbonyl
complexes.7,8
An Eyring analysis for the migratory insertion reac-
tion of (dppp)Pd(Me)(CO)+ (3e) has been previously
carried out and provides activation parameters for this
process of ∆Hq ) 14.8 ( 0.7 kcal/mol and ∆Sq) 0.1 ( 3
eu. The near-zero value of ∆Sq is consistent with the
intramolecular nature of the migratory insertion, with
no assistance from the solvent or additional equivalents
of carbon monoxide.34 That ∆Sq is ca. 0 implies that
there is little temperature dependence of ∆Gq and thus
over the limited temperature ranges given in Table 5
the ∆Gq values can be directly compared for the various
methyl carbonyl complexes and, in addition, ∆Hq values
will be close to ∆Gq values.
(33) Kayaki, Y.; Shimizu, I.; Yamamoto, A. Bull. Chem. Soc. J pn.
1997, 70, 917.
(34) Collman, J . P.; Hegedus, L. S.; Norton, J . R.; Finke, R. G.
Principles and Applications of Organotransition Metal Chemistry;
University Science Books: Mill Valley, CA, 1987; Chapters 6, 11, 12.
(35) Steffen, W. L.; Palenik, G. J . Inorg. Chem. 1976, 15, 2432.
(36) Marcone, J . E.; Moloy, K. G. J . Am. Chem. Soc. 1998, 120, 8527.

5272 Organometallics, Vol. 20, No. 25, 2001
Ta ble 6. Selected NMR Da ta a for (P -P )P d CH3(η2-CH2CH2)+BAr ′4
Ledford et al.
-
5a
dppee
5b
5c
5d
5e
5f
5g
P-P
dpbz
dppe
dmpe
1H NMR
dppp
dippp
dppb
b
[Pd]-CH3
0.85 (dd, 4, 6)
4.95 (br)
0.75 (dd, 3, 7)
5.42 (br)
0.61 (dd, 5, 5)
4.98 (br)
0.51 (dd, 5, 5)
5.32 (br)
0.32 (dd, 4,7)
5.19 (br)
0.35 (dd, 3, 6)
5.54 (br)
0.57 (dd, 7, 7)
5.13 (br)
[Pd](η2-CH2CH2)
31P NMR
(P-P)[Pd]d
66.4 (d, 16)
75.8 (d, 16)
51.4 (d, 30)
62.6 (d, 30)
35.5 (br)
34.8 (d, 26)
51.1 (d, 26)
17.6 (d, 56)
-1.6 (d, 56)
37.6 (d, 41)
7.3 (d, 41)
35.3 (d, 55)
13.9 (d, 55)
13.5 (d, 32)
a
Spectra were collected in CD2Cl2 at -80 °C. Chemical shift values (δ) are relative to CD2Cl2. All couplings (given in parentheses) are
b
3
2
d
2
in Hz. Couplings are J HP
.
c Couplings are J CP
.
Couplings are J PP.
Ta ble 7. Ra te Da ta for th e Migr a tor y In ser tion
Rea ction s of th e Meth yl Eth ylen e Com p lexes 5a -g
state structures. The expected result would be an
acceleration in C-C bond formation.
complex
(ligand)
temp
(K)
correlation
rate constant
∆Gq
(kcal/mol)
4. Eth ylen e Dim er iza tion Stu d ies: Syn th esis
a
coeff (r2)
k (104 s-1
)
-
a n d Kin etics of th e (P -P )P d (η2-C2H4)(R)+BAr ′4
5a (dppee)
5b (dpbz)
5c (dppe)
5d (dmpe)
5e (dppp)b
5f (dippp)
5g (dppb)
237
232
231
233
228
229
233
0.991
0.997
0.997
0.996
0.997
0.997
0.995
4.2
8.1
1.9
0.35
4.9
10.2
7.8
17.4(2)
16.8(2)
17.4(1)
18.3(1)
16.6(1)
16.5(2)
16.8(1)
Com p lexes (R ) Me, 5a -e,g; R ) Et, 6a -e,g). The
reaction of ether complexes 2a -e,g with C2H4 in CD2-
Cl2 at -78 °C quantitatively generates the (P-P)PdMe-
-
(η2-C2H4)+BAr′4 complexes 5a -e,g (eq 7). The treat-
a
The rates were observed with 5-15 equiv of ethylene added;
values are the average of at least two runs at each temperature.
Data previously reported.12
b
the exception of the dippp system (see below), the ethyl
ethylene complex is the catalyst resting state for a
catalytic cycle which dimerizes ethylene to butenes. This
cycle is analogous to the one previously reported for the
phenanthroline-based systems.7,8 Quantitative rate mea-
surements of the migratory insertion reactions of the
methyl ethylene and ethyl ethylene complexes were
made as described below.
The rates of the migratory insertions of the methyl
ethylene complexes 5a -e,g were followed by monitoring
the decrease in the Pd-CH3 1H{31P} NMR resonance.
Clean first-order kinetics were observed, and rates were
independent of the concentration of free ethylene in each
case.37 Kinetic data are summarized in Table 7.
As shown in eq 8, following migratory insertion of the
methyl ethylene species, propylene is released and in
the presence of excess ethylene, ethyl ethylene com-
plexes are formed. NMR data are summarized in Table
8. These complexes are the only Pd species present in
solution and are the catalyst resting states.38The turn-
over frequencies of the catalytic cycle are controlled by
the rates of migratory insertion of these complexes,
which can be easily obtained by measuring the rate of
formation of butenes. These kinetic data are sum-
marized in Table 9.
ment of 2f with ethylene is discussed later. These
species are too unstable to be isolated but were com-
pletely characterized by NMR spectroscopy at low
temperatures. Data are summarized in Table 6.
For spectral characterizations, slightly less than 1
equiv of C2H4 was employed. If more than 1 equiv of
C2H4 was added, exchange between complexed and free
1
C2H4 results in broadened NMR signals. The H NMR
resonances of the Pd-CH3 groups appear in the range
of 0.3-0.8 ppm with a small coupling (4-7 Hz) to both
31P nuclei. The 13C resonances appear in the range of
-2 to 19 ppm and exhibit a large trans J PC (76-91 Hz)
and small to unobservable cis J PC. The 1H NMR
resonances of the η2-ethylene ligands all appear as broad
singlets (possibly due to exchange) between 5.0 and 5.5
ppm; the corresponding 13C resonances are all in the
range 99-110 ppm; in certain cases a small trans J PC
(5-10 Hz) can be observed.
When the methyl ethylene complexes are warmed to
ca. -40 °C in the presence of excess ethylene, the clean
conversion of 5a -e,g to propylene and the ethyl ethyl-
ene complexes 6a -e,g can be monitored (eq 8). With
The ∆Gq values for the methyl ethylene and ethyl
ethylene migratory insertion processes are very similar,
all being within 0.4 kcal/mol for the same ligand.
Collectively the barriers for these alkyl ethylene com-
plexes exhibit much less sensitivity to the P-Pd-P
angle than do the methyl carbonyl complexes. For
example, in the series [(C6H5)2P(CH2)nP(C6H5)2]Pd-
(37) For example, 5e(d p p p ): kinetic runs repeated at 227.6 K with
4, 9, and 19 equiv of ethylene give kobs ) 5.1 × 10-4, 4.8 × 10-4, and
4.9 × 10-4 s-1. [Pd] ) 9.7 mM in each case. [C2H4] ) 39, 87, and 184
mM, respectively. [C2H4] was determined by integration.
(38) The dmpe complex 6d exchanged with propylene at low ethylene
concentrations and 6a -c,e,f exchanged with 1-butene at low ethylene
concentrations. The identity of (dmpe)PdEt(C3H7)+ was confirmed by
adding propylene to a sample of 6d , (dmpe)PdEt(C2H4)+
δ 41.0 (br), 23.0 (br)).
(
31P{1H} NMR

Migratory Insertion into Pd(II)-Methyl Bonds
Organometallics, Vol. 20, No. 25, 2001 5273
Ta ble 8. Selected NMR Da ta a for (P -P )P d CH2CH3(η2-CH2CH2)+BAr ′4
-
6a
dppee
6b
6c
6d
6e
dppp
6g
dppb
P-P
dpbz
dppe
dmpe
1H NMR
0.46 (br)
1.74 (br)
5.52 (br)
[Pd]-CH2CH3
[Pd]-CH2CH3
[Pd](η2-CH2CH2)
0.43 (br)
2.06 (br)
5.50 (br)
0.43 (br)
1.98 (br)
5.42 (br)
0.42 (br)
1.43 (br)
5.14 (br)
0.40 (br)
1.42 (br)
N/O
0.32 (br)
1.91 (br)
5.68 (br)
31P NMR
48.8 (d, 38)
23.8 (d, 38)
(P-P)[Pd]b
75.5 (d, 24)
62.9 (d, 24)
59.0 (br)
44.0 (br)
38.8 (br)
23.0 (br)
17.7 (d, 65)
-2.1 (d, 65)
24.2 (d, 52)
22.8 (d, 52)
a
Spectra were collected in CD2Cl2 at -80 °C. Chemical shift values (δ) are relative to CD2Cl2. All couplings (given in parentheses) are
b
2
in Hz. Couplings are J PP
.
Ta ble 9. Ra tes a n d ∆Gq Va lu es for th e Migr a tor y
In ser tion Rea ction s of Eth yl Eth ylen e Com p lexes
6a -c,e,g
ligand lies perpendicular to the square plane; however,
in the transition state for insertion, the ethylene ligand
must rotate into the square plane and move three (not
two as for aryl carbonyl complexes) carbon atoms within
bonding distance of Pd. The square plane is now
occupied by Ca, Cb, and Cc. This results in little reduc-
tion of steric crowding in the transition state relative
to the ground state and may explain the reduced spread
in barriers for insertions in the alkyl ethylene com-
plexes. The calculated widening in the P1-Pd-Ca bond
angle in the transition sate relative to the ground state
is only 12° (85 to 97°) relative to the change of 27° in
the methyl carbonyl complex.
complex
(ligand)
temp
(K)
correlation
TOFa
∆Gq
(kcal/mol)
coeff (r2)
(103 s-1
)
6a (dppee)
6b (dpbz)
6c (dppe)
6d (dmpe)
6e (dppp)b
6g (dppb)
237
238
244
250
237
238
0.994
0.996
0.994
0.999
0.997
0.993
0.83
1.6
1.6
1.3
5.5
2.5
17.2(2)
16.9(2)
17.3(1)
17.9(2)
16.3(1)
16.6(1)
a
The turnover frequencies were observed with 5-15 equiv of
ethylene added. Data previously reported.12
b
-
5. Gen er a tion of (d ip p p )P d (η2-C2H4)(CH3)+BAr ′4
(5f) a n d (d ip p p )P d (CH2CH2-µ-H)+BAr ′4 (8f) a n d
-
th e Kin etics of Th eir Migr a tor y In ser tion Rea c-
tion s. Analogous to the other systems employed in this
study, the methyl ethylene complex 5f containing the
dippp ligand is cleanly generated by treating the ether
adduct 2f with ethylene at -78 °C. 5f is too unstable to
isolate but was completely characterized by NMR
spectroscopy at low temperatures (Table 6).
When 5f is warmed to ca. -44 °C in the presence of
excess ethylene, the conversion to propylene and the
â-ethyl agostic complex 8f can be monitored (eq 9). Upon
F igu r e 8. Transition state structures calculated by Ziegler
et al.19 for migratory insertions of (H2PCHdCHPH2)Pd(CH2-
CH3)(CO)+ and (H2PCHdCHPH2)Pd(CH2CH3)(CH2CH2)+.
(CH3)(C2H4)+, the difference in ∆Gq values between 5g
(n ) 4, 16.8 kcal/mol) and 5c (n ) 2, 17.4 kcal/mol) is
only ca. 0.6 kcal/mol, whereas the difference is ca. 2.9
kcal/mol between the barriers in the analogous methyl
carbonyl complexes. A similar trend is seen in the ethyl
ethylene complexes 6g and 6c. The barrier to insertion
in 5d , (dmpe)Pd(CH3)(C2H4)+, is somewhat higher than
that observed for the other complexes, no doubt due to
a more electron rich diphosphine with reduced steric
bulk.
The reduced response of the alkyl ethylene insertion
barriers to change in the P-Pd-P bond angle can be
rationalized on the basis of the contrasting ground- and
transition-state structures for the methyl carbonyl and
alkyl ethylene insertion reactions. Ziegler et al.18,19 have
calculated ground- and transition-state structures for
model complexes in each system, employing H2PCHd
CHPH2 as the bidentate phosphine. These transition-
state models are illustrated in Figure 8. For the
migratory insertion of the methyl carbonyl complex the
Ca-Pd-Cb angle is reduced 31° going from the ground
state (86°) to the transition state (55°), while the P1-
Pd-Ca angle widens by 27° (85° to 112°). Thus, sub-
stantial sensitivity of the insertion barrier to the
P-Pd-P bond angle would be expected through relief
of steric crowding in the transition state. In the ground
state of the ethyl ethylene complexes, the ethylene
the initial migratory insertion of 5f a transient species
is generated which disappears as propylene is produced.
We have assigned this transient complex as an agostic
propyl complex, 7f. If the solution is cooled to -120 °C,
1
then a H shift characteristic of an agostic hydrogen is
observed at -2.71 ppm (br d, J PH ) 61 Hz). A detailed
characterization of complex 7f is described below. The
rate of migratory insertion of complex 5f was followed

5274 Organometallics, Vol. 20, No. 25, 2001
Ledford et al.
F igu r e 9. Characterization of n-propyl agostic complex 7f.
by the decrease in the Pd-CH3 1H{31P} NMR resonance.
This process was cleanly first order in 5f and zero order
in free ethylene. Kinetic data are summarized in Table
7.
labeled due to rapid isomerization.23,39 If 7f* is an
agostic n-propyl species, the carbon of the agostic
methylene group (the central carbon) will be 100% 13C
labeled, since the isomerization process does not result
in exchange of this carbon to a new site. The 1H
spectrum of 7f* at -120 °C again showed the resonance
at -2.70 ppm, now appearing to have triplet character.
Decoupling 31P from this spectrum caused this reso-
The dippp system behaves very differently relative
to the other systems examined with regard to ethylene
dimerization. While the initial methyl ethylene complex,
5f, undergoes migratory insertion to produce propylene
at a rate comparable to that for 5e,g, the second phase
of the reaction, catalytic production of butenes, exhibits
very different behavior, specifically (1) the dimerization
reaction occurs at much slower rates than that for the
other complexes examined, (2) the catalyst resting state
is not an ethyl ethylene complex, and (3) the turnover
frequency is now first-order in ethylene concentration.
At 3 °C the second-order rate constant determined from
the production of butenes is 1.58 × 10-3 M-1 s-1. The
catalyst resting state was successfully isolated from a
catalytic reaction (see below) and shown to be the
â-agostic species (dippp)Pd(CH2CH2-µ-H)+ (8f). Thus,
the catalytic cycle shown in eq 8 must be modified to
account for these observations and is illustrated in eq
9. In this system, the bulky isopropyl substituents result
in the less crowded agostic species 8f being more stable
under the reaction conditions than the much more
crowded ethyl ethylene complex 6f, in contrast to all
other systems examined.
1
nance to collapse to a doublet with J CH ) 64 Hz,
confirming the agostic nature of this hydrogen. No
central peak is observed, indicating that Hagostic is bound
exclusively to a labeled carbon atom. This observation
shows that the n-propyl agostic isomer of 7f is the major
isomer (>90%) in solution. This is in contrast to â-agos-
tic (diimine)Pd-propyl+ complexes where the agostic
isopropyl species is favored over the n-propyl agostic
species by ca. 20:1.23,39 See the Supporting Information
for full 13C data of these complexes, which provide
additional support for the proposed structures.
7. Isola tion a n d Dyn a m ic Solu tion Beh a vior of
t h e Agost ic Com p lex (d ip p p )P d (CH 2CH 2-µ-H )+-
-
BAr ′4 (8f). Not only could the agostic complex 8f be
observed in situ during NMR-monitored catalytic dimer-
ization reactions but it could also be isolated as a stable
salt for spectroscopic studies. Purging a CH2Cl2 solution
of 2f with ethylene at 0 °C followed by solvent removal
produced a yellow solid which, after recrystallization,
gave 8f as an analytically pure material. Cooling a
CDCl2F solution of 8f to -143 °C resulted in observation
of a static 1H{31P} NMR spectrum with a resonance at
-2.85 ppm assigned to an agostic hydrogen. Warming
the solution results in rapid broadening of this signal
due to rotation of the agostic methyl group (process I,
Figure 10). A rate constant for methyl rotation can be
estimated as 330 s-1 at -133 °C, corresponding to a ∆Gq
value of ca. 6.4 kcal/mol. Warming to -68 °C results in
coalescence (process I) of the two â-methylene hydrogens
with the agostic hydrogen into a broad triplet (J ) 7
Hz) at 0.05 ppm, integrated for three hydrogens. The
R-methylene hydrogens appear at 1.75 ppm at these
temperatures. Further warming to 5 °C results in
coalescence of these two signals to a broad band at ca.
6. Ch ar acter ization of th e n -P r opyl Agostic Com -
p lex (d ip p p )P d (CH2CH-µ-HCH3)+BAr ′4 (7f). Vari-
-
1
able-temperature H spectra of complex 7f were com-
plicated by the presence of free diethyl ether and the
resonances of the dippp ligand in addition to broadening
that arose from the low temperatures required to obtain
static spectra (ca. -110 °C). Nevertheless, enough data
could be extracted from NMR spectra of 7f and 7f*
(generated using doubly 13C labeled ethylene) to assign
1
a structure to this complex. The H spectrum of 7f at
2
-135 °C showed a resonance at -2.70 ppm (d, J HP
)
61 Hz) that is consistent with an agostic hydrogen. 7f*
was next generated from doubly 13C labeled ethylene.
As shown in Figure 9, two possibilities arise with respect
to the structure and labeling pattern of the propyl group
of 7f*. If 7f* is an agostic isopropyl species, then the
carbon of the agostic methyl group will be 50% 13C
(39) Tempel, D. J .; J ohnson, L. K.; Huff, R. L.; White, P. S.;
Brookhart, M. J . Am. Chem. Soc. 2000, 122, 6686.

Migratory Insertion into Pd(II)-Methyl Bonds
Organometallics, Vol. 20, No. 25, 2001 5275
1
F igu r e 10. Fluxional processes observed for 8f by 13C{1H} and H NMR.
0.7 ppm, a result of the operation of both dynamic
processes I and II shown in Figure 10. These processes
together result in exchange of all five hydrogens of the
agostic ethyl group. The barrier to â-H elimination and
reinsertion (process II) could be estimated from broad-
ening of the averaged agostic methyl group at both -58
and -44 °C as ca. 11(1) kcal/mol.
The activation barrier for â-H elimination and rein-
sertion (process II) could be more accurately determined
using 13C NMR spectroscopy, where the spectra are not
affected by process I. The doubly 13C-labeled agostic
complex 8f*, prepared from 13CH2d13CH2, exhibits two
relatively sharp 13C resonances at δ 27.7 and 6.9 ppm
copolymerization of ethylene and CO have been exam-
ined. The kinetic barriers for the migratory insertion
reactions of the series of complexes [Ph2P(CH2)nPPh2]-
Pd(CH3)(CO)+ (n ) 2-4) decrease with increasing
P-Pd-P bond angle of the complex for the ligands dppe
(n ) 2), dppp (n ) 3), and dppb (n ) 4). The rigidity of
the ligand backbone appears to have little to no effect
on this same process as is seen for the series of ligands
dppe ≈ dpbz ≈ dppee. The steric bulk of the bisphos-
phine ligand had a significant effect on the methyl
carbonyl insertion barriers, with the most bulky ligand
(dippp) having the smallest activation barrier (14.1 kcal/
mol) and the least bulky ligand (dmpe) having the
largest activation barrier (18.4 kcal/mol).
In contrast to the methyl carbonyl case, the alkyl
ethylene migratory insertion reactions experience very
little if any significant increase in rate due to steric
crowding or increase in the P-Pd-P bond angle. Fol-
lowing migratory insertion of (P-P)Pd(CH3)(C2H4)+
(5a -e,g) in the presence of ethylene, (P-P)Pd(CH2CH3)-
(C2H4)+ (6a -e,g) are generated and are catalyst resting
states for the zero-order dimerization of ethylene to
butenes. The (dippp)Pd(CH3)(C2H4)+ complex, 5f, has
one of the lowest barriers to insertion, yet the subse-
quent rate of ethylene dimerization is greatly slowed
due to a now first-order dependence on C2H4 concentra-
tion. The fact that the agostic ethyl complex 8f is now
the catalyst resting state explains the added dependence
on ethylene. That the agostic ethyl complex is favored
over the ethyl ethylene complex is likely due to the large
steric bulk of the dippp ligand. These observations
dramatically illustrate that subtle structural changes
in a transition-metal complex can alter the catalyst
resting state and greatly affect the rate and kinetic
order of a catalytic reaction. The dynamic behavior of
the agostic ethyl complex is notable in that the barriers
to â-hydride elimination (∆Gq ) 10.3(2) kcal/mol, inter-
change of CR, Câ) is substantially higher than the simple
rotation of the agostic methyl group (∆Gq ) ca. 6.4 kcal/
mol).
(-68 °C), assigned to Ca and Cb of 8f*, respectively.
1
Decoupling experiments established that
J
) 157
CaH
Hz, J
) 34 Hz, J
) 35 Hz and J CH(av) ) 123 Hz
PCa
Ca-Cb
for Câ. The averaged J CH of the agostic methyl group of
123 Hz agrees with literature reports of analogous
agostic complexes.40-42 The coalescence temperature of
the two 13C resonances was observed to be -28 °C, and
the rate constant for exchange was determined to be
3500 s-1 at -28 °C, with ∆Gq ) 10.3(2) kcal/mol.
There is substantial precedent for species such as 8f
in the work of Spencer et al., who have observed similar
Pt(II), Pd(II), and Ni(II) agostic ethyl complexes con-
taining bidentate phosphine ligands.40-43 Additionally,
we have recently observed three other ethyl agostic
complexes, two nickel(II) complexes of the general
formula [(P-P)Ni(CH2CH2-µ-H)][X]44,45 and the palla-
dium(II) complex [(N-N)Pd(CH2CH2-µ-H)][X].46 The
complex with the most similar dynamic NMR behavior
to 8f, however, is the ((t-Bu)2P(CH2)3P(t-Bu)2)Pt(CH2-
CH2-µ-H)+ complex.40 The agostic proton resonance was
observed as a broad signal at -2.6 ppm at -118 °C, and
the 13C NMR spectrum at -78 °C indicated CR and Câ
resonances at 28.3 and 6.1 ppm. As in 8f, all five
hydrogens of the agostic ethyl group appear as a broad
singlet at 1.6 ppm at 25 °C due to rapid interchange
via processes I and II.
Con clu sion s
Exp er im en ta l Section
A series of bidentate phosphine Pd(II) complexes that
model the migratory insertion steps involved in the
Gen er a l Meth od s. All reactions, except where indicated,
were carried out in flame-dried glassware under a dry, oxygen-
free argon atmosphere using standard Schlenk and drybox
techniques. Acetone was degassed with argon and dried over
4 Å molecular sieves. All other solvents were either freshly
distilled from sodium benzophenone or deoxygenated and then
dried by passing them through a large column of activated
alumina under nitrogen.47 CD2Cl2 was purchased from Cam-
(40) Conroy-Lewis, F. M.; Mole, L.; Redhouse, A. D.; Litster, A.;
Spencer, J . L. J . Chem. Soc., Chem. Commun. 1991, 1601.
(41) Spencer, J . L.; Mhinzi, G. S. J . Chem. Soc., Dalton Trans. 1995,
3819.
(42) Mole, L.; Spencer, J . L.; Carr, N.; Orpen, A. G. Organometallics
1991, 10, 49.
(43) Carr, N.; Mole, L.; Orpen, A. G.; Spencer, J . L. J . Chem. Soc.,
Dalton Trans. 1992, 2653.
(44) Shultz, C. S.; DeSimone, J . M.; Brookhart, M. Organometallics
2001, 20, 16.
bridge Isotope Laboratories, Inc. and dried over CaH2 or P4O10
.
(45) Shultz, C. S.; White, P. S.; DeSimone, J . M.; Brookhart, M.
Manuscript in preparation.
(46) Shultz, L. H.; Brookhart, M. Organometallics 2001, 20, 3975.
(47) Pangborn, A. B.; Giardello, M. A.; Grubbs, R. H.; Rosen, R. K.;
Timmers, F. J . Organometallics 1996, 15, 1518.

5276 Organometallics, Vol. 20, No. 25, 2001
Ledford et al.
1
The CD2Cl2 was subjected to three freeze-pump-thaw cycles
and vacuum-transferred into glass Schlenk tubes fitted with
high-vacuum Teflon plugs and then stored under Ar. Air-
sensitive complexes were handled in an Ar-filled drybox and
stored under Ar at -30 °C. CP grade CO and ethylene were
purchased from National Welders Supply and used as received.
Ethylene-1,2-13C2, 99% 13C, was purchased from Cambridge
Isotope Laboratories, Inc., and used as received.
For complete details of the kinetics experiments see the
Supporting Information. Kinetics experiments were carried out
under Ar in NMR tubes equipped with septa. CD2Cl2 or
CDCl2F was added to samples at -78 °C, after which solids
were dissolved at the lowest temperature possible. Kinetics
experiments were carried out on a Bruker AMX-300 or Avance
300 spectrometer. NMR probe temperatures were measured
using an anhydrous methanol sample, except for temperatures
below -95 °C, which were determined using a thermocouple.48
1H and 13C chemical shifts were referenced to residual 1H
signals and to the 13C signals of the deuterated solvents,
respectively.
complex 1a (0.198 g, 55%). H NMR (300 MHz, CD2Cl2, -60
°C): δ 7.32-7.51 (m, 22 H, (C6H5)2P(C2H2)P(C6H5)2), 0.25 (m,
6H, Pd(CH3)2). 13C{1H,31P} NMR (75 MHz, CD2Cl2, -60 °C):
δ 147.0 (PCHdCHP), 132.4, 132.3, 130.1, 128.6 ((C6H5)2P-
(C2H2)P(C6H5)2), -0.65 (Pd-CH3). 31P{1H} NMR (121 MHz,
CD2Cl2, -60 °C): δ 50.8. Anal. Calcd for C28H28P2Pd: C, 63.11;
H, 5.30. Found: C, 63.01; H, 5.35.
2. (P -P )P d Me(OEt2)+(BAr ′4)- (2a -f). The preparation of
compound 2a is given as a representative procedure for the
synthesis of the series of 1a -f. The preparation of the complex
containing the dmpe ligand, 1d , via this method was unsuc-
cessful. See the Supporting Information for specific details
regarding the synthesis of compounds 1b,c,e,f.
(d p p ee)P d Me(OEt2)+(BAr ′4)- (2a ). The dimethyl complex
1a (0.125 g, 0.235 mmol) and H(OEt2)2BAr′4 (0.255 g, 0.237
mmol) were suspended in ether (0.25 mL) and CH2Cl2 (0.75
mL) at -30 °C. The reaction mixture was stirred for 2 h,
followed by brief warming to 25 °C in order to dissolve all
solids. The resulting yellow solution was cooled slowly to -78
°C and stored overnight. The resulting precipitate was filtered
and washed with pentane (2 × 3 mL) to afford complex 2a
Compounds 1b, 1c,24,25,49 1d ,24,27 1e,12,26 1f,28 and 1g26 were
synthesized in the same manner as 1a (see the Supporting
Information for details and spectroscopic analysis). Compound
2f 2 was synthesized in the same manner as 2e (see the
Supporting Information for details and spectroscopic analysis).
The syntheses and spectroscopic characterization of com-
pounds 3e and 4e have been previously reported.12 Elemental
analyses were obtained from Oneida Research Services Inc.,
Whitesboro, NY, and Atlantic Microlabs Inc., Norcross, GA.
H(OEt2)2BAr′4,50 1,3-bis(diisopropylphosphino)propane (dip-
pp),51 (TMEDA)Pd(Me)2 (TMEDA ) N,N,N′,N′-tetramethyl-
ethylenediamine),24 and CDCl2F52 were synthesized using
published methods. The following phosphine ligands (>98%
purity) were purchased from either Aldrich or Strem Chemi-
cals and used as received: cis-1,2-bis(diphenylphosphino)-
ethylene (dppee), 1,2-bis(diphenylphosphino)benzene (dpbz),
1,2-bis(diphenylphosphino)ethane (dppe), 1,2-bis(dimethylphos-
phino)ethane (dmpe), 1,3-bis(diphenylphosphino)propane (dppp),
and 1,4-bis(diphenylphosphino)butane (dppb).
1
(0.236 g, 69%) as a yellow powder. H NMR (300 MHz, CD2-
Cl2, -60 °C): δ 7.46-7.62 (m, 20H, (C6H5)2PCHCHP(C6H5)2),
7.32 (br, 1H, PCHCHP), 7.15 (br, 1H, PCHCHP), 3.56 (br, 4H,
3
O(CH2CH3)2), 1.06 (br, 6H, O(CH2CH3)2), 0.63 (d, J HP ) 7.0,
3H, Pd-CH3). 13C{1H,31P} NMR (75 MHz, CD2Cl2, -60 °C): δ
146.4, 143.6 (PCHCHP), 132.6, 132.2, 132.1, 131.6, 129.8,
129.4, 128.1, 127.6 ((C6H5)2PCHCHP(C6H5)2), 70.2 (O(CH2-
CH3)2), 15.5 (O(CH2CH3)2), 12.3 (Pd-CH3). 31P{1H} NMR (121
MHz, CD2Cl2, -60 °C): δ 61.2 (br), 49.2 (br). Anal. Calcd for
C
63H47BF24OP2Pd: C, 52.00; H, 3.26. Found: C, 52.14; H, 3.28.
(d ip p p )P d (CH2CH2-µ-H)+(BAr ′4)- (8f). Complex 2f (0.270
g, 0.202 mmol) was dissolved in CH2Cl2 (1.5 mL) at 0 °C and
purged with ethylene for 10 min. The solvent was removed in
vacuo to afford a yellow solid. This yellow solid was redissolved
in CH2Cl2 (1.0 mL) at 0 °C, and then the solution was slowly
cooled to -78 °C and stored overnight. Colorless crystals of 8f
(0.18 g, 70%) were obtained. 1H{31P} NMR (300 MHz, CDCl2F,
-140 °C; at this temperature the 1H resonances for the ligand,
the R-methylene, and the two nonagostic hydrogens of the
â-methyl group are all very broad and indistinguishable): δ
-2.85 (br, 1H, Pd-CH2CH2-µ-H). 1H{31P} NMR (300 MHz,
CDCl2F, -68 °C): δ 2.07 (sept, 2H, P(CH(CH3)2)2), 1.93 (sept,
The 1H and 13C NMR data attributed to the counterion
-
BAr′4 (Ar′ ) 3,5-(CF3)2C6H3) are consistent for all cationic
complexes examined and are not included in each compound
characterized below. Full spectral details have been previously
reported.53
3
2H, P(CH(CH3)2)2), 1.75 (br q, 2H, J HH ) 7, Pd-CH2), 1.66
(br m, 2H, PCH2CH2CH2P), 1.57 (br m, 2H, PCH2CH2CH2P),
1.13 (br m, 2H, PCH2CH2CH2P), 0.9 (br d, 24H, ((CH3)2-
CH)2PCH2CH2CH2P(CH(CH3)2)2, 0.05 (br t, 3H, 3J HH ) 7, Pd-
CH2CH3). 1H{31P} NMR (300 MHz, CDCl2F, 5 °C): δ 2.07 (sept,
2H, P(CH(CH3)2)2), 1.93 (sept, 2H, P(CH(CH3)2)2), 1.66 (m, 2H,
PCH2CH2CH2P), 1.57 (m, 2H, PCH2CH2CH2P), 1.13 (br m, 2H,
PCH2CH2CH2P), 0.9 (d, 24H, ((CH3)2CH)2PCH2CH2CH2P(CH-
(CH3)2)2), 0.7 (br, 5H, Pd-CH2CH3). 31P{1H} NMR (121 MHz,
P r ep a r a tion s. 1. Dim eth yl Com p lexes 1a -f. The prepa-
ration of compound 1a is given as a representative procedure
for the synthesis of the series of dimethyl complexes 1a -f. See
the Supporting Information for specific details regarding the
synthesis of compounds 1b-f.
(d p p ee)P d Me2 (1a ). A method similar to van Koten’s,24
with only minor variations, was used to prepare all dimethyl
complexes used in this study. A solution of dppee (0.266 g,
0.672 mmol) in acetone (15 mL) was added to a solution of
(TMEDA)PdMe2 (0.169 g, 0.671 mmol) in acetone (5 mL), and
the mixture was stirred for 18 h. The solvent was removed in
vacuo at 0 °C to yield a white precipitate. The precipitate was
washed with hexanes (3 × 5 mL) and dried in vacuo to give
2
2
CD2Cl2, -68 °C): δ 50.5 (d, J PP ) 42), 23.8 (d, J PP ) 42).
Anal. Calcd for C49H51BF24P2Pd: C, 46.16; H, 4.03. Found: C,
46.49; H, 3.97.
Ack n ow led gm en t. This work was supported by the
U.S. Department of Energy and the National Science
Foundation (Grant No. CHE-0107810).
(48) Ammann, C.; Meier, P.; Merbach, A. E. J . Magn. Reson. 1982,
46, 319.
(49) Calvin, G.; Coates, G. E. J . Chem. Soc. 1960, 2008.
(50) Brookhart, M.; Grant, B.; Volpe, A. F., J r. Organometallics 1992,
11, 3920.
(51) Tani, K.; Tanigawa, E.; Tatsuno, Y.; Otsuka, S. J . Organomet.
Chem. 1985, 279, 87.
(52) Siegel, J . S.; Anet, F. A. L. J . Org. Chem. 1988, 53, 2629.
(53) LaPointe, A.; Brookhart, M. Organometallics 1998, 17, 1530.
Su p p or tin g In for m a tion Ava ila ble: Text, figures, and
tables giving full experimental details, sample kinetic plots,
NMR data, and X-ray data. This material is available free of
charge via the Internet at http://pubs.acs.org.
OM010489K