Mechanism of palladium-catalyzed diene cyclization/hydrosilylation: Direct observation of intramolecular carbometalation

Published on Web 04/29/2004

Mechanism of Palladium-Catalyzed Diene Cyclization/

Hydrosilylation: Direct Observation of Intramolecular

Carbometalation

Nicholas S. Perch and Ross A. Widenhoefer*

Contribution from the P. M. Gross Chemical Laboratory, Duke UniVersity,

Durham, North Carolina 27708-0346

Received January 12, 2004; E-mail: rwidenho@chem.duke.edu

Abstract: The results of kinetic, deuterium-labeling, and low-temperature NMR studies have established

a mechanism for the palladium-catalyzed cyclization/hydrosilylation of dimethyl diallylmalonate (1) with

triethylsilane involving rapid, irreversible conversion of the palladium silyl complex [(phen)Pd(SiEt₃)(NCAr)]⁺

[BAr₄]^-[Ar ) 3,5-C₆H₃(CF₃)₂] (4b) and 1 to the palladium 5-hexenyl chelate complex {(phen)Pd[η¹,η²-

CH(CH₂SiEt₃)CH₂C(CO₂Me)₂CH₂CHdCH₂]}⁺[BAr₄]^-(5), followed by intramolecular carbometalation of 5

to form the palladium cyclopentylmethyl complex trans-{(phen)Pd[CH₂CHCH₂C(CO₂Me)₂CH₂CHCH₂SiEt₃]-

(NCAr)}⁺[BAr₄]^-(6), and associative silylation of 6 to release 3 and regenerate 4b.

Introduction

carbocycle that possesses one or more functionalized C-X

bonds that can be manipulated in a subsequent transformation.¹⁰

Functionalized carbocycles are among the most common

structural components of naturally occurring and/or biologically

active molecules and, for this reason, considerable effort has

been directed toward the development of general and efficient

methods for the synthesis of functionalized carbocycles.¹

Transition metal-catalyzed methods have demonstrated particular

utility in the synthesis of functionalized carbocycles due to the

ability of transition metal complexes to facilitate transformations

not possible using traditional approaches and due to the high

levels of selectivity, efficiency, and atom-economy often realized

via transition metal catalysis.²Notable among these transition

metal-catalyzed carbocyclization processes are the cyclization/

addition of enynes,^3,4dienes,^5,6diynes,^7,8or bis(dienes)⁹with a

hydrosilane, hydrostannane, or bimetallic reagent to form a

We have developed a number of effective transition metal-

catalyzed cyclization/addition protocols,^4,6,8including the pal-

ladium-catalyzed cyclization/hydrosilylation of functionalized

1,6-dienes to form silylated cyclopentanes.⁶For example,

reaction of dimethyl diallylmalonate (1) and HSiEt₃catalyzed

by [(phen)Pd(Me)(OEt₂)]⁺[BAr₄]^-[Ar ) 3,5-C₆H₃(CF₃)₂] (2a)

(5 mol %) at 0 °C for 5 min formed silylated cyclopentane 3 in

92% isolated yield with g98% trans selectivity (Scheme 1).¹¹

This protocol displayed good functional group compatibility,

high regio- and diastereoselectivity, and low air- and moisture-

sensitivity and was applicable to the synthesis of cyclohexanes,

fused and tethered polycyclic compounds, and nitrogen hetero-

cycles.⁶A notable extension of this chemistry was the asym-

metric cyclization/hydrosilylation of 1,6-dienes catalyzed by

enantiomerically pure palladium pyridine-oxazoline complexes

to form silylated cyclopentanes with up to 95% ee.⁶

In contrast to our thorough exploration of the scope, limita-

tions, and extensions of palladium-catalyzed diene cyclization/

hydrosilylation,⁶we have generated little information regarding

the mechanism of this transformation, nor has detailed informa-

tion regarding the mechanism of a late transition metal-catalyzed

cyclization/addition process been reported. Although the mech-

anisms of zirconocene-catalyzed diene carboalumination¹²and

(1) (a) Hudlicky, T.; Price, J. D. Chem. ReV. 1989, 89, 1467. (b) Trost, B. M.

Chem. Soc. ReV. 1982, 11, 141.

(2) (a) Ojima, I.; Tzamarioudaki, M.; Li, Z.; Donovan, R. J. Chem. ReV. 1996,

96, 635. (b) Lautens, M.; Klute, W.; Tam, W. Chem. ReV. 1996, 96, 49.

(c) Trost, B. M. Angew. Chem., Int. Ed. Engl. 1995, 34, 259.

(3) (a) Ojima, I.; Donovan, R. J.; Shay, W. R. J. Am. Chem. Soc. 1992, 114,

6580. (b) Muraoka, T.; Matsuda, I.; Itoh, K. Tetrahedron Lett. 1998, 39,

7325. (c) Ojima, I.; Vu, A. T.; Lee, S.-L.; McCullagh, J. V.; Moralee, A.

C.; Fujiwara, M.; Hoang, T. H. J. Am. Chem. Soc. 2002, 124, 9164. (d)

Molander, G. A.; Retsch, W. H. J. Am. Chem. Soc. 1997, 119. 8817.

(4) Chakrapani, H.; Liu, C.; Widenhoefer, R. A. Org. Lett. 2003, 5, 157.

(5) (a) Negishi, E.-i.; Jensen, M. D.; Kondakov, D. Y.; Wang, S. J. Am. Chem.

Soc. 1994, 116, 8404. (b) Shaughnessy, K. H.; Waymouth, R. M. J. Am.

Chem. Soc. 1995, 117, 5873. (c) Molander, G. A.; Hoberg, J. O. J. Am.

Chem. Soc. 1992, 114, 3123. (d) Onozawa, S.; Sakakura, T.; Tanaka, M.

Tetrahedron Lett. 1994, 35, 8177. (e) Molander, G. A.; Nichols, P. J. J.

Am. Chem. Soc. 1995, 117, 4415. (f) Molander, G. A.; Dowdy, E. D.;

Schumann, H. J. Org. Chem. 1998, 63, 3386. (g) Molander, G. A.; Corrette,

C. P. J. Org. Chem. 1999, 64, 9697.

(6) Widenhoefer, R. A. Acc. Chem. Res. 2002, 35, 905.

(7) (a) Tamao, K.; Kobayashi, K.; Ito, Y. J. Am. Chem. Soc. 1989, 111, 6478.

(b) Tamao, K.; Kobayashi, K.; Ito, Y. Synlett 1992, 539.

(8) (a) Madine, J. W.; Wang, X.; Widenhoefer, R. A. Org. Lett. 2001, 3, 385.

(b) Wang, X.; Chakrapani, H.; Madine, J. W.; Keyerleber, M. A.;

Widenhoefer, R. A. J. Org. Chem. 2002, 67, 2778. (c) Liu, C.; Widenhoefer,

R. A. Organometallics 2002, 21, 5666.

(9) (a) Obora, Y.; Tsuji, Y.; Kakehi, Kobayashi, M.; Shinkai, Y.; Ebihara, M.;

Kawamura, T. J. Chem. Soc., Perkin Trans. 1 1995, 599. (b) Takacs, J.

M.; Chandramouli, S. Organometallics 1990, 9, 2877. (c) Takacs, J. M.;

Zhu, J.; Chandramouli, S. J. Am. Chem. Soc. 1992, 114, 773.

(10) (a) These secondary transformations include the oxidation of organosilanes^10b

and organoboranes and the cross-coupling of alkenylsilanes,^10calkenylstan-

nanes,^10dand organoboranes.^10e(b) Jones, G. R.; Landais, Y. Tetrahedron

1996, 52, 7599. (c) Hiyama, T.; Hatanaka, Y. Pure Appl. Chem. 1994, 66,

1471. (d) Mitchell, T. N. Synthesis 1992, 803. (e) Suzuki, A. Pure Appl.

Chem. 1991, 63, 419.

(11) DeCarli, M. A.; Widenhoefer, R. A. J. Am. Chem. Soc. 1998, 120, 3805.

(12) Shaughnessy, K. H.; Waymouth, R. M. Organometallics 1998, 17, 5728.

9

6332

J. AM. CHEM. SOC. 2004, 126, 6332-6346

Pd-Catalyzed Diene Cyclization/Hydrosilylation

A R T I C L E S

Scheme 1

Scheme 2

carbomagnesiation have been reported,¹³these systems are poor

models for synthetically significant late transition metal-

catalyzed cyclization/addition processes, including palladium-

catalyzed diene cyclization/hydrosilylation. Rather, our mecha-

nistic understanding of palladium-catalyzed diene cyclization/

hydrosilylation has been restricted to insights gleaned from our

synthetic investigations⁶or from the mechanistic studies of

related palladium-catalyzed transformations including olefin

hydrosilylation,¹⁴dimerization,¹⁵and copolymerization¹⁶and

diene cycloisomerization.¹⁷Due to the synthetic potential of

palladium-catalyzed diene cyclization/hydrosilylation specifi-

cally, and late transition metal-catalyzed cyclization/addition in

general, we initiated a study directed toward elucidating the

mechanism of palladium-catalyzed diene cyclization/hydrosi-

lylation. Here we report a full account of our mechanistic

investigation of the cyclization/hydrosilylation of 1 and HSiEt₃

to form 3 catalyzed by [(phen)Pd(Me)(NCAr)]⁺[BAr₄]^-[Ar

) 3,5-C₆H₃(CF₃)₂] (2b). This study has produced both the first

detailed mechanism of a late transition metal-catalyzed cycliza-

tion/addition process and the first direct observation of the

â-migratory insertion of a coordinated olefin into the M-C bond

of a transition metal alkyl olefin chelate complex (intramolecular

carbometalation).¹⁸

2b (δ 1.26) and by the appearance of the Pd-SiEt₃resonances

of 4b [δ 0.97 (q), 1.06 (t), J ) 7.5 Hz] and free methane (δ

0.21) in the H NMR spectrum.

1

Formation of Alkyl Olefin Chelate Complex 5. Addition

of 1 (1 equiv) to a solution of 4b (42 mM) at -62 °C led to

rapid (t_1/2e 5 min) formation of the palladium 5-hexenyl chelate

complex {(phen)Pd[η¹,η²-CH(CH₂SiEt₃)CH₂C(CO₂Me)₂CH₂-

1

CHdCH₂]}⁺[BAr₄]^-(5) in 84 ( 10% yield by H NMR

spectroscopy as a single diastereomer (Scheme 2).¹⁹Thermally

Results

sensitive 5 was characterized in solution at -62 °C by NMR

spectroscopy. The H NMR spectrum of 5 displayed a one-

Brookhart has shown that the cationic palladium etherate

complex 2a reacts rapidly with HSiEt₃at -80 °C to form the

palladium silyl silane complex [(phen)Pd(SiEt₃)(HSiEt₃)]⁺

[BAr₄]^-[Ar ) 3,5-C₆H₃(CF₃)₂] (4a), which reacts with simple

olefins to form palladium silyl olefin complexes.¹⁴On the basis

of these results, we initiated our mechanistic investigation of

palladium-catalyzed diene cyclization/hydrosilylation with the

stoichiometric reaction of dimethyl diallylmalonate (1) and the

palladium silyl complex [(phen)Pd(SiEt₃)(NCAr)]⁺[BAr₄]^-[Ar

) 3,5-C₆H₃(CF₃)₂] (4b). Complex 4b was employed in prefer-

ence to 4a to avoid potential complications arising from the

silane ligand of 4a and due to the enhanced thermal stability of

4b relative to that of 4a. Silyl complex 4b was generated in

1

proton multiplet at δ 2.66 (Pd-CH), a one proton doublet of

doublets at δ 1.23 (J ) 3, 13 Hz, -CHHSiEt₃), and a one-

proton triplet at δ 1.11 (J ) 13 Hz, -CHHSiEt₃), which together

established insertion of an olefin of 1 into the Pd-Si bond of

4b. Olefin coordination was supported by the large difference

1

of the H NMR chemical shifts of the olefinic protons of 5 [δ

6.41 (dddd, J ) 4, 8, 9, 16 Hz), 5.44 (d, J_cis) 9 Hz), and 4.22

(d, J_trans) 16 Hz)] relative to the corresponding olefinic

resonances of uncomplexed 1 [δ 5.53, 5.09, and 5.06]. Coor-

dination of the pendant olefin of 5 was further supported by

the large difference of the ¹³C NMR chemical shifts of the

olefinic carbons of the ¹³C-labeled isotopomer {(phen)Pd[η¹,η²-

¹³CH(¹³CH₂SiEt₃)¹³CH₂C(CO₂Me)₂¹³CH₂¹³CHd¹³CH₂]}⁺[BAr₄]^-

(5-¹³C₆) (δ 103.7 and 87.5) relative to the corresponding olefinic

resonances of uncomplexed 1-1,2,3,5,6,7-¹³C₆(δ 131.4 and

119.7) and by the much smaller CdC coupling constant of

5-¹³C₆(J_CdC) 47 Hz) relative to free 1-1,2,3,5,6,7-¹³C₆(J_CdC

) 69 Hz).²⁰

1

quantitative yield (101 ( 10% by H NMR) by treatment of a

CD₂Cl₂solution of 2b (42 mM) with triethylsilane (1 equiv) at

-81 °C for 5 min (Scheme 2). Conversion of 2b to 4b was

established by the disappearance of the Pd-CH₃resonance of

(13) Knight, K. S.; Wang, D.; Waymouth, R. M.; Ziller, J. J. Am. Chem. Soc.

1994, 116, 1845.

(14) LaPointe, A. M.; Rix, F. C.; Brookhart, M. J. Am. Chem. Soc. 1997, 119,

906.

(15) Rix, F. C.; Brookhart, M. J. Am. Chem. Soc. 1995, 117, 1137.

(16) Rix, F. C.; Brookhart, M.; White, P. S. J. Am. Chem. Soc. 1996, 118, 4746.

(17) Goj, L. A.; Widenhoefer, R. A. J. Am. Chem. Soc. 2001, 123, 11290.

(18) A portion of these results have been communicated: Perch, N. S.;

Widenhoefer, R. A. Organometallics 2001, 20, 5251.

(19) Neither formation of byproducts nor decomposition was observed in any

of these transformations. The formation of 6 in 93% yield from 3 suggests

that the yield for the conversion of 2 to 4 was higher than indicated by ¹H

NMR.

(20) Benn, R.; Rufin˜ska, A. J. Organomet. Chem. 1982, 238, C27.

9

J. AM. CHEM. SOC. VOL. 126, NO. 20, 2004 6333

A R T I C L E S

Perch and Widenhoefer

Figure 1. ¹H NOESY spectrum of 5. Cross-peaks corresponding to through-space interactions between the ortho phenanthroline proton at δ 8.35 with the

internal (a) and trans-terminal olefinic protons (b), and between the ortho phenanthroline proton at δ 8.71 with the Pd-CH methine (c), and the -CHHSiEt₃

methylene protons (d) are denoted.

Low-temperature ¹H-¹H NOESY analysis of 5 revealed

through-space interactions between the ortho phenanthroline

proton at δ 8.35 with both the internal (δ 6.41) and trans-

terminal olefinic protons (δ 5.44) (Figure 1). These interactions

established orientation of the olefin approximately perpendicular

to the coordination plane, as is typically observed with non-

chelated square planar Pd(II) and Pt(II) olefin complexes.²¹

NOESY analysis of 5 also revealed through-space interactions

between the ortho phenanthroline proton at δ 8.71 with both

the methine proton of the palladium-bound alkyl group (δ 2.66)

and with one of the diastereotopic methylene protons of the

exocyclic triethylsilylmethyl group (δ 1.11) (Figure 1). Analysis

of molecular models indicated that interaction of the ortho

phenanthroline proton with both the Pd-CH and -CHHSiEt₃

protons can be achieved only if the triethylsilylmethyl group

adopts an axial or pseudoaxial position with respect to the

hexenyl chelate. Although this result was initially somewhat

surprising, further analysis of molecular models revealed that

in an equatorial or pseudoequatorial position, the triethylsilyl-

methyl group experiences a pronounced, unfavorable interaction

with the phenanthroline ligand. Given the axial orientation of

the triethylsilylmethyl group, 5 likely adopts a boatlike con-

formation to avoid unfavorable 1,3-diaxial interaction between

the triethylsilylmethyl group and one of the carbomethoxy

groups.²²

To probe for reversible formation of 5, a solution of 5 (25

mM) that contained NCAr (25 mM) was treated with 4,4-

dicarbomethoxy-2,6-dideuterio-1,7-heptadiene (1-2,6-d₂) (50

1

mM) and monitored periodically by H NMR spectroscopy.

Formation of neither free 1 nor 5-d₂was detected after 90 min

at -80 °C and 30 min at -60 °C, which established irreversible

conversion of 4b to 5 under these conditions. In a similar

manner, ¹H NMR analysis of solutions of 5 that contained either

NCAr (42 mM) or both NCAr (42 mM) and 1 (42 mM) at -62

°C revealed no evidence for displacement of the chelated olefin

of 5 to form the 5-hexenyl species 5-L (L ) NCAr, 1) (Scheme

3). The failure to form detectable quantities of 5-L under these

(22) We cannot rule out a structure for 5 in which the R-(triethylsilyl)methyl

group and the terminus of the complexed olefin have a cis relationship.

However, if this were the case, cis to trans isomerization must precede

conversion of 5 to 6, and this isomerization must be fast relative to the

conversion of 5 to 6 as the rate of conversion of 5 to 6 was independent of

[NCAr].

(21) Albright, T. A.; Hoffmann, R.; Thibeault, J. C.; Thorn, D. L. J. Am. Chem.

Soc. 1979, 101, 3801 and ref 23 therein.

9

6334 J. AM. CHEM. SOC. VOL. 126, NO. 20, 2004

Pd-Catalyzed Diene Cyclization/Hydrosilylation

A R T I C L E S

Scheme 3

Table 1. First-Order Rate Constants for the Conversion of 5 ([5]₀

) 42 mM) to 6 in CD₂Cl₂as a Function of Temperature

entry

temp (°C)

(10⁴)k_5f6(s^-1

)

1

2

3

4

5

6

7

-62

-55

-51

-41

-30

0.288 ( 0.022

0.711 ( 0.002

1.41 ( 0.01

5.04 ( 0.05

4.35 ( 0.03

6.33 ( 0.04

27.2 ( 0.5

Figure 2. First-order plot for the conversion of 5 ([5]₀) 42 mM) to 6 at

-41 °C in CD₂Cl₂that contained NCAr (42 mM).

conditions is not surprising, given the likelihood of a large

negative entropy of reaction for the conversion of 5 to 5-L,^16,23,24

and does not rule out rapid and reversible formation of 5-L. To

the contrary, given the extreme facility of the associative

exchange of ethylene and R-olefins at cationic square planar

palladium complexes,^14-16,25,26reversible formation of 5-L from

5 in the presence of 1 or NCAr appears likely. Because the

relative stereochemistry of 5 is lost upon olefin displacement,

the diastereoselective formation of 5 is likely under thermody-

namic control.

Figure 3. Eyring plot for the conversion of 5 to 6 over the temperature

range -30 to -62 °C.

Formation of Palladium Cyclopentylmethyl Complex 6.

Warming a solution of palladium alkyl olefin chelate complex

5 at -41 °C for 2 h led to â-migratory insertion and formation

of the palladium cyclopentylmethyl complex trans-{(phen)Pd-

constants for the conversion of 5 to 6 were determined as a

function of temperature from -30 to -62 °C (Table 1). An

Eyring plot of these data provided the activation parameters

for the conversion of 5 to 6 (Figure 3): ∆H^q) 13.5 ( 0.6 kcal

mol^-1and ∆S^q) -15 ( 2 eu.

Conversion of 6 to Palladium Carbonyl Chelate Complex

7. Warming a solution of palladium cyclopentylmethyl complex

6 at -9 °C for 2 h in the absence of silane led to rearrangement

and formation of the thermally stable cyclopentyl carbonyl

[CH₂CHCH₂C(CO₂Me)₂CH₂CHCH₂SiEt₃](NCAr)}⁺[BAr₄]^-

1

(6) in 96 ( 10% yield by H NMR spectroscopy as a single

diastereomer (Scheme 2). Thermally sensitive 6 was character-

ized in solution by ¹H- and ¹³C NMR spectroscopy at -41 °C.

The ¹H NMR spectrum of 6 displayed two one-proton multiplets

at δ 1.67 and 1.80, assigned to the cyclopentyl methine protons,

and two one-proton doublets of doublets at δ 2.09 (J ) 8, 11

Hz) and 2.52 (J ) 4, 8 Hz), assigned to the palladium-bound

methylene group, which together established â-migratory inser-

tion and cyclopentyl ring formation. The trans stereochemistry

of 6 was established by the exclusive formation of trans-3 from

reaction of 6 with HSiEt₃, given the likelihood that conversion

of 5 to 6 is irreversible under reaction conditions (see below).

chelate complex trans,trans-{(phen)Pd[CHCH(Me)CH(CH₂-

SiEt₃)CH₂C(COOMe)(COOMe)]}⁺[BAr₄]^-(7) in 110 ( 10%

yield (93 ( 10% from 4b) by ¹H NMR spectroscopy as a single

diastereomer (Scheme 2).¹⁹Complex 7 was subsequently

isolated in 49% yield from the preparative-scale reaction of an

equimolar mixture of [(phen)Pd(Me)(NCCH₃)]⁺[BAr₄]^-(2c),

1, and HSiEt₃and was characterized by spectroscopy and

Disappearance of 5 (∼42 mM) at -41 °C in the presence of

NCAr (∼42 mM) obeyed first-order kinetics to >3 half-lives

with a rate constant of k ) 5.2 ( 0.8 × 10^-4s^-1; ∆G^q

)

232 K

16.9 ( 0.1 kcal mol^-1, as the average of three separate

experiments (Table 1, entries 4-6; Figure 2). Because NCAr

was completely consumed during the conversion of 5 to 6, the

first-order decay of 5 established the zero-order dependence of

the rate of conversion of 5 to 6 on [NCAr]. First-order rate

1

elemental analysis. The H NMR spectrum of 7 displayed a

one-proton doublet at δ 2.47 (J ) 10.5 Hz, Pd-CH) and a three-

proton doublet at δ 1.20 (J ) 6.5 Hz, CH-CH₃), which together

established migration of the palladium atom from the exocyclic

methylene carbon to the C(2) carbon atom of the cyclopentyl

ring. The relative stereochemistry of 7 was established by

degradation with DSiEt₃(see below) and by comparison of the

¹H NMR spectrum of 7 to that of the structurally characterized

(23) Johnson, L. K.; Mecking, S.; Brookhart, M. J. Am. Chem. Soc. 1996, 118,

267.

(24) Mecking, S.; Johnson, L. K.; Wang, L.; Brookhart, M. J. Am. Chem. Soc.

1998, 120, 888.

(25) Johnson, L. K.; Killian, C. M.; Brookhart, M. J. Am. Chem. Soc. 1995,

117, 6414.

(26) Rix, F. C.; Brookhart, M.; White, P. S. J. Am. Chem. Soc. 1996, 118, 2436.

analogue

trans,trans-{(phen)Pd[CHCH(Me)CH(Et)CH₂C-

9

J. AM. CHEM. SOC. VOL. 126, NO. 20, 2004 6335

A R T I C L E S

Perch and Widenhoefer

Scheme 4

Table 2. Silane Concentration Dependence of the Rate of

Conversion of 7 ([7]₀) 20-36 mM) to 3 at -14 °C in CD₂Cl₂

entry

[HSiEt₃] (M)

(10⁴)k_obs(s^-1

)

1

2

3

4

5

6

7

0.18

0.20

0.36

0.38

0.93

0.59 ( 0.01

0.50 ( 0.01

1.33 ( 0.01

1.45 ( 0.02

2.94 ( 0.05

3.0 ( 0.2

2.38 ( 0.03

(COOMe)(COOMe)]}⁺[BAr₄]^-(7a).¹⁷Conversion of palladium

cyclopentylmethyl complex 6 to palladium carbonyl chelate

complex 7 did not obey first-order kinetics (Figure S1),

presumably due to the increasing concentration of free NCAr

with increasing conversion.

Silylation of Palladium Alkyl Complexes. Reaction of

palladium cyclopentylmethyl complex 6 (32 mM) with trieth-

ylsilane (50 mM) at -41 °C for 10 min led to complete (g95%)

consumption of 6 to form a 1:1 mixture of the silylated

carbocycle 3 and palladium silyl complex 4b as the exclusive

products in 90 ( 10% yield by ¹H NMR spectroscopy (Scheme

2). Assuming that silylation of 6 obeyed a second-order rate

law, as was established for the silylation of 7 (see below), we

estimate a lower limit for the second-order rate constant for the

silylation of 6 of k g 0.19 M^-1s^-1at -41 °C (∆G^qe 14.2

kcal mol^-1).²⁷Although the extreme facility of this transforma-

tion precluded detailed kinetic analysis, silylation of 6 was

inhibited by excess nitrile, and reaction of 6 (40 mM) and HSiEt₃

(50 mM) in the presence of CD₃CN (0.5 M) required g30 min

to reach completion at 0 °C. Reaction of 6 with DSiEt₃at -41

°C formed 1,1-dicarbomethoxy-3-deuteriomethyl-4-triethysilyl-

methylcyclopentane (3-d₁) as the exclusive isotopomer as

determined by ¹³C NMR and GC/MS analysis (Scheme 4).

Selective incorporation of deuterium into the exocyclic methyl

group of 3-d₁was established by the 1:1:1 triplet at δ 17.1 (J

) 19.1 Hz, isotopic shift ) 300 ppb), corresponding to the

exocyclic -CH₂D group in the ¹³C NMR spectrum. Noteworthy

is that catalytic cyclization/deuteriosilylation of 1 with DSiEt₃

also formed 3-d₁as the exclusive isotopomer (Scheme 4).

Treatment of palladium alkyl olefin chelate complex 5 with

HSiEt₃formed none of the product resulting from silylation of

the Pd-C bond of 5 but instead formed silylated carbocycle 3

as the exclusive organic product (Scheme 2). This result is in

accord with the failure to form significant amounts (e2%) of

silylated uncyclized products in the palladium-catalyzed cy-

clization/hydrosilylation of 1 and HSiEt₃.¹¹Reaction of 5 (43

mM) with HSiEt₃(128 mM) to form 3 at -51 °C obeyed first-

order kinetics to >2 half-lives with a rate constant of k ) 1.59

( 0.01 × 10^-4s^-1(Figure S2), which does not differ

significantly from the first-order rate constant for the conversion

of 5 to 6 at -51 °C (k ) 1.41 ( 0.01 × 10^-4s^-1) (Table 1,

Figure 4. Plot of k_obsversus silane concentration for the reaction of 7

([7]₀) 20-36 mM) and HSiEt₃in CD₂Cl₂at -14 °C.

Scheme 5

entry 3). These observations are consistent with conversion of

5 to 3 via rate-limiting conversion of 5 to 6 followed by rapid

silylation of 6 to give 3.

In contrast to the facile silylation of cyclopentylmethyl

complex 6, reaction of palladium carbonyl chelate complex 7

(30 mM) with HSiEt₃(40 mM) to form 3 as the exclusive

organic product required 2 h at room temperature (Scheme 2).

Conversion of 7 (36 mM) to 3 at -14 °C in the presence of a

large excess of HSiEt₃(0.36 M) obeyed pseudo-first-order

kinetics through g4 half-lives with an observed rate constant

of k_obs) 1.33 ( 0.01 × 10^-4s^-1(Table 2, entry 3, Figure S3).

To determine the dependence of the rate of the silylation of 7

on silane concentration, pseudo-first-order rate constants for

reaction of 7 and HSiEt₃at -14 °C were determined as a

function of silane concentration from 0.18 to 0.93 M (Table 2).

A linear plot of k_obsversus [HSiEt₃] established the first-order

dependence of the rate of the conversion of 7 to 3 on silane

concentration and the second-order rate law: rate ) k[7][HSiEt₃],

where k ) 3.3 ( 0.3 × 10^-4M^-1s^-1at 22 °C (∆G^q

)

259K

22.0 ( 0.1 kcal mol^-1) (Figure 4).

Treatment of 7 with DSiEt₃(0.42 M) at room temperature

for 2 h led to formation of trans,trans-1,1-dicarbomethoxy-2-

deuterio-4-(triethylsilylmethyl)-3-methylcyclopentane (3-d₁′) in

46% yield with 83% isotopic purity as determined by GC/MS

analysis (Scheme 5). The regiochemistry of 3-d₁′ was established

by the 1:1:1 triplet at δ 42.0 (J_CD) 20.6 Hz, isotopic shift )

320 ppb), assigned to the C(2) carbon atom of the cyclopentane

(27) (a) The rate constant k for the silylation of 6 was calculated for the rate

law: rate ) k[6][HSiEt₃] employing the following equation: kt )

{1/([HSiEt₃]₀- [6]₀)} × ln{([6]₀× [HSiEt₃]_t)/([HSiEt₃]₀× [6]_t)}, where

t ) 600 s, [6]₀) 0.032 M, [HSiEt₃]₀) 0.050 M, [6]_t) 0.0016 M (95%

conversion) [HSiEt₃]_t) 0.0196 M (95% conversion).^27b(b) Frost, A. A.;

Pearson, R. G. Kinetics and Mechanism; Wiley: New York, 1961; p 16.

9

6336 J. AM. CHEM. SOC. VOL. 126, NO. 20, 2004

Pd-Catalyzed Diene Cyclization/Hydrosilylation

A R T I C L E S

Table 3. Ratio of Silylated Cyclopentanes Formed from Reaction

of 1 with a Mixture of HSiEt₃(5 equiv) and HSiMe₂R (R ) Et,

n-octyl, OSiMe₃, and Ph) (5 equiv) (entries 1-4) or with a Mixture

of HSiMe₂Ph (5 equiv) and HSiMe₂(4-C₆H₄R) (R ) NMe₂, OMe, F,

and CF₃) (5 equiv) (entries 5-8) Catalyzed by 2b (5 mol %) in

DCE at Room Temperature

carbocycle

entry

base silane

HSiMe₂R

−SiMe₂R

3:3a−3d

3d:3e−3h

1

2

3

4

5

6

7

8

HSiEt₃

HSiMe₂Ph

Et

n-octyl

OSiMe₃

Ph

4-C₆H₄NMe₂

4-C₆H₄OMe

4-C₆H₄F

4-C₆H₄CF₃

3a

3b

3c

3d

3e

3f

1:5.9

1:4.1

1:48

1:15

-

1:3.2

1:1.7

2.7:1

6.1:1

-

3g

3h

-

Figure 5. Plot of log 3d:3e-3h formed in the reaction of 1 with a 1:1

mixture of HSiMe₂Ph (3d) and HSiMe₂(4-C₆H₄R) [R ) NMe₂(3e), OMe

(3f), F (3g), and CF₃(3h)] catalyzed by 2b versus the Hammet σ-param-

eter.²⁸

ring in the ¹³C NMR spectrum. The relative stereochemistry of

3-d₁′ was established by the absence of the doublet of doublets

1

at δ 2.48 in the H NMR spectrum of 3-d₁′ that had been

previously assigned to the methylene proton trans to the methyl

1

group via two-dimensional H-¹H COSY and NOESY spec-

troscopy of unlabeled 3.

Two sets of experiments were performed to probe the effect

of silane structure on the silyl-incorporation step in the catalytic

cyclization/hydrosilylation of 1. In one set of experiments, 1

was treated with a catalytic amount of 2b and an excess of an

equimolar mixture of HSiEt₃(5 equiv) and HSiMe₂R (R ) Et,

n-octyl, OSiMe₃, or Ph) (5 equiv) in 1,2-dichloroethane (DCE)

to form a mixture of carbocycles 3 and 3a, 3b, 3c, or 3d,

respectively (Table 3, entries 1-4). These data revealed that

the extent of silane incorporation increased with both the

decreasing steric bulk and increasing electron density of the

silane. As examples, palladium-catalyzed reaction of 1 with a

1:1 mixture of HSiEt₃and HSiMe₂Et formed a 1:5.9 mixture

of carbocycles 3:3a (Table 3, entry 1), while reaction of 1 with

a 1:1 mixture of HSiEt₃and the small, electron-rich pentam-

ethyldisiloxane formed a 1:48 mixture of carbocycles 3:3c

(Table 3, entry 3). To better quantify the effect of the electron

density of the silane on the silyl-incorporation step in the

catalytic cyclization/hydrosilylation of 1, 1 was treated with a

catalytic amount of 2b and a mixture of HSiMe₂Ph (5 equiv)

and HSiMe₂(4-C₆H₄R) (R ) NMe₂, OMe, F, or CF₃) (5 equiv)

to form a mixture of carbocycles 3d and 3e, 3f, 3g, or 3h,

respectively (Table 3, entries 5-8). The corresponding plot of

log 3d:3e-3h versus the Hammett σ-parameter was roughly

linear with a slope of F ) -1.1 (Figure 5).²⁸

Figure 6. Concentration versus time plots for the reaction of 1 ([1]₀) 85

mM) with HSiEt₃catalyzed by 2b ([2b]₀) 12 mM) at -41 °C in CD₂Cl₂

that contained NCAr (∼42 mM) where [HSiEt₃]₀) 0.16 M (4) and 90

mM (O).

Table 4. Observed Rate Constants for the Cyclization/

Hydrosilylation of 1 ([1]₀) 85 mM) and HSiEt₃([HSiEt₃]₀) 0.16

M) Catalyzed by 2b in CD₂Cl₂as a Function of Temperature, [2b],

and [NCAr]

entry

[2b] (mM)

temp (°C)

[NCAr] (mM)

(10⁶) k_obs(M s^-1

)

1

2

3

4

5

6

7

8

9

10

11

12

13

12

13

12

6.0

6.2

24

25

12

13

-41

-25

-31

-51

-57

42

12

74

42

8.84 ( 0.09

8.20 ( 0.04

8.0 ( 0.2

7.42 ( 0.03

8.9 ( 0.1

3.93 ( 0.06

2.36 ( 0.02

19.5 ( 0.5

18.6 ( 0.3

26 ( 2

21.7 ( 0.3

1.08 ( 0.01

0.58 ( 0.01

Kinetics of Catalytic Cyclization/Hydrosilylation. We

sought to establish the resting state and rate behavior of the

catalytic cyclization/hydrosilylation of 1 and HSiEt₃under

conditions that approximated the relative and absolute concen-

trations of diene and silane employed in preparative-scale

reactions. To this end, a solution of 1 ([1]₀) 85 mM), HSiEt₃

([HSiEt₃]₀) 0.16 M),²⁹NCAr ([NCAr] ) 42 mM), and a

catalytic amount of 2b ([2b]₀) 12 mM) in CD₂Cl₂was

Throughout complete conversion of 1 to 3, alkyl olefin chelate

complex 5 was the only palladium species detected by ¹H NMR

spectroscopy. A plot of [1] versus time was linear to 3 half-

lives with a pseudo-zero-order rate constant of k_obs) 8.3 (

0.6 × 10^-6M s^-1, determined as the average of three separate

experiments (Figure 6, Table 4, entries 1-3).²⁹The linear decay

of [1] versus time established the zero-order dependence of the

rate of conversion of 1 to 3 on [1] over the range 85-∼6 mM

and on [HSiEt₃] over the range 0.16-∼0.08 M. However,

saturation kinetics were not maintained at lower silane concen-

tration, and a plot of [1] versus time for reaction of 1 ([1]₀)

1

monitored periodically by H NMR spectroscopy at -41 °C.

(28) (a) Exner, O. in Correlation Analysis in Chemistry; Recent AdVances;

Chapman, N. B., Shorter, J., Eds.; Plenum: New York, 1978; pp 439-

540. (b) Matsui, K.; Hepler, Can. J. Chem. 1974, 52, 2906.

(29) The principle source of error in our kinetic measurements was determination

of catalyst concentration, which stemmed from the difficulty in accurately

weighing 2b into the NMR tube. Efforts to improve the accuracy of these

measurements by employing stock solutions of 2b were unsuccessful due

to the short lifetime of 2b in solution at ambient temperature.

9

J. AM. CHEM. SOC. VOL. 126, NO. 20, 2004 6337

A R T I C L E S

Perch and Widenhoefer

Figure 8. Eyring plot for the reaction of 1 ([1]₀) 85 mM) and HSiEt₃

([HSiEt₃]₀) 0.16 M) catalyzed by 2b ([2b]₀) 12 mM) over the temperature

range -57 to -25 °C.

Figure 7. Plot of k_obsversus catalyst loading for the reaction of 1 ([1]₀)

85 mM) and HSiEt₃([HSiEt₃]₀) 0.16 M) catalyzed by 2b ([2b]₀) 12

mM) to form 3 at -41 °C in CD₂Cl₂.

capture with NCAr would form the observed palladium cyclo-

pentylmethyl complex 6. Associative silylation of 6 via the

palladium-silane intermediate IV would release carbocycle 3

and regenerate the palladium silyl complex 4b. Because cationic,

coordinatively unsaturated palladium,^30-32rhodium,³³and co-

balt^34,35alkyl olefin complexes are stabilized by dynamic

â-agostic interactions, coordinately unsaturated complexes II

and III are also likely stabilized by â-agostic interactions.

Conversion of 4b to 5. The failure of alkyl olefin chelate

complex 5 to exchange with 1-2,6-d₂at -60 °C established

irreversible conversion of 4b and 1 to 5, which requires

irreversibility in one or more of the microscopic steps in the

conversion of 4b to 5. Because the nitrile ligand of catalyst

precursor 2b exchanges rapidly with a double bond of 1 at -80

°C (K_eq≈ 0.25),³⁶it is likely that conversion of 4b to I is also

rapid and reversible. Although â-migratory insertion of ethylene

into the Pd-Si bond of the cationic palladium silyl ethylene

complex [(phen)Pd(SiEt₃)(H₂CdCH₂)]⁺is reversible at -60

°C,¹⁴we propose that rapid, exothermic coordination of the

pendant olefin of II to form 5 renders both the conversion of I

to II and the conversion of II to 5 irreversible. We estimate

that conversion of II to 5 is exothermic by ∼12 kcal mol^-1on

the basis of DFT calculations for displacement of the â-agostic

interaction of the palladium ethyl complex [(H₂PCHdCHPH₂)-

PdCH₂CH₃]⁺(8) with ethylene to form [(H₂PCHdCHPH₂)Pd-

(H₂CdCH₂)(CH₂CH₃)]⁺(9).³⁷Irreversible silylpalladation of

1 with 4b is further supported by the irreversible hydropalla-

dation of 1 with a cationic palladium hydride complex.¹⁷

Alkyl Olefin Chelate Complexes. â-Migratory insertion of

the coordinated olefin into the M-C bond of a transition metal

alkyl olefin chelate complex is often invoked as the key C-C

bond-forming process in transition metal-mediated and -cata-

lyzed cyclization protocols.²Although â-migratory insertion of

85 mM) and HSiEt₃([HSiEt₃]₀) 90 mM) catalyzed by 2b

([2b]₀) 12 mM) at -41 °C displayed pronounced positive

curvature (Figure 6).²⁹Point-by-point analysis of concentration

versus time plots for catalytic cyclization/hydrosilylation of 1

([1]₀) 85 mM) at low silane concentration ([HSiEt₃]₀) 90

mM) indicated that the reaction rate decreased ∼50% as the

silane concentration decreased from 90 to ∼5 mM (95%

conversion).

The rate of palladium-catalyzed cyclization/hydrosilylation

under saturation conditions was independent of nitrile concen-

tration over the range of 12-74 mM (Table 4, entries 1-5).

To determine the dependence of the rate of catalytic cyclization/

hydrosilylation on catalyst concentration, pseudo-zero-order rate

constants for reaction of 1 and HSiEt₃([HSiEt₃]₀) 0.16 M)

catalyzed by 2b at -41 °C were determined as a function of

precatalyst concentration (Table 4, entries 1-3, 6-9). A linear

plot of k_obsvs [2b]₀over the range of 0-25 mM established

the first-order dependence of the rate of catalytic cyclization/

hydrosilylation on precatalyst concentration and overall the first-

order rate law under saturation conditions: rate ) k_sat[2b], where

k_sat) 7.74 ( 0.06 × 10^-4s^-1; ∆G^q) 16.8 ( 0.1 kcal mol^-1

(Figure 7). Pseudo-zero-order rate constants for catalytic cy-

clization/hydrosilylation of 1 and HSiEt₃([HSiEt₃]₀) 0.16 M)

were measured as a function of temperature from -57 to -25

°C (Table 4, entries 1-3, and 10-13). An Eyring plot of the

corresponding first-order rate constants provided the activation

parameters: ∆H^q) 13 ( 1 kcal mol^-1and ∆S^q) -15 ( 3 eu

(Figure 8).

Discussion

Mechanism of Cyclization/Hydrosilylation. The mechanism

for the cyclization/hydrosilylation of 1 and HSiEt₃catalyzed

by 2b to form 3 depicted in Scheme 6 is consistent with all of

our experimental observations. Silylation of precatalyst 2b with

triethylsilane forms the observed palladium silyl complex 4b.

Displacement of the nitrile ligand of 4b with one of the double

bonds of 1 would form the unobserved palladium silyl olefin

species I. â-Migratory insertion of the coordinated olefin into

the Pd-Si bond of I to form the coordinatively unsaturated

palladium alkyl intermediate II followed by coordination of the

pendant olefin of II would form the observed alkyl olefin chelate

complex 5. â-Migratory insertion of the coordinated olefin into

the Pd-C bond of 5 to generate the coordinately unsaturated

palladium cyclopentylmethyl complex III followed by ligand

(30) Temple, D. J.; Johnson, L. K.; Huff, R. L.; White, P. S.; Brookhart, M. J.

Am. Chem. Soc. 2000, 122, 6686.

(31) Shultz, L. H.; Temple, D. J.; Brookhart, M. J. Am. Chem. Soc. 2001, 123,

11539.

(32) (a) Shultz, L. H.; Brookhart, M. Organometallics 2001, 20, 3975. (b)

Temple, D. J.; Brookhart, M. Organometallics 1998, 17, 2290.

(33) Hauptman, E.; Sabo-Etienne, S.; White, P. S.; Brookhart, M.; Garner, J.

M.; Fagan, P. J.; Calabrese, J. C. J. Am. Chem. Soc. 1994, 116, 8038.

(34) Brookhart, M.; Volpe, A. F.; Lincoln, D. M.; Horvath, I. T.; Millar, J. M.

J. Am. Chem. Soc. 1990, 112, 5634.

(35) (a) Brookhart, M.; Lincoln, D. M.; Bennett, M. A.; Pelling, S. J. Am. Chem.

Soc. 1990, 112, 2691. (b) Brookhart, M.; Grant, B. E. J. Am. Chem. Soc.

1993, 115, 2151.

(36) Perch, N. S.; Widenhoefer, R. A. Unpublished results.

(37) Margl, P.; Ziegler, T. J. Am. Chem. Soc. 1996, 118, 7337.

9

6338 J. AM. CHEM. SOC. VOL. 126, NO. 20, 2004

Pd-Catalyzed Diene Cyclization/Hydrosilylation

A R T I C L E S

Scheme 6

Scheme 7

Scheme 8

the coordinated olefin into the M-C bond of a nonchelated

transition metal alkyl olefin complex has been directly ob-

served,^{15,16,25,26,30,31,34,35,38,39}olefin â-migratory insertion of the

coordinated olefin into the M-C bond of a transition metal alkyl

olefin chelate complex has not. Rather, the majority of well-

characterized transition metal alkyl olefin chelate complexes

decompose via â-hydride elimination without undergoing

â-migratory insertion.⁴⁰For example, the nickel 4-pentenyl

chelate complex CpNi[η¹,η²-CH₂CH₂CH₂CHdCH₂] (Cp )

C₅H₅) isomerized at 70 °C in THF to form a mixture of π-allyl

complexes CpNi(η³-CH₃CHCHCHCH₃) and CpNi(η³-CH₂-

CHCHCH₂CH₃) (Scheme 7).⁴¹Similarly, the cationic platinum

4-pentenyl chelate complex [(PMe₃)₂Pt(η¹,η²-CH₂CMe₂CH₂-

CHdCH₂)]⁺[BF₄]^-rearranged at -10 °C to form a mixture

of the isomeric π-allyl complexes [(PMe₃)₂Pt(η³-CMe₂CH-

CHMe)]⁺[BF₄]^-and [(PMe₃)₂Pt(η³-CH₂CMeCHMeCH₂CH₃)]⁺

[BF₄]^-.⁴²The neutral rhodium 3-butenyl bis(phosphine) carbonyl

chelate complex (PPh₃)₂Rh(CO)[η¹,η²-CH₂C(Ph)₂CHdCH₂]

decomposed at 50 °C to form 1,1-diphenyl-1,3-butadiene in 44%

yield.⁴³Although treatment of the palladium norbornyl chloride

dimer 10 with dppe [dppe ) bis(diphenylphosphino)ethane] led

to rapid â-migratory insertion to form the palladium tricyclo-

[2.2.1.0^0,0]heptyl complex 11, the reactive alkyl olefin phosphine

complex was not detected (Scheme 8).⁴⁴

In two cases, â-migratory insertion of the olefin into the M-C

bond of a 4-pentenyl chelate complex has been directly

implicated through scrambling of deuterium atoms between the

C(1) and C(3) carbon atoms of the pentenyl ligand. For example,

Hallenbeck and Casey have reported that the yttrocene 4-pen-

tenyl chelate complex Cp*₂Y[η¹,η²-CD₂CH₂CH₂CHdCD₂] (12-

1,1,5,5-d₄) (Cp* ) C₅Me₅) isomerized over 2 h at -78 °C to

form a 1:1 mixture of 12-1,1,5,5-d₄and 12-3,3,5,5-d₄(Scheme

9).⁴⁵Likewise, Flood has reported that thermolysis of the

platinum 4-pentenyl chelate complex (dmpe)Pt[η¹,η²-CD₂CMe₂-

CH₂CHdCH₂] (13-1,1-d₂) [dmpe ) bis(dimethylphosphino)-

ethane] at 125 °C for 8 h formed a 1:1 mixture of 13-1,1-d₂

and 13-3,3-d₂.⁴⁶However, in neither case was the cyclobutyl-

methyl intermediate observed and, for this reason, conversion

of 5 to 6 represents the first direct observation of the â-migratory

insertion of the coordinated olefin into the M-C bond of an

alkyl olefin chelate complex. Furthermore, due to the substantial

(38) (a) Shultz, C. S.; Ledford, J.; DeSimone, J. M.; Brookhart, M. J. Am. Chem.

Soc. 2000, 122, 6351. (b) Svejda, S. A.; Johnson, L. K.; Brookhart, M. J.

Am. Chem. Soc. 1999, 121, 10634. (c) Brookhart, M.; Lincoln, D. M. J.

Am. Chem. Soc. 1988, 110, 8719.

(39) (a) Brookhart, M.; Hauptman, E.; Lincoln, D. M. J. Am. Chem. Soc. 1992,

114, 10394. (b) Wang, L.; Flood, T. C. J. Am. Chem. Soc. 1992, 114, 3169.

(40) Lehmkuhl, H. Pure Appl. Chem. 1986, 58, 495 and references therein.

(41) Lehmkuhl, H.; Naydowski, C.; Benn, R.; Rufin˜ska, A.; Schroth, G.; Mynott,

R.; Kru¨ger, C. Chem. Ber. 1983, 116, 2447.

(42) Flood, T. C.; Statler, J. A. Organometallics 1984, 3, 1795.

(43) Jun, C.-H. Organometallics 1996, 15, 895.

(44) Coulson, D. R. J. Am. Chem. Soc. 1969, 91, 200.

(45) (a) Casey, C. P.; Hallenbeck, S. L.; Pollock, D. W.; Landis, C. R. J. Am.

Chem. Soc. 1995, 117, 9770. (b) Casey, C. P.; Hallenbeck, S. L.; Wright,

J. M.; Landis, C. R. J. Am. Chem. Soc. 1997, 119, 9681.

(46) (a) Flood, T. C.; Bitler, S. P. J. Am. Chem. Soc. 1984, 106, 6076. (b) Ermer,

S. P.; Struck, G. E.; Bitler, S. P.; Richards, R.; Bau, R.; Flood, T. C.

Organometallics 1993, 12, 2634.

9

J. AM. CHEM. SOC. VOL. 126, NO. 20, 2004 6339

A R T I C L E S

Perch and Widenhoefer

Scheme 9

5 to 6 is irreversible, and because the stereochemistry of alkyl

olefin chelate complex 5 is likely determined thermodynamically

(see above), conversion of 5 to 6 represents the stereochemistry-

determining step in the palladium-catalyzed cyclization/hydrosi-

lylation of 1.

Effect of the Chelate on Olefin â-Migratory Insertion.

Activation parameters for olefin â-migratory insertion have been

determined for a number of nonchelated, cationic palladium

alkyl olefin complexes.^{15,16,25,26,30,31}Notable among these is the

â-migratory insertion of the ethyl ethylene complex [(phen)-

Pd(Et)(H₂CdCH₂)]⁺[BAr₄]^-(14) to form the n-butyl complex

[(phen)Pd(CH₂CH₂CH₂CH₃)]⁺[BAr₄]^-(15),¹⁵as complexes 5,

6, 14, and 15 all possess the identical (phen)Pd⁺core. Activation

parameters for conversion of 5 to 6 differ significantly from

those for conversion of 14 to 15, pointing to the influence of

the hexenyl chelate on the energetics of olefin â-migratory

insertion. For example, the entropy of activation for the

conversion of 5 to 6 (∆S^q) -15 ( 2 eu) is significantly more

negative than is the entropy of activation for conversion of 14

to 15 (∆S^q) -3.7 ( 2 eu); the latter value is typical for the

olefin â-migratory insertion of nonchelated, late transition metal

complexes.^39,52,53Because the olefin ligand of 5 is oriented

perpendicular to the coordination plane in the ground state but

must lie parallel to the coordination plane in the transition state

for â-migratory insertion,⁵⁴the negative entropy of activation

for the conversion of 5 to 6 indicates that the hexenyl chain

adopts a more conformationally rigid orientation when the olefin

is oriented in the coordination plane as opposed to perpendicular

to the coordination plane. Although the olefin of 14 must also

undergo 90° rotation prior to olefin â-migratory insertion,

rotation of the ethylene ligand of 14 requires no reorganization

elsewhere in the molecule.

strain that accompanies cyclobutyl ring formation, the â-migra-

tory insertion of 12 and 13 are not likely relevant to the

â-migratory insertion processes that occur in synthetically

relevant transition metal-catalyzed carbocyclization protocols.^2,47

Conversion of 5 to 6. The zero-order dependence of the rate

of conversion of 5 to 6 on nitrile concentration rules out a

mechanism for the conversion of 5 to 6 initiated by attack of

nitrile on 5 or a mechanism involving rapid and reversible

conversion of 5 to III followed by rate-limiting ligand capture

to form 6. Rather, our kinetic data are in accord with a

mechanism involving rate-limiting, irreversible conversion of

5 to III followed by rapid, exothermic ligand capture to form

6 (Scheme 6). DFT calculations indicate that olefin â-migratory

insertion of ethylene into the Pd-C bond of 9 to form the

palladium â-agostic butyl complex [(H₂PCHdCHPH₂)Pd(CH₂-

CH₂CH₂CH₃)]⁺is exothermic by ∼8 kcal mol^-1and, as noted

above, complexation of ethylene to the â-agostic complex 8 is

exothermic by ∼12 kcal mol^-1. Because NCAr and ethylene

possess comparable ligating ability with respect to cationic Pd(II)

complexes,^16,24,36we estimate that the conversion of 5 to 6 is

Despite the less favorable ∆S^qfor the conversion of 5 to 6

relative to ∆S^qfor the conversion of 14 to 15, the free energy

of activation for the conversion of 5 to 6 (∆G^q) 16.9 ( 0.1

kcal mol^-1) is 2.6 kcal mol^-1lower than is the free energy of

activation for the conversion of 14 to 15 (∆G^q) 19.5 ( 0.5

kcal mol^-1) due to the significantly lower enthalpy of activation

also exothermic by ∼20 kcal mol^{-1 48}

. From this value and from

the enthalpy of activation for the conversion of 5 to 6 (∆H^q)

13.5 kcal mol^-1), we estimate that the enthalpy of activation

for the reverse reaction (6 f 5) is ∆H^q≈ 33 kcal mol^-1, which

clearly renders conversion of 5 to 6 irreversible under reaction

conditions. Consistent with this conclusion, â-alkyl elimination

has been observed only in the case of electrophilic d⁰-

metallocene complexes⁴⁹and in strained cyclobutylmethyl^45,46

and cyclopropylmethyl complexes.^50,51Because conversion of

for the conversion of 5 to 6 (∆H^q) 13.5 ( 0.6 kcal mol^-1

)

relative to the conversion of 14 to 15 (∆H^q) 18.5 ( 0.6 kcal

mol^-1).⁵⁵This large difference in the enthalpy of activation

(∆∆H^q) 5 kcal mol^-1) represents the most conspicuous

difference in the activation parameters for the olefin â-migratory

insertion of 5 and 14. We attribute this large ∆∆H^qto ground-

state destabilization of the alkyl olefin chelate complex 5.⁵⁶

Assuming the preferred conformation of a transition metal alkyl

olefin chelate complex mirrors that of a substituted cyclohexane

ring, destabilization of 5 relative to 14 is suggested by the

pseudoaxial orientation of the R-triethylsilylmethyl group of 5.

Presumably, unfavorable steric interactions within the hexenyl

chain of 5 that result from the pseudoaxial orientation of the

(47) Flood also noted that “incorporation of substantial ring strain and

conformational restrictions of the Pt-CH₂group in the transition state [for

the â-migratory insertion of 13], especially the former, make the transition

state potentially quite atypical.”^46b

(48) Although this analysis does not account for the likely destabilization of

both 5 and 6 due to the presence of the hexenyl chelate and cyclopentyl

ring, respectively, 5 and 6 are likely destabilized by a comparable amount

relative to the corresponding acyclic derivatives. Destabilization of 5 by

∼5 kcal mol^-1relative to a nonchelated alkyl olefin complex is suggested

by the lower ∆H^qfor â-migratory insertion of 5 relative to that for

unchelated derivative 14 (∆∆H^q) 5 kcal mol^-1) and by the pseudoaxial

orientation of the R-triethylsilyl group of 5 in solution. Similarly, because

there is approximately 6.5 kcal mol^-1of strain energy associated with

formation of a cyclopentyl ring, cyclopentylmethyl complex 6 is likely

destabilized by ∼6 kcal mol^-1relative to the corresponding acyclic alkyl

complex.

(52) (a) Doherty, N. M.; Bercaw, J. E. J. Am. Chem. Soc. 1985, 107, 2670. (b)

Berger, B. J. Santarsiero, B. D.; Trimmer, M. S. Bercaw, J. E. J. Am. Chem.

Soc. 1988, 110, 3134.

(49) (a) Watson, P. L. J. Am. Chem. Soc. 1982, 104, 6472. (b) Horton, A. D.

Organometallics 1996, 15, 2675. (c) Guo, Z.; Swenson, D. C.; Jordan, R.

F. Organometallics 1994, 13, 1424. (d) Hajela, S.; Bercaw, J. E.;

Organometallics 1994, 13, 1147.

(53) Although the entropy of activation for the reversible â-migratory insertion

of the platinum 4-pentenyl chelate complex 13-1,1-d₂was near zero (∆S^q

) -0.5 ( 1.6 eu), ¹H-¹H NOESY analysis of 13-1,1-d₂was consistent

with orientation of the olefin in or near the coordination plane, and for this

reason, no significant olefin rotation or reorganization of the pentenyl chain

was required for olefin â-migratory insertion.^46b

(54) (a) Thorn, D. L.; Hoffmann, R. J. Am. Chem. Soc. 1978, 100, 2079. (b)

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.

(50) Wang, X.; Stankovich, S. Z.; Widenhoefer, R. A. Organometallics 2002,

21, 901.

(51) In one notable exception, thermolysis of [(dmpe)Pd(PMe₃)CH₂CMe₂Ph]⁺

[BAr₄]^-[dmpe ) Me₂PCH₂CH₂PMe₂; Ar ) 3,5-C₆H₃(CF₃)₂] at 60 °C for

24 h led to â-phenyl elimination to form [(dmpe)Pd(PMe₃)Ph]⁺[BAr₄]^-:

Ca´mpora, J.; Gutie´rrez-Puebla, E.; Lo´pez, J. A.; Monge, A.; Palma, P.; del

R´ıo, D.; Carmona, E. Angew. Chem., Int. Ed. 2001, 40, 3641.

9

6340 J. AM. CHEM. SOC. VOL. 126, NO. 20, 2004

Pd-Catalyzed Diene Cyclization/Hydrosilylation

A R T I C L E S

Scheme 10

R-triethylsilylmethyl group are less severe than is the steric

interaction between the R-triethylsilylmethyl group and the

phenanthroline ligand when the former adopts a pseudoequa-

torial orientation. Ground-state destabilization has been previ-

ously invoked to rationalize the decreasing ∆G^qfor â-migratory

insertion of propylene into the Pd-CH₃bond of the palladium

diimine complexes {[ArNdC(Me)C(Me)dNAr]Pd(Me)CH₂d

CHMe}⁺with the increasing steric bulk of the diimine ligand.³⁰

We considered two mechanisms by which the ground-state

destabilization of 5 could decrease ∆H^qfor the conversion of 5

to 6 relative to that of 14 to 15. The in-plane rotamer of cationic

palladium(II) phenanthroline olefin complexes is ∼10 kcal

mol^-1less stable than is the perpendicular rotamer;^14-16

calculations attribute destabilization of the in-plane rotamer to

unfavorable steric interaction of the olefin ligand with the cis

ligands.⁵⁷Because it appears likely that the energy barrier for

olefin rotation contributes significantly to the energy barrier for

olefin â-migratory insertion,⁵⁴destabilization of the perpen-

dicular rotamer of 5 relative to the in-plane rotamer due to strain

within the hexenyl chain would lead to lower ∆H^qfor

conversion of 5 to 6 relative to the conversion of 14 to 15.

Alternatively, DFT calculations for the â-migratory insertion

of ethylene into the Pd-CH₃bond of [(HNdCHCHdNH)Pd-

(Me)(H₂CdCH₂)]⁺(16) to form [(HNdCHCHdNH)Pd(CH₂-

CH₂CH₃)]⁺(17) indicate that the CH₃-Pd-ethylene_(centroid)

angle decreases from 95° in the in-plane rotamer of 16 to 75°

in the transition state for conversion of 16 to 17.⁵⁸Therefore,

compression of the C-Pd-olefin_(centroid)angle of 5 due to strain

within the hexenyl chain could increase the extent of C-C bond

formation in the transition state for the conversion of 5 to 6

relative to the transition state for the conversion of 14 to 15,

leading to a lower ∆H^qfor the conversion of 5 to 6 relative to

the conversion of 14 to 15.

Formation of Carbonyl Chelate Complex 7. In the absence

of silane, palladium cyclopentylmethyl complex 6 rearranges

within 2 h at -9 °C to form the carbonyl chelate complex 7

with retention of stereochemistry (Scheme 2). Stereoselective

conversion of 6 to 7 likely occurs via iterative â-hydride

elimination/addition to form the palladium cyclopentyl inter-

mediate IX without displacement of the olefin ligand from

intermediates VII and VIII, followed by highly exoergonic

complexation of the pendant carbonyl group to form 7 (Scheme

10). An analogous mechanism was proposed for formation of

carbonyl chelate complex 7a from stoichiometric reaction of 1

with 2c.¹⁷Because both the formation and silylation of 7 are

much slower than is the silylation of 6 under catalytic conditions

(see below), we can conclusively rule out formation of 7 in the

palladium-catalyzed cyclization/hydrosilylation of 1.

Silylation of Palladium Alkyl Complexes. The first-order

dependence of the rate of silylation of 7 on silane concentration

rules out a mechanism for the silylation of 7 initiated by rate-

limiting dissociation of the chelated carbonyl group. Further-

more, the much lower rate for the silylation of 7, which

possesses a tightly coordinated carbonyl chelate ligand, relative

to the silylation of 6, which possesses a labile NCAr ligand,

argues against mechanisms for the silylation of 7 involving

Pd-C bond cleavage without prior silane coordination or Pd-C

bond cleavage from a five-coordinate palladium silane complex.

Rather, our data are in accord with a mechanism for the

conversion of 7 to 3 involving associative displacement of the

carbonyl ligand of 7 with silane to form the four-coordinate

palladium silane intermediate X followed by Pd-C bond

cleavage to form 3, perhaps via σ-bond metathesis (Scheme

10).⁵⁹

(55) The free energy of activation for the â-migratory insertion of propene into

the Pd-CH₃bond of the cationic palladium propylene complex [(N-N)-

Pd(Me)CH₂dCHMe]⁺[N-N ) ArNdC(Me)C(Me)dNAr, Ar ) 2,6-

C₆H₃i-Pr₂] was only 0.5 kcal mol^-1higher than was ∆G^qfor the insertion

of ethylene into the Pd-CH₃bond of the corresponding palladium ethylene

complex [(N-N)Pd(Me)CH₂dCH₂]⁺.²⁵For this reason, it appears unlikely

that the lower ∆G^qand ∆H^qfor olefin â-migratory insertion for 5 relative

to that for 14 is due to the insertion of an R-olefin in the case of 5 and

ethylene in the case of 14. Furthermore, ∆G^qfor â-migratory insertion of

ethylene into the Pd-CH₃bond of [(phen)Pd(CH₃)(H₂CdCH₂)]⁺[BAr₄]^-

was ∼1 kcal mol^-1lower than was â-migratory insertion of ethylene into

the Pd-CH₂CH₃bond of 14.^15,16This comparison suggests that ∆G^qfor

olefin â-migratory insertion increases with increasing substitution of the

palladium σ-bound carbon atom. For this reason, it appears highly unlikely

that the lower ∆G^qfor the conversion of 5 to 6 relative to the conversion

of 14 to 15 is due to migration of a Pd-secondary alkyl group in the case

of 5 and a palladium-primary alkyl group in the case of 14.

(56) Enthalpic destabilization of 5 relative to 14 is not inconsistent with the

failure to generate detectable amounts of the nonchelated palladium hexenyl

complex 5-L from reaction of 5 with 1 or NCAr (Scheme 3). An entropy

of reaction for the conversion of 5 to 5-L of ∆S ≈ -30 eu can be estimated

from the values determined for displacement of the six-membered carbonyl

chelate from the cationic palladium complex [(phen)Pd[CH₂-

CH₂CH₂COOMe]⁺[BAr₄]^-with ethylene and similar ligands (∆S ) -27

to -34 eu).¹⁶Because NCAr and an R-olefin are of comparable ligating

ability with respect to Pd(II),³⁶the enthalpy for the conversion of 5 to 5-L

would be ∆H ≈ 0 if there were no significant destabilization of 5. However,

if we attribute the 5 kcal mol^-1decrease in ∆H^qfor â-migratory insertion

of 5 relative to that for 14 solely to ground-state destabilization of 5, the

enthalpy of reaction for the conversion of 5 to 5-L would be ∆H ≈ -5

kcal mol^-1. From these values (∆H ≈ -5 kcal mol^-1, ∆S ≈ -30 eu) we

estimate a free energy of reaction for the conversion of 5 to 5-L of ∆G ≈

2 kcal mol^-1, which corresponds to an equilibrium constant of K_eq≈ 8 ×

10^-3at -62 °C. Therefore, 5-L should constitute less than 0.1% of a mixture

of 5 (42 mM), 1 (42 mM), and NCAr (42 mM) at -62 °C and, therefore,

5-L should be undetectable by ¹H NMR analysis under these conditions.

(57) Hay, P. J. J. Am. Chem. Soc. 1981, 103, 1390.

(58) Svensson, M.; Matsubara, T.; Morokuma, K. Organometallics 1996, 15,

5568.

(59) Alternatively, Pd-C bond cleavage could occur via an oxidative addition/

reductive elimination sequence.

9

J. AM. CHEM. SOC. VOL. 126, NO. 20, 2004 6341

A R T I C L E S

Perch and Widenhoefer

Scheme 11

All available evidence regarding the silylation of 6 points to

an associative mechanism involving palladium silane intermedi-

ate IV (Scheme 6). The strong inhibition of the rate of the

silylation of 6 by CD₃CN and the much higher rate of the

silylation of 6 relative to the silylation of 7 argues against

mechanisms for the silylation of 6 involving Pd-C bond

cleavage without prior silane coordination or Pd-C bond

cleavage from a five-coordinate palladium silane complex.

Although the first-order dependence of the rate of the silylation

of 6 on silane concentration was not directly established, silane

competition experiments support this contention. Provided that

scrambling of the silyl group of 4b′ with free silane is slow

relative to irreversible conversion of 4b′ to 5′,^60,61the ratio of

silylated carbocycles formed in the catalytic cyclization/hy-

drosilylation of 1 with an excess 1:1 mixture of two different

silanes corresponds to the relative rate at which the two silanes

react with palladium cyclopentylmethyl complex 6′. Therefore,

silane competition experiments establish that the rate of sily-

lation of 6′ increases with both the decreasing steric bulk and

increasing electron density of the silane. Both of these observa-

tions are consistent with a mechanism for silylation involving

direct attack of the silane on the palladium atom of 6′.⁶²

eq 2 when the conversion of 4 to 5 is much faster than is the

conversion of 5 to 6, and to the single-term rate law depicted

in eq 3, when both the conversion of 4 to 5 and the silylation

of 6 are fast relative to the conversion of 5 to 6.⁶⁴

All of our experimental observations point to turnover-limiting

conversion of alkyl olefin chelate complex 5 to cyclopentyl-

methyl complex 6 in the catalytic conversion of 1 to 3 at high

silane concentration ([HSiEt₃]₀) 0.16 M). For example, the

empirical rate law for the catalytic cyclization/hydrosilylation

of 1 and HSiEt₃at high silane concentration (rate ) k_sat[2b]₀)

is of the same form as the derived rate law depicted in eq 3

(rate ) k₂[Pd]_tot). Likewise, the activation parameters for the

catalytic conversion of 1 to 3 at high silane concentration

(∆G^q_232K) 16.8 ( 0.1 kcal mol^-1; ∆H^q) 13 ( 1 kcal mol^-1

;

∆S^q) -15 ( 3 eu) do not differ significantly from those

determined independently for the conversion of 5 to 6 (∆G^q

232

^K) 16.9 ( 0.1 kcal mol^-1; ∆H^q) 13.5 ( 0.6 kcal mol^-1; ∆S^q

) -15 ( 2 eu). Furthermore, alkyl olefin chelate complex 5

was the only palladium species detected during catalytic

cyclization/hydrosilylation of 1 at high silane concentration,

which establishes 5 as the catalyst resting state under these

conditions.

Kinetics of Catalytic Cyclization/Hydrosilylation. The

kinetics of the catalytic cyclization/hydrosilylation of 1 and

HSiEt₃were interpreted in the context of the simplified

mechanism depicted in Scheme 11, which was constructed on

the basis of the following assumptions: (1) conversion of 4b

to 5, 5 to 6, and 6 to 4b is irreversible, (2) the total palladium

The pronounced positive curvature of the concentration versus

time plots for the catalytic cyclization/hydrosilylation of 1 at

lower silane concentration ([HSiEt₃]₀) 90 mM, Figure 6)

indicates that under these conditions, the rate of silylation of

cyclopentylmethyl complex 6 is both comparable to the rate of

conversion of 5 to 6 and is dependent on silane concentration.

Iterative fitting of two sets of concentration versus time data

for the catalytic cyclization/hydrosilylation of 1 at low silane

concentration ([HSiEt₃]₀) 90 mM) to the rate law depicted in

eq 2 provided a best fit with k₂) 7.6 ( 0.2 × 10^-4s^-1and k₃

) 0.12 ( 0.03 M^-1s^-1. Noteworthy is that the value for k₂

extracted from this analysis is in good agreement with the value

obtained under saturation conditions (k_sat) k₂) 7.74 ( 0.06

× 10^-4s^-1). Furthermore, because silylation of 6 is likely

concentration equals the initial concentration of 2b ([Pd]_tot

)

[2b]₀), (3) conversion of 2b to 4b is fast relative to catalyst

turnover,⁶³and (4) the rate of silylation of 6 depends linearly

on silane concentration (see above). Steady-state treatment of

intermediates 4b and 6 produced the three-term rate law depicted

in eq 1, which simplifies to the two-term rate law depicted in

(60) The designators 4b′, 5′, and 6′ refer to the palladium silyl, alkyl olefin

chelate, and cyclopentylmethyl derivatives, respectively, with an undefined

silyl group -SiMe₂R (R ) Et, n-octyl, OSiMe₃, Ph, 4-C₆H₄NMe₂, 4-C₆H₄-

OMe, 4-C₆H₄F, or 4-C₆H₄CF₃) generated in the catalytic silane competition

experiments.

(61) Brookhart has shown that reaction of palladium silyl olefin complex [(phen)-

Pd(SiEt₃)(CH₂dCHt-Bu)]⁺[BAr₄]^-[Ar ) 3,5-C₆H₃(CF₃)₂] (4e) with

HSiPh₃formed (3,3-dimethylbutyl)triethylsilane to the exclusion of (3,3-

dimethylbutyl)triphenylsilane,¹⁴which indicates that olefin insertion/

silylation of 4e is much faster than is silyl exchange. Because NCAr and

an olefin of 1 are of comparable ligating ability with respect to cationic

Pd(II) complexes,³⁶this result strongly suggests that 4b does not undergo

silyl exchange prior to conversion to 5.

(63) Available evidence supports each of these assumptions. The failure of 1-2,6-

d₂to exchange with the 5-hexenyl group of 5 establishes irreversible

conversion of 4b to 5, analysis of DFT data supports the irreversible

conversion of 5 to 6,³⁷and Brookhart has established the irreversible

silylation of a Pd-C bond with a hydrosilane.¹⁴Visual inspection of

catalytically active mixtures of 1, 2b, and HSiEt₃revealed no darkening

of the solution prior to complete consumption of 1, which argues against

catalyst decomposition during cyclization/hydrosilylation. The much faster

conversion of 2b and HSiEt₃to 4b (5 min at -80 °C) relative to the

conversion of 5 to 6 (t_1/2) 22 min at -41 °C) ensures rapid activation of

precatalyst 2b relative to catalyst turnover.

(62) Silane competition experiments are in accord with either rate-limiting

conversion of 6 to IV followed by rapid conversion of IV to 3, or rapid

and reversible conversion of 6 to IV followed by rate-limiting conversion

of IV to 3. The transition states for both the conversion of 6 to IV (6^q) and

VI to 4b (IV^q) should possess a greater degree of covalent Pd-Si bond

character than do 6 and IV, respectively. Likewise, the Pd-Si bond distance

should decrease upon conversion of both 6 to 6^qand IV to IV^q. Therefore,

both 6^qand IV^qshould be destabilized by the decreasing electron density

and increasing steric bulk of the silane.

(64) Although the mechanism for cyclization/hydrosilylation depicted in Scheme

6 likely involves nitrile inhibition of both the conversion of 4b to 5 and

the conversion of 6 to 4b, both these transformations are fast relative to

the conversion of 5 to 6 at high silane concentration ([HSiEt₃]₀) 0.16M).

Although conversion of 6 to 4b becomes kinetically relevant at low silane

concentration, nitrile concentration was invariant ([NCAr] ) 42 mM)

throughout complete conversion of 1 to 3 and between multiple experiments.

Therefore, k₃corresponds to a macroscopic rate constant for the conversion

of 6 to 4b for the specific case where [NCAr] ) 42 mM.

9

6342 J. AM. CHEM. SOC. VOL. 126, NO. 20, 2004

Pd-Catalyzed Diene Cyclization/Hydrosilylation

A R T I C L E S

Figure 10. Atom-labeling schemes for phenanthroline, NCAr, BAr₄^-, 6,

and 7.

Figure 9. Theoretical plot of k_obsversus [HSiEt₃] for the catalytic

cyclization/hydrosilylation of 1 and HSiEt₃calculated from eq 2 where [1]

exothermic ligand capture with NCAr forms 6. Conversion of

5 to 6 is likely irreversible and is therefore the stereochemistry-

determining step in catalytic cyclization/hydrosilylation. The

enthalpy of activation for â-migratory insertion of the olefin

into the Pd-C bond of 5 is 5 kcal mol^-1lower than is ∆H^qfor

â-migratory insertion of the olefin into the Pd-C bond of the

analogous nonchelated alkyl olefin complex 14. We attribute

this large ∆∆H^qto ground-state destabilization of 5 relative to

14. Associative silylation of 6 via palladium-silane intermediate

IV releases carbocycle 3 and regenerates palladium silyl

complex 4b. Under catalytic conditions, the second-order rate

constant for the silylation of 6 (k₃) 0.12 ( 0.03 M^-1s^-1) is

∼160 times greater than is the first-order rate constant for

conversion of 5 to 6 (k₂) 7.6 ( 0.2 × 10^-4s^-1), and both of

these processes are slow relative to the conversion of 4b to 5.

Therefore, at high silane concentration ([HSiEt₃] > 85 mM),

silylation of 6 is fast relative to conversion of 5 to 6, and the

latter step alone determines the rate of catalytic cyclization/

hydrosilylation. Conversely, at low silane concentration ([HSi-

Et₃] < 85 mM), the rate of silylation of 6 becomes competitive

with the rate of conversion of 5 to 6 and the rate of catalytic

cyclization/hydrosilylation depends on silane concentration.

Although cyclopentylmethyl complex 6 rearranges to oxo

chelate complex 7 in the absence of silane, neither the

conversion of 6 to 7 nor the silylation of 7 with HSiEt₃to form

3 is fast enough to compete with the silylation of 6 under

catalytic conditions.

) 85 mM, [Pd]_tot) 12 mM, and [NCAr] ) 42 mM, k₂) 7.6 × 10^-4s^-1

,

and k₃) 0.12 M^-1s^-1

.

inhibited by nitrile, the value for k₃extracted from this analysis

([NCAr] ) 42 mM) is in accord with the rate constant estimated

for the silylation of 6 under stoichiometric conditions ([NCAr]

≈ 0 mM, k g 0.19 M^-1s^-1).²⁷

To better visualize the dependence of the rate of catalytic

cyclization/hydrosilylation on silane concentration, a hypotheti-

cal plot of k_obsversus [HSiEt₃] for the catalytic cyclization/

hydrosilylation of 1 and HSiEt₃was generated by solving eq 2

for k_obs(k_obs) rate/[1]) as a function of [HSiEt₃] in 0.01 M

increments from 0 to 0.50 M, where [1] ) 85 mM, [Pd]_tot

)

12 mM, and [NCAr] ) 42 mM, k₂) 7.6 × 10^-4s^-1, and k₃)

0.12 M^-1s^-1(Figure 9). This theoretical plot indicates that the

rate of catalytic cyclization/hydrosilylation should decrease less

than 3% as the silane concentration decreases from 0.16 to 0.080

M, which is in accord with the observed linearity of the

concentration versus time plots for catalytic cyclization/hy-

drosilylation at high silane concentration ([HSiEt₃]₀) 0.16 M)

through 90% conversion (Figure 6). Likewise, the theoretical

plot predicts that the rate of catalytic cyclization/hydrosilylation

should decrease ∼50% as the silane concentration decreases

from 90 to 5 mM, which is in accord with the pronounced

positive curvature of the concentration versus time plots for

catalytic cyclization/hydrosilylation at low silane concentration

([HSiEt₃]₀) 90 mM) through 95% conversion (Figure 6).

Experimental Section

Conclusions

General Methods. Low-temperature NMR spectra were recorded

1

at 500 MHz for H and 125 MHz for ¹³C, except where noted; room-

All of our observations support the mechanism for the

cyclization/hydrosilylation of 1 and HSiEt₃catalyzed by 2b

depicted in Scheme 6. Rapid and reversible complexation of

one of the double bonds of 1 with palladium silyl intermediate

4b would form the unobserved palladium olefin complex I.

â-Migratory insertion of the olefin into the Pd-Si bond of I to

form the palladium 5-hexenyl complex II, followed by rapid,

irreversible complexation of the pendant olefin would form alkyl

olefin chelate complex 5. Although not detected, associative

displacement of the chelated olefin of 5 by exogenous ligand

is likely and, for this reason, the diastereoselective formation

of 5 is under thermodynamic control. â-Migratory insertion of

the pendant olefin into the Pd-C bond of 5 to form the

palladium cyclopentylmethyl complex III followed by rapid,

1

temperature NMR spectra were recorded at 400 MHz for H and 100

MHz for ¹³C, except where noted. NMR Probe temperatures were

measured with a methanol thermometer and were maintained to within

-

( 0.5 °C. Atom-labeling schemes for phenanthroline, NCAr, BAr₄

,

6, and 7 are shown in Figure 10. ¹H NMR yields for the conversion of

4b to 5, 5 to 6, and 6 to 7 were determined employing a 10 s delay

between pulses. Unless noted otherwise, error limits for rate constants

refer to the standard deviation of the slope of the respective kinetic

plot or to the standard deviation of the average of multiple experiments.

The volumes of low-temperature NMR solutions were calculated from

the room-temperature volume and the temperature dependence of the

density of CH₂Cl₂; volumes given refer to the volume at the specified

reaction temperature. Elemental analysis was performed by Complete

Analysis Laboratories, Inc. (Parsippany, NJ).

9

J. AM. CHEM. SOC. VOL. 126, NO. 20, 2004 6343

A R T I C L E S

Perch and Widenhoefer

[(phen)Pd(SiEt₃)(NCAr)]⁺[BAr₄]^-(4b). HSiEt₃(2.62 mg, 22.5

µmol) was added via syringe to an NMR tube containing a solution of

[(phen)Pd(Me)(NCAr)]⁺[BAr₄]^-(2b) (31.6 mg, 22.5 µmol) in CD₂-

Cl₂(0.54 mL) at -78 °C. The tube was shaken briefly and placed into

the probe of an NMR spectrometer pre-cooled at -81 °C. Reaction

progress was determined by measuring the disappearance of the Pd-

CH₃resonance of 2b (δ 1.26) relative to the para phenyl protons of

the [BAr₄]^-counterion (δ 7.46) in the ¹H NMR spectrum. After 5 min,

2b was completely consumed to form 4b in 101 ( 10% yield. Complex

4b was thermally sensitive and characterized in solution at -81 °C by

CH₂], 32.5 (d, J ) 41 Hz, -CH₂CHdCH₂), 20.3 [d, J ) 30 Hz, Pd-

CH(CH₂SiEt₃)CH₂].

Exchange of 1-2,6-d₂with 5. Diene 1-2,6-d₂(5.88 µL, 29.2 mmol)

was added via syringe to an NMR tube containing a solution of 5 (∼25

mmol) in CD₂Cl₂(0.58 mL) at -80 °C. The tube was shaken briefly

and placed in the probe of an NMR spectrometer cooled at -80 °C.

The sample was maintained at -80 °C for 1.5 h and then warmed at

-60 °C for 30 min. At this time, ¹H NMR analysis revealed

approximately 5% conversion to 6 without formation of either free 1

or 5-d₂.

1

¹H NMR spectroscopy. H NMR (CD₂Cl₂, -81 °C): δ 9.02 (d, J )

trans-{(phen)Pd[CH₂CHCH₂C(CO₂Me)₂CH₂CHCH₂SiEt₃]-

(NCAr)}⁺[BAr₄]^-(6). An NMR tube containing a solution 5 (39 mM)

and NCAr (39 mM) in CD₂Cl₂was warmed at -41 °C and monitored

4.9 Hz, 1 H, H_phen), 8.83 (d, J ) 4.5 Hz, 1 H, H_phen), 8.55 (d, J ) 8.2

Hz, 1 H, H_phen), 8.43 (d, J ) 8.3 Hz, 1 H, H_phen), 8.39 (s, 2 H, H_o),

8.36 (s, 1 H, H_p), 7.93 (m, 3 H, H_phen), 7.78 (dd, J ) 4.8, 8.0 Hz, 1 H,

H_b), 7.73 (s, 8 H, H_o′), 7.46 (s, 4 H, H_p′), 1.06 (t, J ) 7.5 Hz, 9 H,

SiCH₂CH₃), 0.97 (q, J ) 7.5 Hz, 6 H, SiCH₂CH₃).

1

periodically by H NMR spectroscopy; reaction progress was deter-

mined by integrating the carbomethoxy resonances of 6 (δ 3.67 and

3.59) and 5 (δ 3.69 and 3.43) relative to the para phenyl resonances of

{(phen)Pd[η¹,η²-CH(CH₂SiEt₃)CH₂C(CO₂Me)₂CH₂CHd

CH₂]}⁺[BAr₄]^-(5). Dimethyl diallylmalonate (1) (4.7 µL, 0.023 mmol)

was added via syringe to an NMR tube containing a solution of [(phen)-

Pd(SiEt₃)(NCAr)]⁺[BAr₄]^-(4b) (0.023 mmol) in CD₂Cl₂(0.55 mL)

at -78° C. The tube was shaken briefly and placed in the probe of an

NMR spectrometer pre-cooled at -81 °C. The probe was warmed at

-62 °C and the solution was analyzed periodically by ¹H NMR

spectroscopy. Reaction progress was determined by integrating the

carbomethoxy resonances of 5 (δ 3.68 and 3.43) relative to that of the

para phenyl proton of the BAr₄^-counterion (δ 7.49). After 20 min, 4b

was completely consumed to form 5 as the exclusive product in 84%

( 10% yield as a single diastereomer characterized by ¹H NMR

spectroscopy. Complex 5 was thermally sensitive and was characterized

-

the BAr₄counterion (δ 7.51). After 2 h, 5 had been completely

consumed to form 6 in 96 ( 10% yield by ¹H NMR analysis. Complex

6 was thermally sensitive and was characterized in solution by ¹H and

¹³C NMR spectroscopy at e-41 °C. Assignment of proton resonances

and J_HHcoupling constants was aided by ¹H COSY analysis and by ¹H

NMR analysis of the labeled derivatives trans-{(phen)Pd[CH₂CDCH₂C-

(CO₂Me)₂CH₂CDCH₂SiEt₃](NCAr)]}⁺[BAr₄]^-(6-d₂) and trans-{-

(phen)Pd[CH₂CHCD₂C(CO₂Me)₂CD₂CHCH₂SiEt₃](NCAr)]}⁺[BAr₄]^-

(6-d₄) (see Supporting Information).

For 6: ¹H NMR (CD₂Cl₂, -41 °C): δ 8.86 (dd, J ) 1.0, 5.4 Hz, 1

H, H_a), 8.84 (dd, J ) 1.5, 4.9 Hz, 1 H, H_a), 8.65 (s, 2 H, H_o), 8.64 (dd,

J ) 1.1, 8.6 Hz, 1 H, H_c), 8.55 (dd, J ) 1.5, 8.4 Hz, 1 H, H_c), 8.38 (s,

1 H, H_p), 8.02 (d, J ) 2.7 Hz, 2 H, H_d), 7.99 (dd, J ) 5.4, 8.2 Hz, 1

H, H_b), 7.86 (dd, J ) 4.8, 8.2 Hz, 1 H, H_b), 7.72 (s, 8 H, H_o′), 7.51 (s,

4 H, H_p′), 3.67 (s, 3 H, CO₂CH₃), 3.59 (s, 3 H, CO₂CH₃), 2.93 (dd, J

) 7.4, 13.5 Hz, 1 H, H₃), 2.62 (dd, J ) 6.3, 13.1 Hz, 1 H, H₅), 2.52

(dd, J ) 4.0, 7.7 Hz, 1 H, H₁), 2.18 (dd, J ) 10.4, 13.4 Hz, 1 H, H₃),

2.09 (dd, J ) 8.3, 10.8 Hz, 1 H, H₁), 1.80 (m, 1 H, H₂), 1.77 (dd, J )

11.9, 12.9 Hz, 1 H, H₅), 1.67 (m, 1 H, H₆), 1.26 (dd, J ) 3.0, 13.4 Hz,

1 H, H₇), 0.92 (t, J ) 7.9 Hz, 9 H, -SiCH₂CH₃), 0.55 (q, J ) 7.9 Hz,

6 H, -SiCH₂CH₃), 0.46 (dd, J ) 11.9, 13.6 Hz, 1 H, H₇). ¹³C{¹H}

NMR (CD₂Cl₂, -62 °C): δ 175.4 (CO₂CH₃), 175.2 (CO₂CH₃), 163.8

1

in solution by H NMR spectroscopy at -62 °C. Assignment of the

1

proton resonances and J_HHcoupling constants of 5 was aided by H-

¹H COSY analysis and by ¹H NMR analysis of the deuterated

isotopomers {(phen)Pd[η¹,η²-CD(CH₂SiEt₃)CH₂C(CO₂Me)₂CH₂CDd

CH₂]}⁺[BAr₄]^-(5-d₂) and {(phen)Pd[η¹,η²-CH(CH₂SiEt₃)CD₂C(CO₂-

Me)₂CD₂CHdCH₂]}⁺[BAr₄]^-(5-d₄) (see Supporting Information).

For 5: ¹H NMR (CD₂Cl₂, -62 °C): δ 8.71 (d, J ) 4.4 Hz, 1 H,

H_a), 8.61 (d, J ) 8.1 Hz, 1 H, H_c), 8.48 (dd, J ) 1.2, 8.2 Hz, 1 H, H_c),

8.35 (dd, J ) 1.2, 4.9 Hz, 1 H, H_a), 8.16 (s, 2 H, H_o), 8.14 (s, 1 H, H_p),

8.05 (dd, J ) 5.2, 8.0 Hz, 1 H, H_b), 7.96 (s, 2 H, H_d), 7.87 (dd, J )

4.9, 8.1 Hz, 1 H, H_b), 7.74 (s, 8 H, H_o′), 7.49 (s, 4 H, H_p′), 6.41 (dddd,

J ) 4.3, 8.2, 9.0, 16.1 Hz, 1 H, -CHdCH₂), 5.44 (d, J ) 9.0 Hz, 1 H,

-CHdCH₂), 4.22 (d, J ) 16.1 Hz, 1 H, -CHdCH₂), 3.68 (s, 3 H,

CO₂Me), 3.43 (s, 3 H, CO₂Me), 3.02 (dd, J ) 8.2, 12.7 Hz, 1 H, -CH₂-

CHdCH₂), 2.66 (m, 1 H, Pd-CH), 2.22 (dd, J ) 4.0, 12.6 Hz, 1 H,

-CH₂CHdCH₂), 1.74, 1.70 [ABX, J_AB) 15.6 Hz, J_AX) 10.9 Hz,

J_BX) 5.4 Hz, 2 H, Pd-CH(CH₂SiEt₃)CH₂], 1.23 [dd, J ) 2.6, 13.3

Hz, 1 H, Pd-CH(CH₂SiEt₃)CH₂], 1.11 [t, J ) 13.3 Hz, 1 H, Pd-CH-

(CH₂SiEt₃)], 0.94 (t, J ) 7.9 Hz, 9 H, SiCH₂CH₃), 0.64 (q, J ) 7.9

Hz, 6 H, SiCH₂CH₃).

10

(q, J_C11_B) 49.7 Hz, superimposed on a septet, J_C) 16.8 Hz, C_i),

B

159.4, 150.5, 149.6, 145.4, 142.4, 141.4 (C_phen), 136.6 (C_o′), 136.1 (C_o),

2

134.9 (q, J_CF) 35.1 Hz, C_m), 132.6, 132.0 (C_phen), 131.3 (C_p), 130.6

2

(q, J_CF) 31.4 Hz, C_m′), 129.9, 129.6, 127.9, 127.8 (C_phen), 126.4 (q,

¹J_CF) 273 Hz, C_r′), 124.1 (q, ¹J_CF) 274 Hz, C_r), 121.5, 113.8 (C_iand

C_q), 119.5 (C_p′), 59.2 (C₄), 55.3 (CO₂CH₃), 55.2 (CO₂CH₃), 53.6 (C₂),

45.2, 44.8, 43.9 (C₃, C₅, and C₆) 32.4 (C₁), 16.0 (C₇), 9.5 (SiCH₂CH₃),

5.2 (SiCH₂CH₃).

trans-{(phen)Pd[CH₂CHCH₂¹³C(¹³CO₂Et)₂CH₂CHCH₂SiEt₃]-

(NCAr)}⁺[BAr₄]^-(6a-¹³C₃). Warming a solution of 5a-¹³C₃(4.66 µL,

0.023 mmol) in CD₂Cl₂(0.60 mL) at -41 °C for 2 h formed 6a-¹³C₃

{(phen)Pd[η¹,η²-CH(CH₂SiEt₃)CH₂¹³C(¹³CO₂Et)₂CH₂CHd

CH₂]}⁺[BAr₄]^-(5a-¹³C₃). Reaction of diethyl diallylmalonate ¹³C

labeled at each of the quaternary carbon atoms (1a-¹³C₃, 4.7 µL, 0.023

mmol) and 4b (0.023 mmol) in CD₂Cl₂(0.56 mL) at -62° C, employing

a procedure analogous to that used in the synthesis of 5, gave 5a-¹³C₃

1

as the exclusive product by H NMR analysis. ¹³C{¹H} NMR (CD₂-

Cl₂, -41 °C, labeled carbon atoms only): δ 175.4 (dd, J ) 1.4, 58.4

Hz, CO₂CH₃), 175.2 (dd, J ) 1.4, 58.4 Hz, CO₂CH₃), 59.2 (t, J ) 58.4

Hz, C₄).

1

as the exclusive product by H NMR analysis. ¹³C{¹H} NMR (CD₂-

Cl₂, -62 °C, labeled carbon atoms only): δ 172.6 (dd, J ) 2.3, 55.3

Hz), 172.5 (d, J ) 58.4 Hz), 65.5 (dd, J ) 55.2, 58.4 Hz).

trans-{(phen)Pd[¹³CH₂¹³CH¹³CH₂C(CO₂Me)₂¹³CH₂¹³CH¹³CH₂SiEt₃]-

(NCAr)}⁺[BAr₄]^-(6-¹³C₆). Warming a solution of 5-¹³C₆(0.021

mmol) in CD₂Cl₂(0.60 mL) at -40 °C for 2 h formed 6-¹³C₆as the

{(phen)Pd[η¹,η²-¹³CH(¹³CH₂SiEt₃)¹³CH₂C(CO₂Me)₂¹³CH₂¹³CHd

¹³CH₂]}⁺[BAr₄]^-(5-¹³C₆). Reaction of 1-1,2,3,5,6,7-¹³C₆(4.2 µL, 0.021

mmol) and 4b (0.022 mmol) in CD₂Cl₂(0.56 mL) at -60 °C employing

a procedure analogous to that used to synthesize 5 gave 5-¹³C₆as the

1

exclusive product by H NMR analysis. ¹³C{¹H} NMR (CD₂Cl₂, 75

MHz -40 °C, labeled carbon atoms only): δ [51.8 (q), 43.3 (q), C₂

and C₆)], [42.7 (d), 42.2 (d), 31.0 (d), C₁, C₃and C₅], 14.8 (d, C₇); all

¹J_CC) 31-34 Hz.

1

exclusive product by H NMR analysis. ¹³C{¹H} NMR (CD₂Cl₂, 75

MHz -60 °C, labeled carbon atoms only): δ 103.7 (dd, J ) 41, 47

Hz, -CHdCH₂), 87.5 (d, J ) 47 Hz, -CH)CH₂), 50.0 [dd, J ) 30,

36 Hz, Pd-CH(CH₂SiEt₃)CH₂], 39.2 [d, J ) 36 Hz, Pd-CH(CH₂SiEt₃)-

Kinetics of the Conversion of 5 to 6. An NMR tube containing an

equimolar solution of 5 and NCAr (42 mM) in CD₂Cl₂was warmed at

1

-41 °C and monitored periodically by H NMR spectroscopy. The

9

6344 J. AM. CHEM. SOC. VOL. 126, NO. 20, 2004

Pd-Catalyzed Diene Cyclization/Hydrosilylation

A R T I C L E S

concentration of 5 was determined by integrating the carbomethoxy

resonances of 5 (δ 3.69 and 3.43) versus the para phenyl proton of the

BAr₄counterion (δ 7.49). A plot of ln[5] vs time was linear to >3

revealed complete (g95%) deuteration of the exocyclic methyl group.

¹³C{¹H} NMR (CDCl₃, 23 °C): δ 17.1 (t, J ) 19.1 Hz, isotopic shift

300 ppb).

-

half-lives with a first-order rate constant of k ) 5.04 ( 0.05 × 10^-4

s^-1; two identical experiments gave an average value of k ) 5.2 ( 0.8

× 10^-4s^-1. The first-order rate constant for the conversion of 5 to 6

was determined at -62, -55, -51, and -30 °C (Table 1). A plot of

ln k/T versus 1/T provided the activation parameters for the conversion

of 5 to 6; ∆H^q) 13.5 ( 0.6 kcal mol^-1, and ∆S^q) -15 ( 2 eu

(Figure 3).

Kinetics of the Silylation of 7. A solution of 7 (22 µmol, 36 mM)

and HSiEt₃(0.21 mmol, 0.36 M) in CD₂Cl₂(0.60 mL) was monitored

periodically by H NMR spectroscopy at -14 °C. The concentration

1

of 7 was determined by integrating the resonance corresponding to one

of the carbomethoxy peaks of 7 (δ 4.24) relative to that of the para

-

phenyl resonances of the BAr₄counterion (δ 7.51). A plot of ln[7]

versus time was linear to >4 half-lives (Figure S3), with a pseudo-

first-order rate constant of k_obs) 1.33 ( 0.01 × 10^-4s^-1. Employing

a similar procedure, pseudo-first-order rate constants for the reaction

of HSiEt₃with 7 at -14 °C in CD₂Cl₂were obtained at silane

concentrations of 0.18, 0.20, 0.38, and 0.93 M (Table 2). For reactions

employing initial silane concentrations of 0.18 and 0.20 M, the initial

concentration of 7 was lowered to 20 mM. The resulting plot of k_obs

versus [HSiEt₃] was linear, which provided the second-order rate

constant for silylation of 7 of k ) 3.3 ( 0.3 × 10^-4M^-1s^-1(Figure

4).

trans,trans-{(phen)Pd[CHCH(Me)CH(CH₂SiEt₃)CH₂C(COOMe)-

(COOMe)]}⁺(BAr₄]^-(7). Triethylsilane (12.3 µL, 0.077 mmol) was

added via syringe to a solution of [(phen)Pd(Me)(NCCH₃)]⁺[BAr₄]^-

(2c) (93 mg, 0.077 mmol) in CH₂Cl₂(25 mL) at -78 °C, stirred for 30

min, and treated with 1 (16 µL, 0.077 mmol). The resulting solution

was warmed to room temperature, stirred for 2 h, and concentrated

under vacuum to form a yellow glass that was triturated with hexanes

(3 × 5 mL) and dried under vacuum to give 7 (52 mg, 49%) as a

1

yellow powder. H NMR resonances were assigned on the basis of

Reaction of 7 with DSiEt₃. DSiEt₃(107 µL, 0.67 mmol) was added

via syringe to a solution of 7 (160 mg, 0.13 mmol) in CH₂Cl₂(1.6

mL) at room temperature, and the resulting solution was stirred for 2

h. The resulting black solution was concentrated under vacuum and

chromatographed (SiO₂, hexanes-EtOAc ) 24:1) to give trans,trans-

1,1-dicarbomethoxy-2-deuterio-4-(triethylsilylmethyl)-3-methylcyclo-

pentane (3-d₁′) as a colorless oil in 46% yield with 83% isotopic purity

as determined by MS analysis. The regio- and stereochemistry of 3-d₁′

¹H-¹H COSY analysis; the relative stereochemistry of 7 was assigned

via degradation with DSiEt₃.

For 7: ¹H NMR (CDCl₃, 23 °C): δ 8.92 (dd, J ) 1.4, 4.9 Hz, 1 H,

H_c), 8.85 (dd, J ) 1.4, 5.4 Hz, 1 H, H_c), 8.64 (dd, J ) 1.2, 8.2 Hz, 1

H, H_a), 8.62 (dd, J ) 1.6, 8.4 Hz, 1 H, H_a), 8.07, 8.03 (ABq, J ) 8.9

Hz, 2 H, H_d), 7.95 (dd, J ) 4.8, 8.3 Hz, 1 H, H_b), 7.89 (dd, J ) 5.3,

8.2 Hz, 1 H, H_b), 7.72 (t, J ) 2.4 Hz, 8 H, H_o′), 7.55 (s, 4 H, H_p′), 4.16

(s, 3 H, H₁₃), 3.78 (s, 3 H, H₁₁), 2.85 (dd, J ) 7.1, 13.0 Hz, 1 H, H₄),

2.47 (d, J ) 10.5 Hz, 1 H, H₁), 1.70 (m, 1 H, H₃), 1.67 (d, J ) 12.8

Hz, 1 H, H₄), 1.47 (m, 1 H, H₂), 1.20 (d, J ) 6.5 Hz, 3 H, H₆), 1.06

(dd, J ) 2.4, 14.4 Hz, 1 H, H₇), 0.97 (t, J ) 7.9 Hz, 9 H, H₉), 0.58 (q,

J ) 7.8 Hz, 6 H, H₈), 0.37 (dd, J ) 11.2, 14.4 Hz, 1 H, H₇). ¹³C{¹H}

1

was determined via H- and ¹³C NMR analysis and by comparison to

the spectroscopy of unlabeled 3; complete assignment of the ¹H NMR

resonances of 3 was achieved via combined ¹H-¹H COSY and NOESY

analysis. ¹³C{¹H} NMR (CD₂Cl₂, 23 °C): δ 42.0 [t, J_CD) 20.6 Hz,

isotopic shift ) 320 ppb, C(1)], 58.3 [s, isotopic shift ) 60 ppb, C(5)],

43.8 [s, isotopic shift ) 100 ppb, C(2)].

NMR (CDCl₃, 23 °C): δ 191.5 (C₁₂), 171.7 (C₁₀), 161.9 (q, J_CB

)

50.0 Hz, C_i′), 151.7, 149.1 (C_a), 147.8, 143.8 (C_e), 139.7, 139.1 (C_b),

Silane Competition Experiments. A mixture of HSiEt₃(4.95 mmol)

and HSiMe₂R (R ) Et, n-octyl, OSiMe₃, or Ph) (4.95 mmol) was added

via syringe to a solution of 1 (100 µL, 0.50 mmol) and a catalytic

amount of 2b (12 µmol) in DCE (10 mL) at 0 °C. The resulting solution

was stirred at room temperature until 1 was completely consumed (as

determined by GC analysis) to form a mixture of silylated carbocycles

3 and 3a-3d, respectively. The ratio of carbocycles was determined

by gas chromatography, correcting for the GC response factors of the

carbocycles (Table 3). A second set of experiments employing a 1:1

mixture of HSiMe₂Ph and HSiMe₂(4-C₆H₄R) (R ) NMe₂, OMe, CF₃,

F) to form a mixture of silylated carbocycles 3d and 3e-3h was

performed in an analogous manner (Table 3).

2

134.9 (C_o′), 130.9, 130.2 (C_f), 129.2 (q, J_CF) 31.7 Hz, C_m′), 128.0,

127.5, 126.0, 125.4 (C_band C_c), 124.6 (q, ¹J_CF) 272.8 Hz, C_r′), 117.6

(C_p′), 68.5, 56.5, 53.8, 53.2, 42.1, (C₁, C₂, C₅, C₁₁, and C₁₃), 42.9, 40.9

(C₃and C₄), 17.3 (C₆), 15.8 (C₇), 7.5 (C₉), 4.0 (C₈). IR (KBr, cm^-1):

1746, 1612 (ν_CdO). Anal. Calcd (found) for C₆₁H₅₁N₂O₄F₂₄PdSiB: H,

3.48 (3.14); C, 49.59 (49.47); N, 1.90 (2.02).

Reaction of 5 with HSiEt₃. A solution of 5 (23 µmol, 43 mM) in

CD₂Cl₂(0.53 mL) was treated with HSiEt₃(10.8 µL, 67.7 µmol) at

-80 °C, shaken briefly, and placed in the probe of an NMR

spectrometer pre-cooled at -81 °C. The NMR probe was warmed at

-51 °C, and the solution was monitored periodically by ¹H NMR

spectroscopy. The concentration of 5 was determined by integrating

the carbomethoxy resonances of 5 (δ 3.69 and 3.43) versus that of the

Catalytic Cyclization/Deuteriosilylation. A solution of 1 (100 µL,

0.50 mmol), (phen)PdMeCl (8 mg, 24 µmol), NaBAr₄(23 mg, 26

µmol), and DSiEt₃(300 µL, 1.9 mmol) in DCE (10 mL) was stirred at

room temperature for 30 min. The resulting dark solution was

concentrated and chromatographed (SiO₂, hexanes-EtOAc ) 24:1) to

give 3-d₁(167 mg, 102%) as a colorless oil. Mass spectral analysis

-

para phenyl proton of the BAr₄counterion (δ 7.49). A plot of ln[5]

versus time was linear to 3 half-lives with a first-order rate constant of

k ) 1.59 ( 0.01 × 10^-4s^-1(Figure S2).

Reaction of 6 with HSiEt₃. A solution of 6 (∼19 mmol, ∼30 mM)

in CD₂Cl₂(0.60 mL) was treated with HSiEt₃(30 µmol, 50 mM) at

-78 °C, shaken briefly, and placed in the probe of an NMR

spectrometer cooled at -80 °C. The NMR probe was warmed at -41

°C, and the solution was monitored periodically by ¹H NMR. After 10

min, resonances corresponding to 6 could no longer be detected (g95%

conversion), and resonances corresponding to a 1:1 mixture of 4b and

3 were observed.

1

indicated an isotopic purity of 98%, and H- and ¹³C NMR analysis

revealed complete (g95%) deuteration of the exocyclic methyl group.

Kinetics of Catalytic Cyclization/Hydrosilylation. CD₂Cl₂(0.58

mL), NCAr (2.98 µL, 17.7 µmol), and 1 (10.0 µL, 49.5 µmol) were

added sequentially via syringe to a 5-mm NMR tube containing 2b

(9.9 mg, 7.0 µmol) that was sealed with a septum. After the tube was

shaken thoroughly and cooled at -81 °C, HSiEt₃(16 µL, 100 µmol)

was added via syringe and the tube was placed in the probe of an NMR

spectrometer pre-cooled at -80 °C. The sample was then warmed at

Reaction of 6 with DSiEt₃. DSiEt₃(5 µL, 31 µmol) was added via

syringe to an NMR tube containing a solution of 6 (19 µmol) in CD₂-

Cl₂(0.60 mL) at -80 °C. The tube was shaken briefly and placed in

the probe of an NMR spectrometer pre-cooled at -41 °C. Upon

complete consumption of 6 (∼10 min), the sample was removed from

the probe, and the solution was filtered through a plug of Celite and

concentrated under vacuum to give 3-d₁. Mass spectral analysis of 3-d₁

1

-41 °C and monitored periodically by H NMR spectroscopy. The

concentration of 1 was determined by integrating the olefinic resonances

of 1 at δ 5.08 and 5.11 relative to the aryl para hydrogen resonance of

-

BAr₄at δ 7.52. A plot of [1] versus time was linear to 3 half-lives

(Figure 6), with an observed rate constant of k_obs) 8.20 ( 0.04 ×

1

indicated an isotopic purity of 98%, and H- and ¹³C NMR analysis

10^-6M s^-1; two identical experiments gave an average value of k_obs

)

9

J. AM. CHEM. SOC. VOL. 126, NO. 20, 2004 6345

A R T I C L E S

Perch and Widenhoefer

8.3 ( 0.6 × 10^-6M s^-1(Table 4, entries 1-3). Resonances attributed

to the olefinic protons [δ 6.41 (dtd, J ) 4.3, 8.4, 16.1 Hz, 1 H, -CHd

CH₂), 5.44 (d, J ) 9.0 Hz, 1 H, -CHdCH₂), 4.22 (d, J ) 16.1 Hz, 1

H, -CHdCH₂)] and carbomethoxy protons [3.68 (s, 3 H), and 3.43

(s, 3 H)] of 5 were observed throughout complete conversion of 1 to

3. Pseudo-zero-order rate constants for the reaction of 1 ([1]₀) 85

mM) and HSiEt₃([HSiEt₃]₀) 0.16 M) catalyzed by 2b were determined

as a function of [NCAr], [2b], and temperature, employing a similar

procedure (Table 4). The first-order rate constant for the catalytic

cyclization/hydrosilylation of 1 ([1]₀) 85 mM) and HSiEt₃([HSiEt₃]₀

) 0.16 M) at -41 °C was determined from the linear plot of pseudo-

zero-order rate constants versus [2b]₀over the range 0-25 mM (Figure

7). The activation parameters for the reaction of 1 ([1]₀) 85 mM) and

HSiEt₃([HSiEt₃]₀) 0.16 M) catalyzed by 2b ([2b]₀) 12 mM) were

obtained from a plot of ln k/T versus 1/T over the range -57 to -25

°C (Figure 8). The dependence of the rate of catalytic cyclization/

hydrosilylation on silane concentration was determined via iterative

fitting of two sets of concentration versus time data for reaction of 1

([1]₀) 85 mM), 2b ([2b]₀) 12 mM), and HSiEt₃([HSiEt₃]₀) 90

mM) in CD₂Cl₂that contained NCAr (42 mM) at -41 °C to the rate

equation depicted in eq 2 employing MacKinetics (version 0.9.1b). This

analysis provided a best fit with k₂) 7.6 ( 0.2 × 10^-4s^-1and k₃)

0.12 ( 0.03 M^-1s^-1. To estimate the error limit for k₃, k₂was held

constant (k₂) 7.6 × 10^-4s^-1), k₃was incrementally varied by (0.01

M^-1s^-1, and the resulting concentration versus time plots were

compared visually to the experimental data. The error limit for k₂

determined via iterative fitting was estimated by employing an

analogous procedure.

Acknowledgment is made to the NIH (GM59830-01) and

PRF (36399-AC4), administered by the American Chemical

Society, for support of this research. We thank Mr. Jeff Cross

(UNC-Chapel Hill) for obtaining the ¹³C NMR spectra of 5-¹³C₆

and 6-¹³C₆, and Dr. Tao Pei for synthesizing 1-1,2,3,5,6,7-¹³C₆

and 1-3,3,5,5-d₄.

Supporting Information Available: Syntheses and spectral

data for 1a-¹³C₃, 1-1,2,3,5,6,7-¹³C₆, 5-d₂, 5-d₄, 6-d₂, and 6-d₄,

spectral and analytical data for 3b, 3e, 3f, 3g, and 3h, and

representative kinetic plots for reaction of 5 and 7 with HSiEt₃

and for the conversion of 6 to 7. This material is available free

of charge via the Internet at http://pubs.acs.org.

JA049806F

9

6346 J. AM. CHEM. SOC. VOL. 126, NO. 20, 2004

Article Doi

Mechanism of palladium-catalyzed diene cyclization/hydrosilylation: Direct observation of intramolecular carbometalation

DOI: 10.1021/ja049806f

Source and publish data:

Authors:

Article abstract of DOI:10.1021/ja049806f

Full text of DOI:10.1021/ja049806f

Products guided by the article

R&D Labs maybe for 388567-20-2

Relevant to this article

Hot Product