Palladium pincer complex catalyzed stannyl and silyl transfer to propargylic substrates: Synthetic scope and mechanism

Article abstract of DOI:10.1021/ja043951b

Pincer complex catalyzed substitution of various propargylic substrates could be achieved using tin- and silicon-based dimetallic reagents to obtain propargyl- and allenylstannanes and silanes. These reactions involving chloride, mesylate, and epoxide substrates could be carried out under mild conditions, and therefore many functionalities (such as COOEt, OR, OH, NR, and NAc) are tolerated. It was shown that pincer catalysts with electron-supplying ligands, such as NCN, SCS, and SeCSe complexes, display the highest catalytic activity. The catalytic substitution of secondary propargyl chlorides and primary propargyl chlorides with electron-withdrawing substituents proceeds with high regioselectivity providing the allenyl product. Opening of the propargyl epoxides takes place with an excellent stereo- and regioselectivity to give stereodefined allenylstannanes. Silylstannanes as dimetallic reagents undergo an exclusive silyl transfer to the propargylic substrate affording allenylsilanes with high regioselectivity. According to our mechanistic studies, the key intermediate of the reaction is an organostannane (or silane)-coordinated pincer complex, which is formed from the dimetallic reagent and the corresponding pincer complex catalyst. DFT modeling studies have shown that the trimethylstannyl functionality is transferred to the propargylic substrate in a single reaction step with high allenyl selectivity. Inspection of the TS structures reveals that the trimethylstannyl group transfer is initiated by the attack of the palladium-tin σ-bond electrons on the propargylic substrate. This is a novel mechanism in palladium chemistry, which is based on the unique topology of the pincer complex catalysts.

Full text of DOI:10.1021/ja043951b

Published on Web 01/25/2005
Palladium Pincer Complex Catalyzed Stannyl and Silyl
Transfer to Propargylic Substrates: Synthetic Scope and
Mechanism
Johan Kjellgren, Henrik Sunde´n, and Ka´lma´n J. Szabo´*
Contribution from the Department of Organic Chemistry, Arrhenius Laboratory,
Stockholm UniVersity, SE-106 91, Stockholm, Sweden
Received October 5, 2004; E-mail: kalman@organ.su.se
Abstract: Pincer complex catalyzed substitution of various propargylic substrates could be achieved using
tin- and silicon-based dimetallic reagents to obtain propargyl- and allenylstannanes and silanes. These
reactions involving chloride, mesylate, and epoxide substrates could be carried out under mild conditions,
and therefore many functionalities (such as COOEt, OR, OH, NR, and NAc) are tolerated. It was shown
that pincer catalysts with electron-supplying ligands, such as NCN, SCS, and SeCSe complexes, display
the highest catalytic activity. The catalytic substitution of secondary propargyl chlorides and primary propargyl
chlorides with electron-withdrawing substituents proceeds with high regioselectivity providing the allenyl
product. Opening of the propargyl epoxides takes place with an excellent stereo- and regioselectivity to
give stereodefined allenylstannanes. Silylstannanes as dimetallic reagents undergo an exclusive silyl transfer
to the propargylic substrate affording allenylsilanes with high regioselectivity. According to our mechanistic
studies, the key intermediate of the reaction is an organostannane (or silane)-coordinated pincer complex,
which is formed from the dimetallic reagent and the corresponding pincer complex catalyst. DFT modeling
studies have shown that the trimethylstannyl functionality is transferred to the propargylic substrate in a
single reaction step with high allenyl selectivity. Inspection of the TS structures reveals that the
trimethylstannyl group transfer is initiated by the attack of the palladium-tin σ-bond electrons on the
propargylic substrate. This is a novel mechanism in palladium chemistry, which is based on the unique
topology of the pincer complex catalysts.
I. Introduction
palladium is largely restricted to +2. Because of feature 2, the
catalytic activity of pincer complexes and commonly used
palladium salts is particularly different,13-16 as usual palladium
sources have at least two coordination sites available for the
catalytic processes.
Palladium catalysis offers an attractive approach for synthesis
of unsaturated organometallic compounds from dimetallic
reagents because of the high selectivity and functional group
tolerance of the process.1-6 Application of pincer complex
catalysts,7-12 such as 1a-e (eq 1), instead of commonly used
palladium sources opens new synthetic routes for selective
synthesis of densely functionalized organometallic reagents due
to three important features: (1) in these pincer complexes, there
is a strong rigid bonding between the central atom and the
terdentate ligand; (2) there is only a single coordination site,
trans to the electron-rich palladium-carbon bond available on
palladium for catalytic processes; and (3) the oxidation state of
(1) Beletskaya, I.; Moberg, C. Chem. ReV. 1999, 99, 3435.
(2) Marshall, J. A. Chem. ReV. 2000, 100, 3163.
(3) Horn, K. A. Chem. ReV. 1995, 95, 1317.
(4) Suginome, M.; Ito, Y. Chem. ReV. 2000, 100, 3221.
(5) Azizian, H.; Eaborn, C.; Pidcock, A. J. Organomet. Chem. 1981, 215, 49.
(6) Azarian, D.; Dua, S. S.; Eaborn, C.; Walton, D. R. M. J. Organomet. Chem.
1976, 117, C55.
(7) Albrecht, M.; Koten, G. v. Angew. Chem., Int. Ed. 2001, 3750.
(8) Slagt, M. Q.; Rodr´ıguez, G.; Grutters, M. M. P.; Gebbink, R. J. M. K.;
Klopper, W.; Jenneskens, L. W.; Lutz, M.; Spek, A. L.; Koten, G. v.
Chem.sEur. J. 2004, 10, 1331.
The palladium-catalyzed reactions of hexaalkylditin with
alkynes represent an interesting example.1,17-20 Application of
(13) Solin, N.; Kjellgren, J.; Szabo´, K. J. Angew. Chem., Int. Ed. 2003, 42,
3656.
(14) Solin, N.; Kjellgren, J.; Szabo´, K. J. J. Am. Chem. Soc. 2004, 126, 7026.
(15) Kjellgren, J.; Sunde´n, H.; Szabo´, K. J. J. Am. Chem. Soc. 2004, 126, 474.
(16) Wallner, O. A.; Szabo´, K. J. Org. Lett. 2004, 6, 1829.
(17) Mitchell, T. N.; Amamria, A.; Killing, H.; Rutschow, D. J. Organomet.
Chem. 1986, 304, 257.
(9) Rodr´ıguez, G.; Albrecht, M.; Schoenmaker, J.; Ford, A.; Lutz, M.; Spek,
A. L.; Koten, G. v. J. Am. Chem. Soc. 2002, 124, 5127.
(10) Boom, M. E. v. d.; Milstein, D. Chem. ReV. 2003, 103, 1759.
(11) Ohff, M.; Ohff, A.; Boom, M. E. v. d.; Milstein, D. J. Am. Chem. Soc.
1997, 119, 11687.
(12) Singleton, J. T. Tetrahedron 2003, 59, 1837.
9
10.1021/ja043951b CCC: $30.25 © 2005 American Chemical Society
J. AM. CHEM. SOC. 2005, 127, 1787-1796
1787

A R T I C L E S
Kjellgren et al.
usual palladium(0) catalysts leads to oxidative addition of
palladium to the tin-tin bond as long as propargylic leaving
groups (such as halogenides, mesylate, epoxide, etc.) are not
present in the substrate. This oxidative addition generates a bis-
stannyl palladium(II) species,19 which then adds to the triple
bond providing a disubstituted olefin product. As only a single
coordination site is available on palladium in a pincer complex
catalyst, formation of a bis-stannyl reagent is unlikely, and
therefore a completely different reactivity is expected. Indeed,
we have found that using hexamethylditin (2a) with propargyl
chloride or mesylate substrates (e.g., 3a-f) and catalytic
amounts (2.5 mol %) of pincer complex catalyst leads to
displacement of the propargylic leaving group instead of addition
of the trimethylstannane groups to the triple bond.15 Thus, this
reaction leads to formation of propargyl or allenylstannanes,
depending on the steric and electronic effects of the substituents
attached to the triple bond.
yield. The neutral reaction conditions and the redox stability of
the catalyst allow a wide variety of substrate functionalities,
including COOEt, OR, OAc, NR, NAc, and even unprotected
OH. The regioselectivity of the catalytic displacement reaction
for primary propargyl chlorides is strongly dependent on the
electronic effects of the propargylic substituents. For bulky
electron-supplying substituents, such as phenyl (3a,b) and alkyl
(3c) groups, the main product of the reaction is propargylstan-
nane (entries 1-8). Propargyl chloride (3d) itself and substrates
with electron-withdrawing substituents (e.g., 3e, 3f, 3i, and 3j)
give exclusively allenyl products (entries 9-12, 16, and 17).
Interestingly, even subtle electronic effects have a strong
influence on the regioselectivity of the stannylation process. For
example, the hydroxymethylene group in 3f has a weak electron-
withdrawing effect, which is sufficient to direct the reaction
toward exclusive formation of the corresponding allenyl product
5e (entry 12). However, benzyl protection (3g) of the hydroxy
group slightly modifies the electronic properties of the substrate
leading to an appearance of the corresponding propargyl product
(4c) as a minor component in the reaction mixture (entry 13).
Furthermore, insertion of a methylene unit renders the OBn
functionality to the γ-position to the triple bond (3h), which
leads to predominant formation of the propargyl product 4d
(entry 15). A similar trend is observed for an aminomethylene-
type of substituents (entries 18-20). As one goes from electron-
supplying (3k) to electron-withdrawing substituents (3l and 3m)
on nitrogen, the allenyl selectivity increases.
Recently, we have communicated our results involving
primary propargyl chloride and mesylate substrates in these
reactions.15 In this paper, we give a full account of our results
on pincer complex catalyzed substitution of propargylic sub-
strates with dimetallic reagents. In addition, we present our new
studies on the following synthetic (i-iii) and mechanistic (iv)
aspects of this reaction:
(i) Possibilities to extend the synthetic scope to secondary
propargylic substrates, including substitution of propargyl
epoxides.
(ii) Application of unsymmetrical dimetal reagents, such as
silylstannanes 2b-d, for chemoselective silyl transfer reactions.
(iii) Possibilities to improve the catalytic activity of the
applied pincer catalyst by varying the heteroatom in the ligand
(cf. complexes 1d and 1e).
(iv) Modeling the transfer of the trimethylstannyl group from
the pincer complex to the propargylic substrate using density
functional theory (DFT).
Secondary Propargylic Substrates. Our new studies on
secondary propargyl chlorides (3n-r) show that NCN catalyst
1a smoothly transforms these substrates to the corresponding
allenylstannane products (entries 21-26). The reaction proceeds
with an excellent regioselectivity for various propargylic sub-
stituents involving phenyl (3n), 2-naphthyl (3o), benzyl (3p),
and N-Boc-aminomethylene (3q) groups to provide exclusively
the corresponding allenyl product with good to excellent yield.
Phenyl substitution of 3a at the propargylic position (3r) leads
to a complete reversal of the regioselectivity, as the catalytic
stannylation of substrate 3r gives exclusively the allenyl product
5q (cf. entries 1 and 26).
We have found that propargyl epoxides (3s-u) can also be
employed in place of propargyl chlorides. The epoxide opening
takes place with an excellent regioselectivity providing exclu-
sively the corresponding allenyl product (entries 27-29). When
the reaction is conducted at 0 °C, the opening of cyclic epoxides
3t and 3u provides single diastereomers 5s and 5t, respectively
(entries 28 and 29). Thus, the pincer complex catalyzed epoxide
opening reaction proceeds with an excellent regio- and stereo-
chemistry and, at least for 3t, an excellent yield. Increasing the
reaction temperature to r.t. leads to formation of small amounts
the other diastereomeric form (up to 25%).
2. Catalytic Trimethyltin Transfer to Propargylic
Substrates
Primary Propargylic Substrates. As we communicated
before,15 NCN complex 1a efficiently catalyzes the transfer of
the trimethyltin group from hexamethylditin (2a) to primary
propargyl chloride and mesylate substrates. The reactions are
conducted under mild and neutral conditions at room temper-
ature. Application of mild reaction conditions is essential to
achieve high yields in these reactions since the functionalized
propargyl and allenylstannane products have a limited thermal
stability. Increase of the reaction temperature over 50 °C usually
leads to substantial decomposition of these products giving poor
Attempts to Increase the Activity of the Pincer Complex
Catalyst. The high electron density on palladium ensured by
the σ-donor amino and phenyl ligands is essential to achieve a
high catalytic activity by NCN complex 1a. A slight decrease
of the electron-donor properties of the phenyl group by nitro
substitution (1b)8 leads to lowering of the catalytic activity of
the pincer complex (cf. entries 1 and 2). Commonly used
catalysts, such as Pd(PPh3)4 (6) and Li2[PdCl4], do not show
any catalytic activity under the applied reaction conditions.
Exchange of the amino groups in 1a with less-efficient σ-donor
(18) Mabon, R.; Richecoeur, A. M. E.; Sweeney, J. B. J. Org. Chem. 1999, 64,
328.
(19) Tsuji, Y.; Nishiyama, K.; Hori, S.-i.; Ebihara, M.; Kawamura, T. Orga-
nometallics 1998, 17, 507.
(20) Manusco, J.; Lautens, M. Org. Lett. 2003, 5, 1653.
9
1788 J. AM. CHEM. SOC. VOL. 127, NO. 6, 2005

Palladium Pincer Complex Catalyzed Stannyl and Silyl Transfer
A R T I C L E S
phosphines (1c)21 also leads to a complete loss of the catalytic
activity (entry 3). Therefore, we considered replacing the amino
ligands of 1a with strongly electron-supplying ligands, such as
sulfur (1d)22 and selenium-based ligands (1e).23 The SCS
complex 1d displayed a high catalytic activity; however, the
catalytic transformations using 1d required longer reaction times
and higher reaction temperatures than the corresponding pro-
cesses with 1a (cf. entries 1 and 4, as well as entries 7 and 8).
On the other hand, SeCSe complex 1e, reported very recently
by Yao and co-workers,23 displayed a catalytic activity higher
than that of 1a. Comparing the corresponding entries in Table
1 (entries 1 and 5, 10 and 11, 13 and 14, as well as 24 and 25)
reveals that the stannylation reactions with SeCSe complex 1e
required typically 50% shorter reaction times than with 1a. On
the other hand, we obtained slightly lower regioselectivities
(entries 5 and 14) and yields (entries 5 and 8) with 1e than
with 1a as catalyst. These results indicate that in the presented
trimethyltin transfer processes, NCN complex 1a is a slightly
more selective catalyst than SeCSe complex 1e; on the other
hand, slow catalytic transformations can be efficiently acceler-
ated by employment of 1e as catalyst.
products were usually higher with reagent 2b (permitting a
reaction temperature of 40 °C) than with 2c (employed at 60
°C). A fast addition of 2b usually leads to a partial decomposi-
tion of this dimetallic reagent, which caused lowering of the
yields. Reagent 2d could also be employed in the silylation
reactions; however, this reagent undergoes extensive decom-
position in the reaction mixture, even when it was added slowly
to the reaction mixture. It was also found that addition of 4 Å
molecular sieves leads to improved yields in the catalytic
silylation reactions.
Since the substitution reactions with 2b are still conducted
under relatively mild and neutral conditions, many functional-
ities, including COOEt, OBn, CH2Ph, and NHBoc, are tolerated.
The regioselectivity of the reaction is excellent, as we obtained
exclusively allenylsilane products (8). The only exception is
the reaction of 3c with 2b affording predominantly the allenyl
product 8c together with 20% of propargylsilane 7 (entry 32).
The high allenyl preference of the reaction with silylstannanes
2b,c is in sharp contrast with the regioselectivity of the
stannylation process with hexamethylditin (2a). For example,
propargyl chloride 3a reacts with 2a in the presence of pincer
complex catalyst 1a, affording mainly the propargyl product
4a and only traces of allenylstannane 5a (entry 1). Conversely,
the same catalytic reaction with silylstannanes 2b and 2c (entries
30 and 31) yields exclusively allenylsilanes 8a and 8b,
respectively. As mentioned, the reaction with alkyl-substituted
substrate 3c gave the propargylic isomer as a minor product
(entry 32). However, 3h bearing a benzyloxy functionality at
the γ-position gives again exclusively allenyl product 8f (entry
35). This regioselectivity (entries 32 and 35) is also the opposite
to the selectivity of the stannylation reactions (entries 7 and
15), which provide the corresponding propargylic isomers (4b
and 4d) as the major products.
3. Application of Unsymmetrical Dimetal Reagents
2b-d
Similar to 2a, unsymmetrical dimetal reagents, such as
silylstannanes (2b-d), are known to undergo addition to triple
bonds in the presence of usual palladium catalysts (such as 6)
providing difunctionalized olefins.1,4,24-26 In the above pincer
complex catalyzed applications, symmetrical distannane 2a was
employed as reagent for the palladium pincer catalyzed trialkyl-
metal transfer process. However, when unsymmetrical reagents
2b-d are employed, either the silyl or the stannyl group can
be transferred to the propargyl substrates.
The silylation reactions proceed smoothly in the presence of
electron-withdrawing substituents at the R- or â-position to the
triple bond (entries 33 and 34). Product 8e proved to be
particularly sensitive to thermal decomposition. When 3g was
treated with 2c in place of 2b at 60 °C, only traces of the
corresponding allenylsilane product (8e) could be isolated from
the reaction mixture. Secondary propargyl chloride derivatives
3p and 3q were also converted smoothly, affording the
corresponding allenylsilanes 8g,h (entries 36 and 37).
We have found that the pincer complex catalyzed substitution
of propargyl chlorides can also be performed by employment
of silylstannanes 2b and 2c (eq 3). A particularly important
feature of this reaction is that exclusively the silyl functionality
is transferred to the propargylic substrates, affording allenylsi-
lanes in good yield (entries 30-37). These catalytic reactions
were conducted at 40 °C with slow addition (over 5 h) of
trimethylstannyl reagent 2b, while a higher reaction temperature
(60 °C) was required when tributylstannyl analogue 2c was
employed (entry 31). Since the allenylsilane products have a
relatively poor thermal stability, the isolated yields of these
4. Mechanistic Considerations
The mechanistic aspects of the catalytic reactions were studied
by performing stoichiometric reactions for the conceivable
elementary steps of the catalytic cycle. These reactions were
conducted in THF-d8, and they were monitored by various NMR
spectroscopy methods. It is well-known that palladium(0)
complexes readily undergo oxidative addition to propargyl
chlorides.27-29 However, we found that NCN complex 1a or
its acetate analogue 1f (in which the counterion dissociates easier
than in 1a) did not react with propargyl chloride 3d at r.t. to 70
°C up to 24 h reaction time (eq 4).
(21) Rimml, H.; Venanzi, L. M. J. Organomet. Chem. 1983, 259, C6.
(22) Dupont, J.; Beydoun, N.; Pfeffer, M. J. Chem. Soc., Dalton Trans. 1989,
1715.
(23) Yao, Q.; Kinney, E. P.; Zheng, C. Org. Lett. 2004, 6, 2997.
(24) Mitchell, T. N.; Killing, H.; Dicke, R.; Wickenkamp, R. J. Chem. Soc.,
Chem. Commun. 1985, 354.
(25) Mitchell, T. N.; Wickenkamp, R.; Amamria, A.; Dicke, R.; Schneider, U.
J. Org. Chem. 1987, 52, 4868.
(26) Chenard, B. L.; Zyl, C. M. V. J. Org. Chem. 1986, 51, 3561.
9
J. AM. CHEM. SOC. VOL. 127, NO. 6, 2005 1789

A R T I C L E S
Kjellgren et al.
Table 1. Pincer Complex Catalyzed Silyl and Stannyl Transfer to Propargylic Substratesa
9
1790 J. AM. CHEM. SOC. VOL. 127, NO. 6, 2005

Palladium Pincer Complex Catalyzed Stannyl and Silyl Transfer
A R T I C L E S
Table 1 (Continued)
a In the stannylation reactions, 2.5 mol % of pincer complex catalyst was applied (General Procedure A), while in the silylation processes, 5 mol %
catalyst was applied and the silylstannane reagent (2b) was added in 5 h (General Procedure B). b Pincer complex catalyst. c Reaction conditions; reaction
temperature [°C]/reaction time [h]; r.t. ) room temperature. d Propargyl-to-allenyl ratio; a.p. ) allenyl product only (indicating that the propargylic isomer
was not detected in the crude and in the isolated product by 1H NMR spectroscopy). e Isolated yield. f Product was not isolated because of its volatility. The
yield was determined by NMR spectroscopy.
Stoichiometric Reactions with (SnMe3)2. In contrast to
the above stoichiometric reaction (eq 4), complex 1f reacted
readily with 2a. The progress of this reaction was clearly
ppm) but significant upfield shift of the 1H protons of the pincer
complex. According to the time-averaged H NMR spectrum,
1
the newly formed species has a C2V symmetry, indicating that
the symmetrical pincer structure is maintained in the reaction
of 1f with 2a. The above assumption for formation of 1g is
confirmed by the 119Sn NMR spectrum of the reaction mixture.
In the 119Sn NMR spectrum, we observed a decrease of the
resonance peak of 2a (-109 ppm) and a simultaneous appear-
ance of a new peak at -0.2 ppm. This latter peak we assigned
to the 119Sn NMR resonance of 1g. This assignment is also in
line with the 119Sn NMR data published for similar palladium-
1
indicated by systematic changes of the H and 119Sn NMR
1
spectrum of the reaction mixture. In the H NMR spectrum of
this stoichiometric reaction, the methyl peak of 2a (0.2 ppm)
gradually decreased, and at the same time, two new singlets
appeared at 0.5 and 0.05 ppm (eq 5). The peak at 0.5 ppm
belongs to the methyl resonance of 9, while the appearance of
the other singlet is indicative for the formation of a new
trimethylstannyl species, 1g. We could also observe a small (0.1
9
J. AM. CHEM. SOC. VOL. 127, NO. 6, 2005 1791

A R T I C L E S
Kjellgren et al.
trimethylstannyl complexes, such as Pd(PMe3)2(SnMe3)2 [δ(119Sn)
) -28 ppm]19 and Pddppe(CONiPr)(SnMe3) [δ(119Sn) ) 45.4
ppm].30 On the basis of the above results, we consider complex
1g as the product of the introducing step of the catalytic process.
The Catalytic Cycle. On the basis of the above stoichiometric
reactions, we consider the transfer of the corresponding metal
functionality from the dimetal reagent to the pincer complex
(eq 7) as the introducing step of the catalytic cycle. This
transmetalation process involves cleavage of the corresponding
Sn-M bond. In the case of symmetrical dimetal reagent 2a,
this reaction leads to transfer of the trimethylstannyl group to
palladium. However, for unsymmetrical reagent 2b, either the
silyl or the stannyl functionality can be transferred to palladium.
Nevertheless, according to the experimental results (entries 30-
37) and the results of the stoichiometric reactions (eq 6), it can
be concluded that preferentially the silyl functionality is
transferred to palladium. The predominant formation of 1h
instead of 1g probably arises from the higher electronegativity
of silicon in 2b. The heterolytic cleavage of the Si-Sn bond
leads to accumulation of the negative charge on the silicon atom
and charge depletion on the more electropositive tin atom. Thus,
in the transmetalation step, the halogenide counterion on
palladium will be replaced by a negatively charged silyl group
instead of a positively charged stannyl functionality.
Stoichiometric Reactions with PhMe2Si-SnMe3. We have
also briefly studied the stoichiometric reaction of complex 1a
with silylstannane 2b (eq 6). Silylstannane 2b in THF-d8 was
added slowly (over 5 h) to pincer complex 1a dissolved in THF-
d8, then the resulting mixture was analyzed by 29Si NMR. In
this stoichiometric reaction, 2b [δ (29Si) ) -13.4 ppm] was
completely consumed, while two new peaks at -21.9 and -5.5
ppm appeared in the 29Si NMR spectrum of the reaction mixture.
The high field shift at -21.9 ppm arose from 2e (lit.31 ) -21.7
ppm) formed by disproportionation of 2b (eq 6). The other shift
at -5.5 ppm could be easily distinguished from the resonance
frequency of the possible hydrolysis product (PhSiMe2)2O
resonating at -1.2 ppm. The shift value at -5.5 ppm is close
to the 29Si NMR shift (-1.1 ppm) reported32 for the (dcpe)2-
PdHSiMe2Ph complex also incorporating a palladium(II)-
SiMe2Ph bond. Accordingly, we assigned this shift value to
pincer complex 1h (eq 6), which is the silyl analogue of 1g
formed in the corresponding reaction with 2a (eq 5).
The final step of the catalytic process is the transfer of the
organometallic functionality (MMe2R) from palladium to the
propargylic substrate. Considering the fact that there is no
accessible coordination site on palladium in 1g (or 1h), this
process is expected to be the most unique reaction step in the
above pincer complex catalyzed metalation process. Thus, to
explore the mechanistic details of this step, we carried out DFT
modeling studies. In these studies, we investigated the mecha-
nistic aspects of the transfer of trimethylstannyl group from
complex 1g to propargyl chloride 3d.
DFT Modeling of the Transfer of the Trimethylstannyl
Group. The DFT studies were focused on the exploration of
the potential energy surface (PES) of the trimethylstannyl
transfer process (Figures 1 and 2). Study of the factors
determining the regiochemistry of the reaction was of particular
interest. Therefore, we studied various possible pathways leading
to formation of the allenyl (5c) and propargyl (11) stannanes
from 1g and 3d.
All geometries were fully optimized employing a Becke-
type33 three-parameter functional B3PW91 using a double-ú-
(DZ)+P basis constructed from the LANL2DZ basis34-36 by
adding one set of polarization functions to the heavy atoms
(exponents: C, 0.63; N, 0.864; Cl, 0.514; Sn, 0.183) and one
set of diffuse functions on palladium (exponent: 0.0628).
It was also observed that stoichiometric addition of 2b to 1a
in one portion (instead of using slow addition) leads to an
extensive disproportionation of 2b yielding 2e and 2a. As
mentioned above, fast addition of 2b to the reaction mixture in
the catalytic reactions (eq 3) usually leads to lowering of the
yield, which can explained by disproportionation or other
degradation processes of the unreacted 2b.
Unfortunately, our efforts to isolate silicon- and tin-organic
pincer complexes 1g and 1h were fruitless because of the poor
stability of these species. This poor stability can probably be
explained by the presence of a highly reactive tin/silicon-
palladium σ-bond (vide infra) in these complexes.
(27) Elsevier, C. J.; Kleijn, H.; Boersma, J.; Vermeer, P. Organometallics 1986,
5, 716.
(28) Tsuji, J.; Mandai, T. Angew. Chem., Int. Ed. Engl. 1995, 34, 2589.
(29) Ma, S.; Zhang, A. J. Org. Chem. 2002, 67, 2287.
(30) Hua, R.; Onozawa, S.-y.; Tanaka, M. Organometallics 2000, 19, 3269.
(31) Fu¨rstner, A.; Weidmann, H. J. Organomet. Chem. 1988, 354, 15.
(32) Boyle, R. C.; Mague, J. T.; Fink, M. J. J. Am. Chem. Soc. 2003, 125, 3228.
(33) Becke, A. D. J. Chem. Phys. 1993, 98, 5648.
(34) Dunning, T. H.; Hay, P. J. Modern Theoretical Chemistry; Plenum: New
York, 1977; Vol. 3.
(35) Hay, P. J.; Wadt, W. R. J. Chem. Phys. 1985, 82, 270.
(36) Hay, P. J.; Wadt, W. R. J. Chem. Phys. 1985, 82, 299.
9
1792 J. AM. CHEM. SOC. VOL. 127, NO. 6, 2005

Palladium Pincer Complex Catalyzed Stannyl and Silyl Transfer
A R T I C L E S
Figure 1. Reaction profiles for the trimethylstannyl transfer reactions from palladium (1g) to propargyl chloride (3d) giving allenyl (5c) or propargyl
product (11). The TS structures (10a-d) are given in Figure 2. All energies are given in kcal mol-1
.
According to the vibrational analysis (performed at the opti-
mization level), 10a-d were characterized as transition-state
structures, while the rest of the structures were characterized
as minima on the PES. All calculations were carried out by
employing the Gaussian 03 package.37
(2.961 Å) and the tin (2.773 Å) atoms accompanied by the
cleavage of the carbon-chlorine bond (2.502 Å). We have also
localized two other TS structures, 10c and 10d, in which the
palladium atom is not directly involved in the trimethyltin
transfer process. In these processes, however, the activation
barrier is unrealistically high (43.5 and 52.6 kcal mol-1), and
therefore it is unlikely that the stannylation reaction would
proceed through these TS structures.
The calculated structure of 1g was characterized by a
relatively long palladium-tin bond (2.749 Å) and with a tight
coordination by the nitrogen and carbon atoms of the pincer
ligand. We considered two distinct pathways for the trimethyltin
transfer reaction. (a) The first pathway involves interaction
between the palladium atom and propargylic substrate. (b) The
other one is performed without involvement of the palladium
atom. An interesting feature of both pathways is that the
formation of allenyl (5c) or propargyl stannane (11) from 1g
and 3d takes place in a single reaction step. This means that
the trimethylstannyl transfer from 1g to 3d involves a simul-
taneous formation of the carbon-tin bond and cleavage of the
carbon-chlorine bond.
The lowest energy path on the PES leads to formation of
allenylstannane via TS 10a requiring a relatively low activation
energy of 20.5 kcal mol-1 (Figure 1). In TS 10a, the distances
of the attacked propargylic carbon to the palladium (2.638 Å)
and to the tin (2.664 Å) atoms (Figure 2) are about equal. The
carbon-chlorine bond (1.952 Å) is much longer than the C-Cl
bond length in 3d (1.810 Å), indicating that a bond-breaking
process also takes place in the TS. Formation of propargyl
stannane (11) proceeds through TS structure 10b requiring 13
kcal mol-1 higher activation energy (33.5 kcal mol-1) than that
required for formation of the allenyl analogue 5c. The geometry
of the reaction center in 10b is similar to that in 10a, as the
attacked propargylic carbon interacts with both the palladium
A closer inspection of TS structures 10a and 10b reveals that
the geometry of the reaction center is similar to that of the
classical SN2′ and SN2 reactions, in which the formation of the
new bond and cleavage of the leaving group take place
simultaneously. There is, however, a clear difference between
the geometry of the TS of the classical nucleophilic substitution
reactions and that of the presented processes (10a,b). In a
classical SN2 reaction, the attacking nucleophile is aligned with
the reaction center and the leaving group closing an angle of
180°. However, the corresponding Sn-C-Cl angle in 10b is
132°, and furthermore, the palladium atom is also involved in
the displacement of the leaving group. On the other hand, the
middle point of the palladium-tin bond, the attacked propargylic
carbon, and the carbon-chloride bond do form an angle close
to 180° in the TS structures. This suggests that the electron pair
of the palladium-carbon σ-bond exerts a nucleophilic attack
on the propargylic carbon, cleaving the carbon-chloride bond.
This hypothesis can also be confirmed by inspection of the high-
energy (-4.5 eV) HOMO orbital of complex 1g (Figure 3),
which is largely localized to the palladium-tin σ-bond. Thus,
the electron pair of the palladium-tin σ-bond is readily available
for electrophiles, such as 3d. It is particularly interesting to
compare the HOMO of 1g to the corresponding HOMO of
hexamethylditin (2a). The HOMO in 2a has also a clear metal-
metal σ-bond character; however, the HOMO energy (-6.3 eV)
is considerably lower in 2a than in 1g. Because of this low
HOMO energy, 2a is not able to perform the nucleophilic
displacement of the chloride of 3d. On the other hand, palladium
pincer complex 1g has a high energy HOMO, and therefore
this complex is more nucleophilic than 2a, which leads to a
successful trimethyltin transfer to the propargylic substrate 3d.
This can also be considered as the main reason for employment
of palladium pincer complex catalysis in the above reaction.
(37) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M.
A.; Cheeseman, J. R.; Montgomery, J. A.; Vreven, T.; Kudin, K. N.; Burant,
J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.;
Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada,
M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima,
T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.;
Hratchian, H. P.; Cross, J. B.; Adamo, C.; Jaramillo, J.; Gomperts, R.;
Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.;
Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.;
Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain,
M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.;
Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.;
Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.;
Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng,
C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.;
Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03; Gaussian,
Inc.: Pittsburgh, PA, 2003.
As far as we know, the above-described palladium-metal
σ-electron transfer-initiated single-step displacement of the
9
J. AM. CHEM. SOC. VOL. 127, NO. 6, 2005 1793

A R T I C L E S
Kjellgren et al.
Figure 2. Calculated geometries for 1g and the TS’s of the trimethylstannane transfer reaction. The ZPV-corrected energies are given in italics. Energies
in kcal mol-1 and bond lengths in Å.
leaving group is a unique process in palladium chemistry. This
can be explained by the special topology of the pincer complex
catalyst. In a typical reactive pincer complex intermediate, such
as in 1g, the high-energy palladium-tin and palladium-carbon
σ-bonds are forced to a trans geometry. This topology does not
permit a reductive elimination involving the Pd-Sn and Pd-C
bonds. Thus, the high-energy Pd-Sn bond becomes available
for reactive external electrophiles, such as 3d (cf. 10a,b). On
the other hand, commonly occurring organopalladium interme-
diates have a considerably more flexible structure than pincer
complexes, and therefore the high-energy Pd-Sn and Pd-C
bonds may easily isomerize to a cis geometry, permitting a facile
intramolecular reductive elimination reaction.38
Formation of the allenyl (5c) and propargyl (11) stannane
products is a highly exothermic process (-47.3 and -53.5 kcal
mol-1), indicating that the trimethyltin transfer from the
palladium atom of 1g to 3d is an irreversible process. The
formation of allenylstannane 5c is both kinetically and thermo-
dynamically more favored than formation of propargyl stannane
11, at least from the parent compound 3d (entry 9). However,
(38) Albright, T. A.; Burdett, J. K.; Whangbo, M.-H. Orbital Interactions in
Chemistry; Wiley: New York, 1985.
9
1794 J. AM. CHEM. SOC. VOL. 127, NO. 6, 2005

Palladium Pincer Complex Catalyzed Stannyl and Silyl Transfer
A R T I C L E S
reactions of allenylstannanes and aryl iodides. To demonstrate
the synthetic utility of this reaction, we reacted compound 5s,
obtained from stannylation of 3t, as a single diastereomer (entry
28) with phenyl iodide in the presence of catalytic amounts of
6 (eq 8). This procedure gave mainly trans-substituted allene
12a50,51 and only traces of its syn-substituted counterpart 12b.51
Since it is known that the analogue Stille-coupling reactions
proceed with retention of the geometry of the vinyl-tin
bond,52,53 this reaction also helped us to assign the stereochem-
istry of 5s. Thus, as the major product of the coupling reaction
(12a) is the anti form, we conclude the same stereochemistry
for 5s.
Figure 3. HOMO orbital of complex 1g (ꢀHOMO ) -4.5 eV) and
hexamethylditin 2a (ꢀHOMO ) -6.3 eV).
substituent effects obviously influence the regioselectivity. For
example, the presence of bulky phenyl (3a) or alkyl (3c)
substituents at the attacked carbon in 10a leads to unfavorable
steric interactions with the pincer complex. This explains the
predominant formation of the propargylic product via 10b-type
TS from 3a,b (entries 1-8). On the other hand, electron-
withdrawing substituents (e.g., 3e, 3f, 3i, and 3j) increase the
electrophilicity of the unsaturated carbon, promoting the forma-
tion of the allenyl product via 10a-type TS. The final allenyl-
to-propargyl ratio of the reaction is determined by the coun-
teracting electronic and steric effects. From secondary propargylic
substrates (3n-u), allenyl products are formed (entries 21-
29) because of the strongly destabilizing steric interactions in
the TS structure of the SN2-type process (cf. 10b).
A further, synthetically interesting feature of the products of
the above-described pincer complex catalyzed reactions (eq 2)
is that these compounds comprise a trimethylstannyl group
instead of the more commonly used tributylstannyl functionality.
Since the trimethylstannyl group is less bulky than its tributyl-
stannyl counterpart, the presented allenylstannanes are expected
to be more reactive in electrophilic substitution and coupling
reactions than their tributylstannyl analogues.
6. Conclusions
Pincer complex catalysis can be employed for regioselective
transfer of the stannyl and silyl functionality to propargylic
substrates. The reactions proceed under mild and neutral
conditions, and therefore many functionalities are tolerated. The
catalytic activity of the employed catalyst strongly depends on
the electronic effects of the employed pincer ligand. Pincer
complexes with electron-supplying ligands, such as NCN, SCS,
and SeCSe complexes, display a very high catalytic activity.
The regioselectivity of the stannylation reaction of primary
propargylic substrates depends on the steric and electronic
effects of the substituents on the triple bond. Catalytic reactions
involving secondary propargylic substrates provide exclusively
the allenyl product. The epoxide opening via stannylation of
the corresponding propargylic substrates takes place with an
excellent regio- and stereochemistry. The silylation reactions
proceed with higher allenyl selectivity than the corresponding
stannylation processes. Interestingly, under the applied catalytic
conditions, exclusively the silyl functionality is transferred from
silylstannanes 2b,c to the propargylic substrates. Mechanistic
studies have shown that the active intermediate of the catalytic
reactions is an organostannyl- or silyl ligand-coordinated pincer
complex. According to DFT modeling, the stannyl group transfer
from palladium to propargyl chloride is a single-step process.
The displacement of the chloride is initiated by nucleophilic
attack of the palladium-tin σ-bond electrons on the propargylic
substrate. Since the above-described pincer complex catalyzed
process is characterized by mild reaction conditions, a high level
of selectivity, and operational simplicity, this transformation
5. Synthetic Utility of the Products
Allenylstannanes and silanes are versatile building blocks in
advanced organic synthesis and in natural product synthesis.39-46
Although many excellent procedures for selective synthesis of
these organometallic compounds are already available in the
literature,39-46 there is a considerable need to broaden the variety
of available functionalized species. Allenylstannanes and silanes
undergo highly stereoselective reactions with various aldehydes
in the presence of Lewis acid catalysts to afford homopropargyl
alcohol products.39 Allenylstannanes can also be utilized in
palladium-catalyzed coupling reactions, such as in Stille-type
coupling47,48 with aromatic halogenides to provide substituted
allenes.49 The palladium pincer catalysts 1a-e do not react with
allenylstannanes under the employed reaction conditions, which
is a prerequisite of the high yields obtained in the catalytic
transformations. On the other hand, commonly used palladium
sources, such as Pd(PPh3)4 (6), readily catalyze the coupling
(39) Marshall, J. A. Chem. ReV. 1996, 96, 31.
(40) Marshall, J. A.; DeHoff, B. S. J. Org. Chem. 1986, 51, 863.
(41) Marshall, J. A.; Wang, X.-J. J. Org. Chem. 1990, 55, 6246.
(42) Marshall, J. A.; Wang, X.-J. J. Org. Chem. 1991, 56, 3211.
(43) Marshall, J. A.; Wang, X.-J. J. Org. Chem. 1992, 57, 1242.
(44) Ranslow, R. B. D.; Hegedus, L. S.; Rios, C. D. L. J. Org. Chem. 2004, 69,
105.
(45) Ruitenberg, K.; Westmijze, H.; Meijer, J.; Elsevier, C. J.; Vermeer, P. J.
Organomet. Chem. 1983, 241, 417.
(46) Jeganmohan, M.; Shanmugasundaram, M.; Cheng, C.-H. Org. Lett. 2003,
5, 881.
(47) Stille, J. K. Angew. Chem., Int. Ed. Engl. 1986, 25, 508.
(48) Espinet, P.; Echavarren, A. M. Angew. Chem., Int. Ed. 2004, 43, 4704.
(49) Huang, C.-W.; Shanmugasundram, M.; Chang, H.-M.; Cheng, C.-H.
Tetrahedron 2003, 59, 3635.
(50) Alexakis, A.; Marek, I.; Mangey, P.; Normant, J. F. Tetrahedron 1991,
47, 1677.
(51) Fu¨rstner, A.; Me´ndez, M. Angew. Chem., Int. Ed. 2003, 42, 5355.
(52) Corriu, R. J. P.; Geng, B.; Moreau, J. J. E. J. Org. Chem. 1993, 58, 1443.
(53) Reginato, G.; Mordini, A.; Caracciolo, M. J. Org. Chem. 1997, 62, 6187.
9
J. AM. CHEM. SOC. VOL. 127, NO. 6, 2005 1795

A R T I C L E S
Kjellgren et al.
centrifugation, and the solvent was evaporated to yield 1f. Complex 1f
was dissolved in 0.7 mL of THF-d8, and then hexamethylditin (63 mg,
0.192 mmol) was added to the solution. The progress of the reaction
provides an attractive new route for preparation of functionalized
allenylstannanes and silanes.
7. Experimental Section54
1
was monitored by using H and 119Sn NMR spectroscopy. Under the
above conditions, about 50% of 1f was converted to tin complex 1g
according to the 1H NMR spectrum of the reaction mixture. NMR data
(ppm) for complex 1g (determined from the obtained reaction mix-
ture): 1H NMR δ 0.05 (s, 9H), 2.87 (s, 12H), 3.98 (s, 4H), 6.64 (d,
3J(H-H) ) 7.9 Hz, 2H), 6.71 (t, 3J(H-H) ) 7.9 Hz, 1H); 119Sn NMR
δ -0.2 ppm.
Catalytic Stannylation with 2a. General Procedure A. The
corresponding propargylic substrates 3a-u (0.16 mmol) and catalyst
1 (0.004 mmol, 2.5 mol %) were dissolved in THF (0.25 mL). To this
mixture was added hexamethylditin 2a (0.168 mmol), resulting in a
dark-yellow solution. This reaction mixture was then stirred for the
allotted temperatures and times given in Table 1. The solvent was
evaporated, and the products were purified by column chromatography.
To avoid decomposition of the allenyl and propargylstannane products,
a rapid chromatography procedure was applied.
Catalytic Silylation with 2b. General Procedure B. The corre-
sponding propargylic substrate (0.16 mmol), catalyst 1a (0.008 mmol,
5 mol %), and 4 Å molecular sieves in THF (0.20 mL) were heated to
40 °C. To this mixture was added (dimethylphenylsilyl)trimethyl-
stannane 2b (0.24 mmol) in THF (0.25 mL) via a syringe-pump over
5 h at 40 °C, and the reaction mixture was stirred for an additional 19
h at 40 °C. Then the solvent was evaporated, and the products were
purified by column chromatography. To avoid decomposition of the
allenylsilane products, a rapid chromatography was applied (see General
Procedure A). The same procedure was employed for silylation with
(trimethylsilyl)tributylstannane (2c), except that this reagent was added
in one portion and the reaction was conducted for 4 h at 60 °C.
Reaction of Complex 1b with 2a Monitored by NMR. The
chloride counterion of 1a was exchanged to acetate (forming 1f) by
using the following procedure. Complex 1a (12 mg, 0.032 mmol) was
added to AgOAc (8 mg, 0.048 mmol) suspended in CHCl3 at 0 °C.
This mixture was stirred 1 h at 0 °C and then an additional hour at 25
°C in a dark reaction vessel. The AgBr precipitation was separated by
Reaction of Complex 1a with Silylstannane 2b Monitored by
NMR. To complex 1a (12 mg, 0.032 mmol) in THF-d8 (0.20 mL) was
added silylstannane 2b (29 mg, 0.096 mmol, 3 equiv) in THF-d8 (0.25
mL) over 5 h via a syringe-pump at room temperature. Then, the
reaction mixture was analyzed by 1H and 29Si NMR spectroscopy.
1
According to H NMR spectroscopy under these reaction conditions,
1a was fully converted to silicon complex 1h. NMR data (ppm) for
complex 1h (determined from the obtained reaction mixture): 1H NMR
3
δ 0.18 (s, 6H), 2.91 (s, 12H), 3.92 (s, 4H), 6.73 (d, J(H-H) ) 8.4
3
Hz, 2H), 6.89 (t, J(H-H) ) 8.4 Hz, 1H), 7.27-7.41 (m, 5H); 29Si
NMR δ -5.5 ppm.
Acknowledgment. This work was supported by the Swedish
Natural Science Research Council (VR).
Supporting Information Available: Detailed experimental
procedures, characterization and 13C NMR spectra of the
prepared allenyl and propargyl silanes and stannanes, 119Sn and
29Si NMR spectra of the stoichiometric reactions with pincer
complexes and dimetal reagents, and computational details of
the DFT studies. This material is available free of charge via
the Internet at http://pubs.acs.org.
(54) General experimental conditions, detailed experimental procedures and
characterization of the products are given in the Supporting Information.
JA043951B
9
1796 J. AM. CHEM. SOC. VOL. 127, NO. 6, 2005