Mechanism of the Stille reaction. 1. The transmetalation step. Coupling of R1I and R2SnBu3 catalyzed by trans-[PdR1IL2] (R1 = C6Cl2F3; R2 = vinyl, 4-methoxyphenyl; L = AsPh3)

Article abstract of DOI:10.1021/ja9742388

The so far accepted mechanism of the Stille reaction (palladium- catalyzed cross-coupling of organotin reagents with organic electrophiles) is criticized. Based on kinetic studies on catalytic reactions, and on reactions with isolated intermediates, a corrected mechanism is proposed. The couplings between R¹I (1) (R¹ = C₆-Cl₂F₃ = 3,5-dichlorotrifluorophenyl) and R²SnBu₃ (R² = CH=CH₂, 2a; C₆H₄-4-OCH₃, 2b), catalyzed by trans- [PdR¹I(AsPh₃)₂] (3a), give R¹-R² and obey a first-order law, r(obs) = a[3a][2a]/(b + [AsPh₃]), with a = (2.31 ± 0.09) x 10^-5 s^-1 and b = (6.9 ± 0.3) x 10^-4 mol L^-1, for [1] = [2a] = 0-0.2 mol L^-1, [3a] = 0-0.02 mol L^-1, and [AsPh₃] = 0-0.07 mol L^-1, at 322.6 K in THF. The only organopalladium(II) intermediate detected under catalytic conditions is 3a. The apparent activation parameters found for the coupling of 1 with 2a support an associative transmetalation step (ΔH((±))(obs) = 50 ± 2 kJ mol^-1, ΔS((±))(obs) = -155 ± 7 J K^-1 mol^-1 in THF; and ΔH((±))(obs) = 70.0 ± 1.7 kJ mol^-1, ΔS((±))(obs) = -104 ± 6 J K^-1 mol^-1 in chlorobenzene, with [1]₀ = [2]₀ = 0.2 mol L^-1, [3a] = 0.01 mol L^-1). The reactions of 2a with isolated trans-[PdR¹X(AsPh₃)₂ (X = halide) show rates Cl > Br > I. From these observations, the following mechanism is proposed: Oxidative addition of R¹X to PdL(n) gives cis- [PdR¹XL₂], which isomerizes rapidly to trans[PdR¹XL₂]. This trans complex reacts with the organotin compound following a S(E)² (cyclic) mechanism, with release of AsPh₃ (which explains the retarding effect of the addition of L), to give a bridge intermediate [PdR¹L(μ-X)(μ-R²)SnBu₃]. In other words, an L-for-R² substitution on the palladium leads R² and R¹ to mutually cis positions. From there the elimination of XXnBu₃ yields a three- coordinate species cis-[PdR¹R²L], which readily gives the coupling product R¹-R².

Full text of DOI:10.1021/ja9742388

8978
J. Am. Chem. Soc. 1998, 120, 8978-8985
Mechanism of the Stille Reaction. 1. The Transmetalation Step.
Coupling of R1I and R2SnBu3 Catalyzed by trans-[PdR1IL2] (R1 )
C6Cl2F3; R2 ) Vinyl, 4-Methoxyphenyl; L ) AsPh3)
Arturo L. Casado and Pablo Espinet*
Contribution from the Departamento de Qu´ımica Inorga´nica, Facultad de Ciencias,
UniVersidad de Valladolid, E-47005 Valladolid, Spain
ReceiVed December 15, 1997
Abstract: The so far accepted mechanism of the Stille reaction (palladium-catalyzed cross-coupling of organotin
reagents with organic electrophiles) is criticized. Based on kinetic studies on catalytic reactions, and on reactions
with isolated intermediates, a corrected mechanism is proposed. The couplings between R1I (1) (R1 ) C6-
Cl2F3 ) 3,5-dichlorotrifluorophenyl) and R2SnBu3 (R2 ) CHdCH2, 2a; C6H4-4-OCH3, 2b), catalyzed by
trans-[PdR1I(AsPh3)2] (3a), give R1-R2 and obey a first-order law, robs ) a[3a][2a]/(b + [AsPh3]), with a )
(2.31 ( 0.09) × 10-5 s-1 and b ) (6.9 ( 0.3) × 10-4 mol L-1, for [1] ) [2a] ) 0-0.2 mol L-1, [3a] )
0-0.02 mol L-1, and [AsPh3] ) 0-0.07 mol L-1, at 322.6 K in THF. The only organopalladium(II) intermediate
detected under catalytic conditions is 3a. The apparent activation parameters found for the coupling of 1 with
2a support an associative transmetalation step (∆Hq ) 50 ( 2 kJ mol-1, ∆Sq ) -155 ( 7 J K-1 mol-1
obs
obs
in THF; and ∆Hq ) 70.0 ( 1.7 kJ mol-1, ∆Sq ) -104 ( 6 J K-1 mol-1 in chlorobenzene, with [1]0 )
obs
obs
[2]0 ) 0.2 mol L-1, [3a] ) 0.01 mol L-1). The reactions of 2a with isolated trans-[PdR1X(AsPh3)2] (X )
halide) show rates Cl > Br > I. From these observations, the following mechanism is proposed: Oxidative
addition of R1X to PdLn gives cis-[PdR1XL2], which isomerizes rapidly to trans-[PdR1XL2]. This trans complex
2
reacts with the organotin compound following a SE (cyclic) mechanism, with release of AsPh3 (which explains
the retarding effect of the addition of L), to give a bridged intermediate [PdR1L(µ-X)(µ-R2)SnBu3]. In other
words, an L-for-R2 substitution on the palladium leads R2 and R1 to mutually cis positions. From there the
elimination of XSnBu3 yields a three-coordinate species cis-[PdR1R2L], which readily gives the coupling product
R1-R2.
Introduction
Scheme 1
The palladium-catalyzed cross-coupling of organotin reagents
with organic electrophiles (Stille reaction) has become an
attractive method in modern organic synthesis,1 mainly due to
the advantages of using trialkylorganotin species: They are
readily available and quite air and moisture stable and tolerate
many functional groups.2 The broadly accepted mechanism is
the catalytic cycle in Scheme 1.1,3 Both the oxidative addition
and the reductive elimination steps are supposed to be fast,
compared to the Sn/Pd transmetalation,4,5 which is the rate-
determining step.6,7 This proposal fits the observations that the
reaction rate is zeroth order in electrophile (the oxidative
addition must be fast) and first order in stannane.
(1) (a) Stanforth, S. P. Tetrahedron 1998, 54, 263-303. (b) Farina, V.;
Roth, G. P. AdV. Metalorg. Chem. 1996, 5, 1-53. (c) Curran, D. P.; Hoshino,
M. J. Org. Chem. 1996, 61, 6480-6481. (d) Mateo, C.; Ca´rdenas, D. J.;
Ferna´ndez-Rivas, C.; Echavarren, A. M. Chem. Eur. J. 1996, 2, 1596-
1606. (e) Roth, G. P.; Farina, V.; Liebeskind, L. S.; Pen˜a-Cabrera, E.
Tetrahedron Lett. 1995, 36, 2191-2194. (f) Mitchell, T. N. Synthesis 1992,
803-815. (g) Scott, W. J.; Stille, J. K. J. Am. Chem. Soc. 1986, 108, 3033-
3040. (h) Stille, J. K. Angew. Chem., Int. Ed. Engl. 1986, 25, 508-524. (i)
Beletskaya, I. P. J. Organomet. Chem. 1983, 250, 551-564.
(2) Lee, A. S.-Y.; Dai, W.-C. Tetrahedron 1997, 53, 859-868, and
references therein.
Scheme 2a
a R1 ) C6Cl2F3, L ) PPh3.
(3) (a) Farina, V. In ComprehensiVe Organometallic Chemistry II; Abel,
E. W., Stone, F. G. A., Wilkinson, G., Eds.; Pergamon: Oxford, U.K., 1995;
Vol. 12, Chapter 3.4. (b) Brown, J. M.; Cooley, N. A. Chem. ReV. 1988,
88, 1031-1046.
(4) Studies on oxidative additions: (a) Amatore. C.; Jutand, A.; Suarez,
A. J. Am. Chem. Soc. 1993, 115, 9531-9541. (b) Amatore, C.; Pflu¨ger, F.
Organometallics 1990, 9, 2276-2282. And references therein.
We have proved recently that the oxidative addition of R1I
to PdL2 is not as simple as it appears in Scheme 1. It gives
first cis-[PdR1IL2] (I), which then isomerizes to trans-[PdR1-
IL2] (II) in a reaction autocatalyzed by both isomers (Scheme
2).8
S0002-7863(97)04238-8 CCC: $15.00 © 1998 American Chemical Society
Published on Web 08/21/1998

Mechanism of the Stille Reaction
J. Am. Chem. Soc., Vol. 120, No. 35, 1998 8979
Scheme 3
one would expect the process to be independent of the
concentration and nature of the stannane, against what is
observed.
Thus, the two plausible mechanisms initiated by a dissociation
of L (whether fast or slow) are inconsistent with the data in the
literature and must be discarded.
Other obscure points reveal that further studies are needed.
In effect, it is not clear that the trans-to-cis isomerizations in
[PdR2L2] complexes are fast. On the contrary, the theoretical
paper usually cited to support this assumption states literally,
“T-shaped trans-[PdR2L], arising from dissociation of L in
[PdR2L2], will encounter a substantial barrier to polytopal
rearrangement to cis-[PdR2L]”.5b Actually, the isomerizations
studied in isolated [PdR2L2] complexes are slow5c or extremely
slow.14 Thus, intermediates of the type trans-[PdR1R2L2] (A)
might be expected to be quite long-lived, but they have never
been detected under catalytic conditions.5g This warns, in our
opinion, against a cursory acceptance of an I-for-R2 substitution
with preserVation of the configuration at the palladium.
Finally, substitution reactions in palladium involving initial
L dissociation are a rarity.15 Since associative models are
perfectly compatible with an eventual neutral ligand dissociation,
they should not be discarded a priori on the basis of the observed
L retarding effect. On the contrary, measurements of activation
parameters, not available so far, seem convenient in order to
better decide which mechanism is more consistent with the
observations.
In the same paper where they proposed the formation of a
three-coordinate intermediate as the general mechanism for the
transmetalation with organotin compounds, Louie and Hartwig
remarked, “Moreover, dissociative ligand substitution typically
occurs by initial loss of the covalent ligand that is being replaced.
It is striking that transmetalation reactions involving organotin
reagents are dissociative and even more unusual that it is a dative
spectator ligand that undergoes dissociation”.10 These puzzling
questions disappear in the light of an associative ligand
substitution of L (which is not an spectator ligand anymore) as
the rate-determining step. Thus, we have considered the
alternative cycle proposed in Scheme 4, which in our opinion
solves all the inconsistencies just analyzed. Differently from
the proposals in the literature, the transmetalation involves an
associative L-for-R2 substitution, which gives directly a cis R1/
R2 rather than a trans R1/R2 arrangement in IV, and therefrom
the cis T-shaped V, from which the coupled product will
immediately be eliminated.
Also, the transmetalation step (framed in Scheme 1) is still
not well understood. Although there is no evidence, it is
generally thought to preserve the trans configuration of complex
II to give a trans-[PdR1R2L2] complex (A), as it occurs in the
case of main-group organometallic transmetalations.9 Since the
reductive elimination of R1-R2 is well established to occur on
cis derivatives,5 a fast isomerization of A to B needs to be
postulated.
Attempts at gaining insight into the transmetalation step have
shown that the addition of neutral ligand L retards the
coupling,1b,10,11 and this has been taken as an indication that L
dissociation from II is a key step in the transmetalation. Thus,
the mechanism in Scheme 3, involving a dissociatiVe X-for-R2
substitution (X ) I, Br) with preservation of the stereochemistry
at the Pd, has been proposed for vinyl- and arylstannanes. It is
assumed that II cannot undergo transmetalation, probably
because it is too electron rich, and a ligand dissociation occurs
previous to the transmetalation; it is the more coordinatively
unsaturated species C (most likely having a coordinated solvent
molecule, S) that is involved in the electrophilic substitution at
tin.
This proposal is qualitatiVely consistent with the observa-
tions: The existence of a (fast) preequilibrium explains the
retarding effect of L, whereas the (slow) transmetalation on C
explains the first-order dependence on the stannane. However,
it is stated in the literature1b that the equilibrium constant for
the dissociation of II in Scheme 3 (for R1 ) C6H5, L ) AsPh3,
THF at 323 K) is Kdis ) (k′1/k′-1) ) 8.6 × 10-4 mol L-1. From
this we calculate 40% dissociation in the experimental condi-
tions,12 a value impossible to accept.13 Thus this mechanism
is inconsistent with the quantitative results.
Our proposal is consistent with the observations in the
literature and is supported by a detailed kinetic study of the
Stille coupling between 1-iodo-3,5-dichlorotrifluorobenzene (C6-
Cl2F3I) and vinyl- or 4-methoxyphenyltributyltin, catalyzed by
trans-[Pd(C6Cl2F3)I(AsPh3)2]. Furthermore, it offers a plausible
picture of the kind of bonding interactions leading from the
reagents to the products (see later) and eliminates the need for
unlikely fast trans-to-cis isomerization after the transmetalation
step.
Alternatively, if the L dependence was attributed to the
dissociation step in Scheme 3 being slow and rate determining,
(5) Studies on reductive eliminations: (a) Ozawa, F.; Fujimori, M.;
Yamamoto, T.; Yamamoto, A. Organometallics 1986, 5, 2144-2149. (b)
Tatsumi, K.; Hoffmann, R.; Yamamoto, A.; Stille, J. K. Bull. Chem. Soc.
Jpn. 1981, 54, 1857-1867. (c) Ozawa, F.; Ito, T.; Nakamura, Y.;
Yamamoto, A. Bull. Chem. Soc. Jpn. 1981, 54, 1868-1880. (d) Moravski,
A.; Stille, J. K. J. Am. Chem. Soc. 1981, 103, 4182-4186. (e) Loar, M. K.;
Stille, J. K. J. Am. Chem. Soc. 1981, 103, 4174-4181. (f) Ozawa, F.; Ito,
T.; Yamamoto, A. J. Am. Chem. Soc. 1980, 102, 6457-6463. (g) Gillie,
A.; Stille, J. K. J. Am. Chem. Soc. 1980, 102, 4933-4941. (h) Komiya, S.;
Albright, T. A.; Hoffmann, R.; Kochi, J. K. J. Am. Chem. Soc. 1976, 98,
7255-7265.
(12) Reference 1b, p 10, gives a dissociation constant at 50 °C in THF:
Kdis ) 8.6 × 10-4 mol L-1. The catalyst concentration was [Pd]total ) 3.2
(6) Labadie, J. W.; Stille, J. K. J. Am. Chem. Soc. 1983, 105, 6129-
6137.
× 10-3 mol L-1
: The concentration of three-coordinate trans-[PdPhI-
(AsPh3)] can be calculated from Kdis ) [PdPhI(AsPh3)][AsPh3]/[PdPhI-
(AsPh3)2] ) x2/[Pd]total - x. This gives [PdPhI(AsPh3)] ) x ) 1.3 × 10-3
mol L-1, corresponding to 40% dissociation.
(7) For transmetalation with platinum complexes, see: (a) Eaborn, C.;
Odell, K. J.; Pidcock, A. J. Chem. Soc., Dalton Trans. 1978, 357-368. (b)
Eaborn, C.; Odell, K. J.; Pidcock, A. J. Chem. Soc., Dalton Trans. 1979,
758-760. (c) Deacon, G. B.; Gatehouse, B. M.; Nelson-Reed, K. T. J.
Organomet. Chem. 1989, 359, 267-283.
(8) Casado, A. L.; Espinet, P. Organometallics 1998, 17, 954-959.
(9) Parshall, G. W. J. Am. Chem. Soc. 1974, 96, 2360-2366.
(10) Louie, J.; Hartwig J. F. J. Am. Chem. Soc. 1995, 117, 11598-11599.
(11) Farina, V.; Krishnan, B. J. Am. Chem. Soc. 1991, 113, 9585-9595.
(13) For instance, for complexes cis-[PdR2L2] (R ) C6F5, C6F3Cl2; L )
tetrahydrothiophene, an easily dissociable ligand), we have estimated the
dissociation of L as 0.13% for a solution 8 × 10-4 mol L-1 in the complex.
See ref 14b.
(14) (a) Minniti, D. Inorg. Chem. 1994, 33, 2631-2634. (b) Casado, A.
L.; Casares, J. A.; Espinet, P. Organometallics 1997, 16, 5730-5736.
(15) Cross, R. J. AdV. Inorg. Chem. 1989, 34, 219-292.

8980 J. Am. Chem. Soc., Vol. 120, No. 35, 1998
Casado and Espinet
Scheme 4
Figure 1. 19F NMR (282 MHz, F4 region) spectra sequence (intervals
in hours) of the coupling of C6Cl2F3I (1, 0.2 mol L-1) with (CH2d
CH)SnBu3 (2a, 0.2 mol L-1) catalyzed by trans-[Pd(C6Cl2F3)I(AsPh3)2]
(3a, 0.01 mol L-1) and AsPh3 (0.02 mol L-1), in THF at 322.6 K. The
product is C6Cl2F3(CHdCH2) (4a).
Scheme 5a
make measurements from the beginning, as it eliminates the
delays needed in other cases, when the catalysts are formed in
situ from Pd(0) precursors.1,11 Some decomposition to Pd(0)
(less than 10% mol after 50-60% conversion) was observed
in couplings with 2a carried out without addition of free AsPh3.
This affected the linearity of the plots, and in these cases, we
measured the initial reaction rate r0 () -d[2a]/dt), from which
first-order constants were estimated as kobs ) r0/[2a]0 (see
Experimental Section).
a R1 ) C6Cl2F3, tht ) tetrahydrothiophene.
Results
Preparation of the Catalyst. The catalyst trans-[Pd(C6-
Cl2F3)I(AsPh3)2] (3a) was prepared in quantitative yield as
shown in Scheme 5.16 Under the standard conditions used in
catalytic couplings (10-2 mol L-1 in THF at 323 K), no
detectable dissociation was found by 19F NMR.17
The kinetic results obtained from catalytic couplings between
1 and 2a in THF at 322.6 K have been analyzed as follows in
order to obtain the experimental rate law.
1. Kinetic Order with Respect to the Reagents. In
agreement with previous results in the literature,1b,g,10,11 the
reaction rate is clearly independent of the electrophile concen-
tration, [1]: A similar value r0 ) 6.8 × 10-5 mol L-1 s-1 (kobs
) 3.4 × 10-4 s-1) was found for [1] ) 0.1 or 0.2 mol L-1
(with [2a] ) 0.2 mol L-1, [3a] ) 0.01 mol L-1 in THF at 322.6
K). Hence, the overall first order observed is assigned to [2a]
and reveals a behavior typical of the Stille reaction. Thus, an
experiment using [2a] ) 0.092 mol L-1 (other conditions the
same as before) resulted in r0 ) 3.2 × 10-5 mol L-1 s-1 (kobs
) 3.5 × 10-4 s-1).
Catalytic Studies. The couplings of C6Cl2F3I (1) with R2-
SnBu3 (R2 ) CHdCH2, 2a; or C6H4-4-OCH3, 2b) catalyzed
by 3a were monitored by 19F NMR (eq 1, Figure 1). The
(1)
2. Retardation by Addition of Free Neutral Ligand. The
reaction rate is minus first order with respect to [AsPh3] (the
slope of ln(kobs) vs ln[AsPh3] is -1.1). Then, a good linear
products R2C6Cl2F3 (R2 ) CHdCH2, 4a; or C6H4-4-OCH3, 4b)
were formed in up to 95% yield (∼1% of (C6Cl2F3)SnBu3 (2c)
was also detected after long reaction periods). Vinyltributyltin
2a reacts much faster than 4-methoxytributyltin, 2b. Both
couplings are retarded by addition of AsPh3. The catalyst 3a
was the only organopalladium(II) species detected under
catalytic conditions.18
-1
dependence kobs vs [AsPh3] is observed, with a slope (4.31
( 0.17) × 106 mol-1 L s in THF at 322.6 K (Figure 2).
3. Catalyst Activity. The kinetic order with respect to the
catalyst concentration [3a] was determined in the presence of
added AsPh3 (0.02 mol L-1). The slope of ln(kobs) vs ln[3a] is
1.0, indicating a first-order dependence with respect to [3a].
Accordingly, the experimental values fit very well a straight
line kobs vs [3a], the slope being (1.16 ( 0.05) × 10-3 mol-1
L s-1 (Figure 3a).
Kinetic Studies. The couplings of C6Cl2F3I (1) with 2a or
2b were followed by monitoring the disappearance of 1 by 19
F
NMR. The reactions followed first-order kinetics, providing
straight lines ln([1]0 - [1]) ) kobst.19 Since the concentration
of 1 is stoichiometrically linked to that of 2a,b, the consum-
mation of the latter necessarily obey a first-order law as well.
The use of 3a (the actual transmetalation catalyst) allows us to
Numerical analysis of the kinetic data leads to the rate law
given in eq 2, with a ) (2.32 ( 0.09) × 10-5 s-1 and b ) (6.9
a[3a]
(16) Espinet, P.; Martinez-Ilarduya, J. M.; Pe´rez-Briso, C.; Casado, A.
L.; Alonso, M. A. J. Organomet. Chem. 1998, 551, 9-20.
robs ) kobs[2a] )
[2a]
(2)
[AsPh3] + b
(17) The integration of the signals agreed with that of the added internal
standard, 1,3,5-C6Cl3F3, and no shift was observed upon successive additions
of AsPh3. This discounts the existence of an appreciable percentage of the
three-coordinate C in fast equilibrium with 3a. In that case, the observed
chemical shifts of a solution of 3a, would be those averaged between C
and 3a, but with an appreciable value for Keq, they should move noticeably
toward those of 3a upon addition of AsPh3, following the displacement of
the equilibrium. Thus, the percentage of dissociation must be very small.
(18) When 3a was replaced by trans-[PdCl2(AsPh3)2], the latter gave
quantitatively 3a in a short time.
(19) A selection of kinetics and mechanisms books is given: (a)
Espenson, J. H. Chemical Kinetics and Reaction Mechanisms, 2nd ed.;
McGraw-Hill: Singapore, 1995. (b) Jordan, R. B. Reaction Mechanisms
of Inorganic and Organometallic Systems; Oxford University Press: New
York, 1991. (c) Katakis, D., Gordon, G. Mechanisms of Inorganic Reactions;
John Wiley & Sons: New York, 1987. (d) Wilkins, R. G. Kinetics and
Mechanism of Reactions of Transition Metal Complexes, 2nd ed.; VCH:
New York, 1991.

Mechanism of the Stille Reaction
J. Am. Chem. Soc., Vol. 120, No. 35, 1998 8981
Figure 4. Eyring plots for the coupling of C6Cl2F3I (1, 0.2 mol L-1
)
Figure 2. Retarding effect of the addition of AsPh3 on the coupling
and R2SnBu3 (2, 0.2 mol L-1) catalyzed by trans-[Pd(C6Cl2F3)I(AsPh3)2]
of C6Cl2F3I (1, 0.2 mol L-1) and (CH2dCH)SnBu3 (2a, 0.2 mol L-1
)
(3a, 0.01 mol L-1): (a) THF, R2 ) vinyl; (b) PhCl, R2 ) vinyl; (c)
catalyzed by trans-[Pd(C6Cl2F3)I(AsPh3)2] (3a, 0.01 mol L-1) in THF
THF, R2 ) 4-anisyl; (d) THF, R2 ) vinyl, [AsPh3] ) 0.02 mol L-1
.
at 322.6 K.
Table 1. Apparent Activation Parameters for the Coupling of
C6Cl2F3I (1, 0.2 mol L-1) with R2SnBu3 (2, 0.2 mol L-1) Catalyzed
by trans-[Pd(C6Cl2F3)I(AsPh3)2] (3a, 0.01 mol L-1
)
R2
solvent
∆Hqobs/kJ mol-1
∆Sqobs/J K-1 mol-1
vinyl
THF
THF
PhCl
THF
50 ( 2
-155 ( 7
-56 ( 15
-104 ( 6
-100 ( 3
vinyla
vinyl
91 ( 4
70.0 ( 1.7
72.8 ( 1.1
4-anisyl
a With [AsPh3] ) 0.02 mol L-1
.
Table 2. Reactions of trans-[Pd(C6Cl2F3)X(AsPh3)2] (0.01 mol
L-1) and (CH2dCH)SnBu3 (2a, 0.2 mol L-1) at 20 °C in THF:
Conversion (%) to CH2dCHsC6Cl2F3 (4a)
Figure 3. kobs vs [3a] plot for the coupling of C6Cl2F3I (1, 0.2 mol
L-1) and (CH2dCH)SnBu3 (2a, 0.2 mol L-1) catalyzed by trans-[Pd(C6-
Cl2F3)I(AsPh3)2] (3a) in THF at 322.6 K: (a) with AsPh3 (0.02 mol
L-1); (b) without AsPh3.
X (complex)
% after 2.3 h
% after 10 h
I (3a)
20
58
90
51
37
92
100
100
Br (3b)
Cl (3c)
Cl (3c)a
( 0.3) × 10-4 mol L-1. The same value of coefficient a was
obtained from Figure 2 and from Figure 3a, supporting the
consistency of our data.20
Although it is not relevant to the catalytic conditions, where
ratios L:Pd > 2:1 are used, it can be noted that in absence of
added AsPh3 (L:Pd ) 2:1) the increase of rate with the catalyst
concentration is not linear (Figure 3b).21
a With [AsPh3] ) 0.02 mol L-1
.
Reactions of Isolated Organopalladium(II) Complexes
with Organotributyltins. The influence of the halide on the
transmetalation rate was studied on reactions using isolated
palladium complexes. Reactions of trans-[Pd(C6Cl2F3)X-
(AsPh3)2] (X ) I, 3a; Br, 3b; Cl, 3c) with (CH2dCH)SnBu3
(2a) in a 1:20 ratio (similar to that used in catalytic reactions)
gave the coupled product CH2dCH-C6Cl2F3 (4a) and were
monitored at similar times in order to have an approximate
estimation of their relative rates (eq 3). The rate depended on
4. Activation Parameters. The temperature dependence of
the rate was examined within the range 295-328 K (lower
temperatures gave very low rates, difficult to measure; the upper
limit is imposed by the boiling point of the solvent). Eyring
plots (Figure 4) provided the apparent activation parameters
given in Table 1.19,22 The apparent activation entropy ∆Sq
obs
is very negative (ranging from -56 to -155 J K-1 mol-1
)
regardless of the presence or not of added neutral ligand AsPh3,
the type of organotributyltin reagent, or the solvent used.23 This
result suggests an associative rate-controlling step.
(3)
(20) Note that the imprecision in the intercept in Figure 2, (-2 ( 6) ×
103 s, is higher than the value measured (in fact a negative intercept makes
no physical sense). Thus the b coefficient has been calculated from those
of a and kobs ) (3.37 ( 0.04) × 10-4 s-1 measured in the absence of added
AsPh3.
X in the order Cl > Br > I (Table 2), and the reactions were
(21) Under these conditions (L:Pd ) 2:1), the kinetic order in respect to
[3a] is 0.27 (Figure 3b). This order is only apparent. The nonlinear behavior
probably comes from the fact that in the absence of added AsPh3 the
concentration of AsPh3 arising from dissociation (otherwise negligible) must
be considered and is not fixed but varies with the concentration of 3a (we
have studied a similar phenomenon before in ref 14b). Moreover, this
dissociation must lead to the formation of other catalytic species (such as
iodo-bridged dimers; see ref 10), contributing to the catalysis. This effect
has never been noticed before, probably because generally an excess of
neutral ligand is used in catalytic couplings (see ref 1). However, Scott
and Stille1g already noticed that the reaction rate did not increase linearly
with the amount of palladium added for higher concentrations of Pd and
attributed this to “increased concentration of free phosphine in solution,
catalyst aggregation, or change in the catalytic species in solution”).
(22) The activation parameters given in the Table 1 cannot be assigned
to any elemental step at this point (see Discussion). For this reason we
refer to them as “apparent”.
retarded by addition of AsPh3.
Discussion
The experiments carried out on isolated perhalophenylpalla-
dium(II) reaction intermediates support that the transmetalation
is the rate-determining process of the coupling cycle (Scheme
4), as expected for a typical Stille reaction. In effect, [Pd(C6-
(23) To detect any solvent participation in the reaction, PhCl was checked,
in addition to the widely used THF; note that both solvents have the same
polarity (ET ) 98.8 kJ mol-1), hence similar solvation effects, but PhCl is
a much worse donor: (a) Catala´n, J.; Lo´pez, V.; Pe´rez, P.; Mart´ın-Villamil,
R.; Rodriguez, J.-G. Liebigs Ann. Chem. 1995, 241-252. (b) Jensen, W.
B. Chem. ReV. 1978, 78, 1-22. (c) Gutmann, V. Coord. Chem. ReV. 1976,
18, 225-255.

8982 J. Am. Chem. Soc., Vol. 120, No. 35, 1998
Casado and Espinet
Cl2F3)(CHdCH2)L2] could not be detected in transmetalation
experiments at moderate temperature (eq 3), suggesting that the
reductive elimination occurs fast once the transmetalation has
taken place. trans-[Pd(C6Cl2F3)I(AsPh3)2] (3a) was the only
organopalladium(II) intermediate detected along the catalytic
couplings of C6Cl2F3I (1) with R2SnBu3 (R2 ) CHdCH2, 2a;
C6H4-4-OCH3, 2b) (eq 1); hence the oxidative addition step and
the subsequent cis-to-trans isomerization (Scheme 4) are also
faster than the transmetalation.24 The kinetic zeroth order in
electrophile 1 agrees with this. Consequently, the kinetic results
of the catalytic runs can be properly assigned to the transmeta-
lation step.
Scheme 6a
The true complexity of the rate law (eq 2) reveals the
concurrence at the transmetalation process of a set of elemental
steps. The mechanistic interpretation establishes that the
elemental composition at the transition state is [2a + 3a -
AsPh3].19a In other words, the interaction between 2a and 3a
takes place with release of one molecule of AsPh3. We will
consider the two general pathways by which AsPh3 can be
released from 3a: before or during the interaction with the
stannane. We will refer to them as dissociative or associative
transmetalation, respectively.
In the so far accepted model, the dissociation of AsPh3
precedes the interaction with the stannane. This has been drawn
in Scheme 3 (in our case R1 ) C6Cl2F3 and L ) AsPh3). If we
consider the dissociation to be rate determining, application of
the steady-state approximation leads to eq 4.
a R1 ) C6Cl2F3, L ) AsPh3, X ) halide.
Consequently, the two possible models for a dissociatiVe
transmetalation, are inconsistent with the obserVations.
We propose the associative transmetalation pathway shown
in Scheme 6, which leads to the transformation trans-II f cis-V
via L-for-R2 substitution at the coordination plane. A nucleo-
philic attack of the Pd-coordinated halide to the organotin
compound makes the palladium center more electrophilic and
the CR atom of R2 more nucleophilic, assisting the electrophilic
attack of Pd to give the activated complex III, from which L is
released.,14b,25,26 Since associative substitutions via pentacoor-
dinate palladium occur with preservation of the stereochemistry
at the palladium,15 intermediate IV, as well as the three-
coordinate V, must necessarily have R2 in the position of the
leaving AsPh3, i.e., cis to R1. Then, V can readily eliminate
the coupling product R1-R2 without the need for further (and
comparatively slow) dissociation or isomerization steps that must
be proposed in the so far accepted mechanism (Scheme 1).
Applying the steady-state approximation to our associative
transmetalation model (Scheme 6), eq 6, with k1 ) 0.034 mol-1
k′1k′2[3a]
robs,ss ) kobs,ss[2a] )
[2a]
(4)
k′-1[AsPh3] + k′2[2a]
For a very low concentration of AsPh3 (as is the case in the
absence of added AsPh3), eq 4 is simplified to robs,ss ≈ k′1[3a],
a rate expression zeroth order with respect to [2a], contrary to
the first-order dependence observed experimentally (eq 2).
Moreover, the reaction rate becomes mathematically independent
of the organotributyltin used, contrary to the marked differences
observed between vinyltin and phenyltin compounds. Therefore,
the mechanistic assumption leading to eq 4 is to be discarded.
Alternatively, if we consider a fast dissociation,1b,10,11 eq 5
is obtained applying the preequilibrium model,19b with Kdis
k′1/k′-1
)
.
k1k2[3a]
robs,ss ) kobs,ss[2a] )
[2a]
(6)
k′2Kdis[3a]
total [2a]
Kdis + [AsPh3]
k-1[AsPh3] + k2
robs,pe ) kobs,pe[2a] )
(5)
L s-1 and k2/k-1 ) 6.9 × 10-4 mol L-1, is obtained (see
Appendix; numeric values given hold for X ) I, in THF at
322.6 K), which also agrees with the experimental rate law (eq
2).
Equation 5 agrees with the rate law (eq 2) and data treatment
gives k′2 ) 1.8 × 10-1 s-1 and Kdis ) 1.3 × 10-4 mol L-1 (at
322.6 K in THF). Since the concentration of catalyst added
was [3a]total ) 10-2 mol L-1, the value of Kdis indicates that
12% of catalyst 3a should be dissociated as C (or its corre-
sponding solvated complex) in the absence of added AsPh3. This
high dissociation should be detectable by 19F NMR, but the
expected effects were not observed.17 Hence the three-
coordinate intermediate C is not formed in significant concen-
tration, and the preequilibrium dissociative model is also
unsatisfactory.
(25) Both the ability of tin(IV) to increase its coordination number and
the formation of single-bridged compounds are well documented in the
following reviews: (a) Davies, A. G. In ComprehensiVe Organometallic
Chemistry II; Abel, E. W., Stone, F. G. A., Wilkinson, G., Eds.;
Pergamon: Oxford, U.K., 1995. Vol. 2, Chapters 6.4 and 6.5. (b) Harrison,
P. G. In ComprehensiVe Coordination Chemistry; Wilkinson, G., Gillard,
R. D., McCleverty, J. A., Eds.; Pergamon: Oxford, U.K., 1987, Vol. 3,
Chapter 26.3.4. (c) Davies, A. G.; Smith P. J. Chapter 11.4.4.2 In
ComprehensiVe Organometallic Chemistry; Wilkinson, G., Stone, F. G. A.,
Abel, E. W., Eds.; Pergamon: Oxford, U.K., 1982; Vol. 2, Chapter 11.4.4.2.
(26) For R-bridging structures observed by X-ray diffraction, see: (a)
Uso´n, R.; Fornie´s, J.; Falvello, L. R.; Toma´s, M.; Casas, J. M.; Mart´ın, A.;
Cotton, F. A. J. Am. Chem. Soc. 1994, 116, 7160-7165. (b) Uso´n, R.;
Fornie´s, J.; Toma´s, M.; Casas, J. M.; Cotton, F. A.; Falvello, L. R.; Feng,
X. J. Am. Chem. Soc. 1993, 115, 4145-4154. (c) Uso´n, R.; Fornie´s, J.;
Toma´s, M.; Casas, J. M.; Navarro, P. J. Chem. Soc., Dalton Trans. 1989,
169-172. (d) Uso´n, R.; Fornie´s, J.; Toma´s, M.; Casas, J. M. Organome-
tallics 1988, 7, 2279-2285.
(24) In fact, the coupling rate of 1 and 2a catalyzed by trans-[Pd(C6-
Cl2F3)I(PPh3)2] (10-2 mol L-1) in THF at 322.6 K is (9.9 ( 0.4) × 10-6
s-1: Casado, A. L. Ph.D. Thesis, Universidad de Valladolid, Spain, March
1998. Under the same conditions, the isomerization rate of cis-[Pd(C6-
Cl2F3)I(PPh3)2] to trans-[Pd(C6Cl2F3)I(PPh3)2] is (1.49 ( 0.09) × 10-3 s-1
,
i.e., ∼150 times faster (the isomerization limits the formation of trans-[Pd-
(C6Cl2F3)I(PPh3)2] and the oxidative addition is faster; see ref 8).

Mechanism of the Stille Reaction
J. Am. Chem. Soc., Vol. 120, No. 35, 1998 8983
Scheme 7
of the Pd complex was disregarded, and the effect of added
ligand was not studied in refs 7 and 30, where a direct X-for-
R2 substitution is implied (Scheme 7). Note that this kind of
substitution would lead to a trans arrangement of the two R
groups after transmetalation and should be little sensitive to the
addition of L, which remains coordinated. On the contrary, our
proposal explains the dependence on L observed and produces
immediately the cis arrangement needed for fast R1-R2
coupling.31
Finally, some comments about the influence of the neutral
ligands, L, can be made in light of our mechanism. It is known
from the literature that there is an inverse relationship between
ligand donicity and transmetalation rate.11 The Stille reaction
runs up to 3 orders of magnitude faster with ligands of modest
donicity (AsPh3) than with good donors (PPh3). This seemed
to support the dissociative proposal because the former should
dissociate more readily. But they will be more easily displaced
also in an associative substitution process. Moreover, the
electrophilicity of the palladium complex will be higher with L
ligands of low donicity, and this will make the activated complex
III more accessible. Furthermore, the L ligands must move to
equatorial positions in III for the substitution to occur. Hence,
if L ligands with similar net donicities are compared, it is
reasonable that the transmetalation should be favored by those
better π acceptors, as they have a stronger preference for the
equatorial sites in d8 bpt complexes.32
Figure 5. Activated complex models for transmetalation with stannanes
and silanes, as proposed in refs 7a and 30.
As k1 depends on the nature of the organotin reagent, different
coupling rates are to be expected for different organotributyltins,
even for [AsPh3] ) 0 (in such case robs,ss ) kobs,ss[2a] ) k1-
[3a][2a]). In fact, the observed rate is vinyl > phenyl.27
Although the interpretation of apparent activation parameters
is not simple in such a multistep reaction,22 the very large
negative values for the apparent ∆Sqobs (Table 1, obtained from
the composite rate constant kobs) found in all cases examined
are in agreement with the associative mechanism proposed. This
entropy argument also rules out solvent (THF) participation in
the AsPh3 replacement (at least as the main pathway), which
should produce only a small increase in order in the transition
state.19d Moreover, the activation entropy found in chloroben-
zene is also negative, although this solvent is much worse ligand
than THF.28
Changes of the halogen atom X are expected to produce little
effect on the activation entropies of the processes concerned;
hence, the variation in rates observed for the transmetalation
of trans-[Pd(C6Cl2F3)X(AsPh3)2] (Table 2) must be related to
changes in activation entalpy. These are probably associated
with a high activation enthalpy for the pentacoordination of the
Pd atom. In fact the II f IV transformation corresponds, with
little variation, to the rate-determining step of an associative
substitution in Pd complexes.15 The nucleophilic attack of the
atom of the R2 group (entering ligand) on the electrophilic Pd
complex must be facilitated as the Pd complex becomes more
electrophilic and the CR more nucleophilic (predicting a rate
variation Cl > Br > I, as observed).29
On the other hand, it seems that there is no clear correlation
between rate and the steric parameters of L,1b contrary to the
expectations for an initial L dissociation, which should be
favored for increasing steric requirement of L. In the mecha-
nism in Scheme 6, the influence of increasing the bulkiness of
L is more difficult to predict since this will destabilize the
ground and the transition states less differently than when these
are four- and three-coordinated (as in the dissociative mecha-
nism). Bulkier ligands will probably make more difficult the
access to III, but at the same time they will favor the dissociation
to IV. Thus, in a concerted process, a less clear influence can
be expected. Note, however, that coming to very bulky ligands,
a dramatic change in behavior has been reported: For the
transmetalation of the dimeric complex [PdArBrL]2 (Ar ) p-Tol;
L ) P(o-Tol)3) with trialkyltin aryls, the rate depends on the
square root of the concentration of dimer and is not affected by
the addition of L. This is consistent with a dissociative
mechanism.10 The scheme proposed in the literature for this
The mechanism in Scheme 6 is a variation of the so-called
SE (cyclic) ligand replacement.15 Thus, for the arylation of
2
[PtX2L2] complexes using aryltrimethylstannanes, four center-
activated complexes (Figure 5a) have been suggested.7a More-
over, Hatanaka and Hiyama have proposed similar activated
complexes for the palladium-catalyzed coupling of organosilicon
compounds with the aid of fluoride ion: This coupling can
course with retention or with inversion of configuration,
depending on the temperature and the solvent. For the reaction
occurring with retention of the configuration, the cyclic transition
2
state (Figure 5b) was proposed, whereas an SE (open) mecha-
nism would be operating when the reaction courses with
inversion (Figure 5c).30
2
mechanism can be accommodated to an SE (cyclic) model as
2
The SE (cyclic) mechanism implies retention of the config-
shown in Scheme 8.
uration at CR. Unfortunately, the effect on the stereochemistry
Conclusions
(27) It must be noted that in order to produce the transmetalation, R2
must become bridging ligand using its CR. This does not discount an initial
π-coordination of R2 helping the bimolecular interaction, when possible
(R2 ) vinyl, alkynyl). This kind of coordination has been proposed by Farina
et al.11 and would account for two facts observed for vinyl derivatives: (i)
A faster reaction rate and (ii) the formation of Heck coupling byproducts
in some cases; see: (a) Liao, J.-H.; Kanatzidis, M. G. J. Am. Chem. Soc.
1990, 112, 7399-7400. (b) Kikukawa, K.; Umekawa, H.; Matsuda, T. J.
Organomet. Chem. 1986, 311, C44-C46.
The studies presented here concern the conditions most
commonly used for the Stille reaction, involving the use of aryl
or vinyl halides in solvents of moderate donicity, palladium
complexes with monodentate ligands of normal steric bulk, and
ratios L:Pd > 2:1. The mechanism proposed in Scheme 4 fits
all the observations made on the reaction, either stoichiometric
or catalytic. First, it depicts the fact that the oxidative addition
(28) For THF complexes of Pd(II), see: Uso´n, R.; Fornie´s, J. AdV.
Organomet. Chem. 1988, 28, 219-297.
(29) For a dissociative ligand substitution, the converse trend should be
expected, as the dissociation of L should be the more difficult (hence Kdis
in eq 5 should be the smaller) the more electronegative the halide.
(30) Hatanaka, Y.; Hiyama, T. J. Am. Chem. Soc. 1990, 112, 7794-
7796.
(31) Pitifully, neglect of the coordination sphere of the metal is quite
common in the literature of metal-catalyzed organic transformations. We
hope that this paper will serve to draw attention on its mechanistic relevance
in some cases.
(32) Rossi, A. R.; Hoffmann, R. Inorg. Chem. 1975, 14, 365-374.

8984 J. Am. Chem. Soc., Vol. 120, No. 35, 1998
Casado and Espinet
Scheme 8
taken on a Hewlett-Packard 5980 spectrometer (70 eV, EI). 1H,
13C{1H}, 19F, and 119Sn{1H} NMR spectra were run on a Bruker
ARX-300 spectrometer equipped with a VT-100 variable-
temperature probe; chemical shifts are reported in ppm from
SiMe4 (1H, 13C), CCl3F (19F) in CDCl3, or net SnMe4 (119Sn) at
room temperature.
R2SnBu3: R2 ) CHdCH2 (2a), C6H4-4-OCH3 (2b). To a
solution below 0 °C of vinylmagnesium bromide (14.0 mmol)
in diethyl ether (60 mL) ClSnBu3 (3.48 mL, 12.8 mmol) was
slowly added. The mixture was stirred overnight while it
reached room temperature. The resulting white suspension was
hydrolyzed with aqueous NaHCO3, washed with water (2 ×
30 mL), and dried over MgSO4. The diethyl ether solution was
evaporated to give a pale yellow oil which was vacuum-distilled
yielding a colorless liquid 2a (3.45 g, 85%): 119Sn{1H} NMR
(0.2 mol L-1 in THF/ClPh) δ -48.96/-49.27 (s). 2b was
similarly prepared from (4-methoxyphenyl)magnesium iodide
in 71% yield. The rest of spectroscopic data are similar to those
reported.2
gives a cis compound, which isomerizes rapidly to the trans
isomer. From the latter, an associative ligand substitution
produces L-for-R2 exchange. This leads directly to a T-shaped
three-coordinate cis-[PdR1R2L] species from which irreversible
R1-R2 elimination must be fast, simplifying noticeably the
course of the mechanism after the transmetalation step. The
dependence of the transmetalation rate on the addition of L, on
the halide, and on the tin derivative is also understood on the
basis of this associative transmetalation. For all these reasons,
the associative model for the Stille reaction is more adequate
than the dissociative so far accepted.
However, one should be careful not to consider our proposal
as the only possible mechanism for the Stille reaction. In very
different conditions (e.g., very coordinating solvents, very bulky
ligands, chelating ligands, or lack of halide in the system) other
mechanisms might be competing with or replacing the one just
discussed. For instance, a case has been reported recently where
the stannane seems to react directly with the palladium(0)
complex.33 Also, our mechanism predicts retention of config-
uration at the carbon, which has been observed in some
reactions,34 but inversion has been reported in other cases.6 This
stresses the idea that other reaction conditions deserve mecha-
nistic studies on their own.
trans-[Pd(C6Cl2F3)Cl(AsPh3)2] (3c). To a stirred suspension
of [Pd2(C6Cl2F3)2(µ-Cl)2(tht)2] (100 mg, 0.116 mmol) in acetone
(10 mL) was added AsPh3 (150 mg, 0.488 mmol). After some
seconds, white 3c precipitated. The mixture was stirred for 1
h, and the solvent evaporated. The solid was then washed with
diethyl ether and air-dried (yield 209 mg, 94%): IR (KBr) 1436
(vs), 1400 (vs), 1078 (m), 1046 (m), 1034 (m), 777 (s), 739
1
(vs), 692 (vs), 480 (s), 460 (m), 331 (m), 314 (m); H NMR
(CDCl3) δ 7.7-7.6 (m, 2CH), 7.5-7.3 (m, 3CH); 19F NMR
(CDCl3/THF) δ -90.62/-86.82 (s, 2F2), -120.27/-116.89 (s,
F4). Anal. Calcd for C42H30As2Cl3F3Pd: C, 52.86; H, 3.17.
Found: C, 52.65; H, 3.35.
trans-[Pd(C6Cl2F3)X(AsPh3)2]: X ) I (3a), Br (3b). To a
colorless solution of 3c (110 mg, 0.115 mmol) in acetone/CH2-
Cl2 (4/4 mL) was added an excess of NaX (0.20 mmol). The
yellow mixture formed was stirred for 1 h and then evaporated
to dryness. The residue was extracted in CH2Cl2 (5 mL) and
evaporated again. The yellow solid was washed with diethyl
ether/n-hexane and vacuum-dried. 3a (93%): IR (KBr) 3073
(m), 1482 (m), 1436 (vs), 1403 (vs), 1047 (m), 775 (m), 735
Finally, a unified view seems to be merging for the trans-
metalations to palladium, whether with Sn (Stille), with silicon
fluoride-assisted (Hiyama), and possibly with other transmet-
allating agents: The reactions occurring with retention of
2
configuration most likely proceed in all cases by an SE (cyclic)
1
(vs), 691 (vs), 481 (s), 467 (s), 335 (s), 322 (s); H NMR
L-for-R replacement at the Pd center. It is plausible that an
(CDCl3) δ 7.58 (m, 2CH), 7.5-7.3 (m, 3CH); 19F NMR (CDCl3/
THF/ClPh) δ -91.81/-87.60/-87.32 (s, 2F2), -120.02/-
116.72/-115.50 (s, F4). Anal. Calcd for C42H30As2Cl2F3IPd:
C, 48.24; H, 2.89. Found: C, 48.12; H, 2.97. 3b (89%): IR
(KBr) 3055 (m), 1482 (m), 1435 (vs), 1404 (vs), 1047 (m), 778
2
SE (open) mechanism (as suggested by Hiyama in Si) can also
be operating for the Stille coupling in those cases where
inversion has been observed (using the very coordinating solvent
mixture HMPA-THF). Further studies are in progress to
explore these aspects.
1
(m), 737 (vs), 693 (vs), 483 (s), 471 (s); H NMR (CDCl3) δ
7.58 (m, 2CH), 7.5-7.3 (m, 3CH); 19F NMR (CDCl3/THF) δ
-91.01/-86.82 (s, 2F2), -120.15/-116.81 (s, F4). Anal. Calcd
for C42H30As2BrCl2F3Pd: C, 50.51; H, 3.03. Found: C, 50.45;
H, 3.16.
Experimental Section
The reactions involving organolithium or organomagnesium
reagents were carried out under N2. Commercial ClSnBu3,
AsPh3, and vinylmagnesium bromide (1 M solution in THF)
were used without further purification. The solvents were
purified using standard methods. C6Cl2F3I (1),8 trans-[PdCl2-
(AsPh3)2],35 [Pd2(C6Cl2F3)2(µ-Cl)2(tht)2],16 and (4-methoxyphe-
nyl)magnesium iodide (from 4-bromoanisole and magnesium
turnings)2 were prepared as reported in the literature.
Infrared spectra (cm-1) were recorded on a Perkin-Elmer FT-
IR 1720 X spectrometer. Combustion analyses were made on
a Perkin-Elmer 2400 CHN microanalyzer. Mass spectra were
Isolation of R2C6Cl2F3 from Catalytic Reaction Mix-
tures: R2 ) CHdCH2 (4a), C6H4-4-OCH3 (4b). After
evaporation of the solvent (THF), product 4a was vacuum-
distilled (70 °C, 3 mm) as a colorless liquid (80% isolated
yield): d 1.19. IR (NaCl) 2928 (m), 1610 (s), 1463 (vs), 1451
(vs), 1411 (vs), 1219 (s), 1081 (s), 991 (s), 936 (s), 860 (s),
1
3,trans
801 (s); H NMR (CDCl3) δ 6.61 (dd,
J
) 18.0 Hz,
HH
3,cis
J
HH
) 11.9 Hz, CHgem), 6.06 (d, 3,transJHH ) 18.0 Hz, CHcis),
5.69 (d, 3,cisJHH ) 11.9 Hz, CHtrans); 13C{1H} NMR (CDCl3) δ
1
4
4
154.94 (ddd, JCF ) 253.9 Hz, JCF ) 9.2 Hz, JCF ) 4.2 Hz,
(33) Shirakawa, E.; Yoshida, H.; Hiyama, T. Tetrahedron Lett. 1997,
CF2), 153.77 (dt, 1JCF ) 251.9 Hz, 4JCF ) 5.5 Hz, CF4), 123.22
38, 5177-5180.
(34) Ye, J.; Bhatt, R. K.; Falck, J. R. J. Am. Chem. Soc. 1994, 116,
4
6
3
(td, JCF ) 7.9 Hz, JCF ) 2.5 Hz, CH2), 121.60 (td, JCF
)
1-5.
17.3 Hz, 5JCF ) 2.3 Hz, CH), 112.62 (td, 2JCF ) 16.5 Hz, 4JCF
(35) Norbury, A. H.; Sinha, A. I. P. J. Inorg. Nucl. Chem. 1973, 35,
1211-1218.
2
2
) 4.6 Hz, C-vinyl), 107.29 (ddd, JCF ) 24.9 Hz, JCF ) 21.0

Mechanism of the Stille Reaction
J. Am. Chem. Soc., Vol. 120, No. 35, 1998 8985
4
Hz, JCF ) 3.9 Hz, C-CF); 19F NMR (CDCl3/THF/ClPh) δ
Reactions of Organopalladium(II) Complexes and Orga-
notributyltins. Samples containing 0.01 mol L-1 in palladium
complex and 0.2 mol L-1 in the corresponding organotributyltin,
prepared as above-described, were allowed to react at 20 °C
(thermostated bath). The reaction products were analyzed by
19F NMR at the reported intervals of time.
-115.41/-112.21/-111.50 (d, 4JFF ) 2.4 Hz, 2F2), -113.29/-
110.62/-109.57 (t, JFF ) 2.4 Hz, F4); MS m/z (%) 226 (50)
4
[M+], 191 (41), 156 (100), 105 (25). Anal. Calcd for C8H3-
Cl2F3: C, 42.33; H, 1.33. Found: C, 42.27; H, 1.32. 4b: After
evaporation of the solvent (THF), the residue was treated with
diethyl ether and a saturated aqueous KF solution. After
vigorous stirring, FSnBu3 separated as a white solid. The diethyl
ether phase was separated, dried over MgSO4, and evaporated
to dryness. The residue was chromatographed (silica gel/
hexane) giving a white solid, which was recrystallized from
pentane at -28 °C (white needles, 95%): mp 98-99 °C; IR
(KBr) 2941 (m), 2839 (m), 1613 (s), 1522 (s), 1445 (vs), 1408
(vs), 1255 (vs), 1186 (s), 1053 (vs), 1034 (s), 795 (vs), 565
Acknowledgment. Dedicated to Prof. Peter M. Maitlis on
the occasion of his 65th birthday. We thank Prof. J. Barluenga
and Prof. A. M. Echavarren for helpful discussions. Financial
support by the Direccio´n General de Investigacio´n Cient´ıfica y
Te´cnica (Project PB96-0363) and the Junta de Castilla y Leo´n
(Project VA 40-96), and a fellowship to A.L.C. from the
Ministerio de Educacio´n y Ciencia are very gratefully acknowl-
edged.
1
(m), 526 (m); H NMR (CDCl3) δ 7.35 (m, 2CH), 7.14 (m,
2CH), 3.87 (s, CH3); 13C{1H} NMR (CDCl3) δ 160.09 (s,
C-OMe), 154.34 (ddd, 1JCF ) 249.7 Hz, 4JCF ) 8.3 Hz, 4JCF
)
Supporting Information Available: Values of the first-
order rate constant kobs for the coupling of C6Cl2F3I (1, 0.2 mol
L-1) with R2SnBu3 (2a,b, 0.2 mol L-1) catalyzed by trans-
[Pd(C6Cl2F3)I(AsPh3)2] (3a) under different experimental condi-
tions (1 page, print/PDF). See any current masthead page for
ordering information and Web access instructions.
4.4 Hz, CF2), 153.80 (dt, 1JCF ) 250.6 Hz, 4JCF ) 5.3 Hz, CF4),
131.30 (s, CH), 118.87 (s, C-C6Cl2F3), 114.07 (s, CH), 116.09
(td, 2JCF ) 20.1 Hz, 4JCF ) 4.4 Hz, C-Ph), 107.34 (ddd, 2JCF
)
25.6 Hz, 2JCF ) 21.0 Hz, 4JCF ) 3.3 Hz, CCF), 55.22 (s, CH3);
19F NMR (CDCl3/THF) δ -115.69/-112.18 (dt, 4JFF ) 2.4 Hz,
5JFH ) 1.3 Hz, 2F2), -113.64/-111.39 (t, JFF ) 2.4 Hz, F4);
4
MS m/z (%) 306 (78) [M+], 263 (69), 193 (100). Anal. Calcd
for C13H7Cl2F3O: C, 50.85; H, 2.30. Found: C, 50.84; H, 2.36.
Kinetics on Palladium-Catalyzed Couplings of C6Cl2F3I
(1) with Organotributyltins (2a,b). NMR tubes (5 mm) were
charged with 1 (39.2 ( 0.1 mg, 120.0 ( 0.3 mmol) and suitable
amounts of palladium catalyst 3a, AsPh3, and organotributyltin
2a,b. The samples were dissolved under N2 at room temperature
(293 K) in THF (or PhCl) to a fixed volume of 600 ( 5 µL,
charged with an acetone-d6 capillary for NMR lock, and placed
into a thermostated probe ((0.2 K; the temperature was
measured by an ethylene glycol standard method). Concentra-
tion-time data were then acquired from 19F NMR signal areas
of 1 and the products (4a,b), and fitted to equation ln([1]0 -
[1]) ) kobst to get first-order constants kobs (standard deviations
are also given). Alternatively, when no good first-order rates
were achieved (couplings with 2a in absence of added AsPh3),
first data points (up to 10% of conversion) were fitted to the
following second-degree Taylor equation [1] ) a0 + a1t + a2t2,
where a1 gives the initial reaction rate in mol L-1 s-1 (a1 ) r0
) -d[1]/dt ) -d[2a]/dt). Since the reaction rate r is first order
in 2a and zeroth order in 1 (eq 2), the first-order constants can
be estimated in these cases as kobs ) r0/[2a]0.
Appendix: Derivation of the Kinetic Eq 6. The steady-
state concentration of intermediate IV (Scheme 6, X ) I), and
the reaction rate, are given in eqs 7-9.
d[IV]/dt ) k1[2a] [3a] - k-1[IV][AsPh3] - k2[IV] ) 0 (7)
k1[2a][3a]
[IV] )
(8)
k-1[AsPh3] + k2
robs,ss ) k2 [IV]
(9)
(6)
k1k2[3a]
k-1[AsPh3] + k2
robs,ss ) kobs,ss[2a] )
[2a]
JA9742388