Palladium-Catalyzed Stille Couplings with Fluorous Tin Reactants

Article abstract of DOI:10.1021/jo9709413

A new class of "fluorous" aryl tin reactants was investigated for use in the Stille coupling. Fluorous compounds partition into a fluorocarbon (fluorous) phase in a fluorous/organic extraction. The coupling of tris[(perfluorohexyl)ethyl]phenyl tin [(C₆F₁₃CH₂CH₂)₃SnPh] with bromobenzene occurs smoothly in DMF/THF (1/1) at 80 °C in the presence of a catalytic amount of PdCl₂(PPh₃)₂ and 3 equiv of LiCl. Partitioning between CH₂Cl₂ and FC-72 (a mixture of perfluorohexanes) provided biphenyl in 90% yield from the organic phase and tris[(perfluorohexyl)ethyl]tin chloride in > 90% yield from the fluorous phase. The tin chloride was reacted with phenylmagnesium bromide to regenerate the starting tin reactant. A study to optimize reaction conditions is described, and the scope of the method is illustrated with 20 coupling reactions. The beneficial effect of lithium choride is an unusual feature of the reaction, but it also promotes the formation of some fluoroalkyl-coupled products. These can be suppressed by adding CuI. The paper describes a prototypical example of how to render a tin reactant fluorous. This process should be advantageous in small- and large-scale synthesis, as well as in automated synthesis.

Full text of DOI:10.1021/jo9709413

J . Org. Chem. 1997, 62, 8341-8349
8341
P a lla d iu m -Ca ta lyzed Stille Cou p lin gs w ith F lu or ou s Tin Rea cta n ts
Masahide Hoshino, Peter Degenkolb, and Dennis P. Curran*
Department of Chemistry, University of Pittsburgh, Pittsburgh, Pennsylvania 15260
Received May 28, 1997X
A new class of “fluorous” aryl tin reactants was investigated for use in the Stille coupling. Fluorous
compounds partition into a fluorocarbon (fluorous) phase in a fluorous/organic extraction. The
coupling of tris[(perfluorohexyl)ethyl]phenyl tin [(C
smoothly in DMF/THF (1/1) at 80 °C in the presence of a catalytic amount of PdCl
equiv of LiCl. Partitioning between CH Cl and FC-72 (a mixture of perfluorohexanes) provided
6
F
13CH
2
CH
2
)
3
SnPh] with bromobenzene occurs
2
(PPh and 3
3 2
)
2
2
biphenyl in 90% yield from the organic phase and tris[(perfluorohexyl)ethyl]tin chloride in >90%
yield from the fluorous phase. The tin chloride was reacted with phenylmagnesium bromide to
regenerate the starting tin reactant. A study to optimize reaction conditions is described, and the
scope of the method is illustrated with 20 coupling reactions. The beneficial effect of lithium choride
is an unusual feature of the reaction, but it also promotes the formation of some fluoroalkyl-coupled
products. These can be suppressed by adding CuI. The paper describes a prototypical example of
how to render a tin reactant fluorous. This process should be advantageous in small- and large-
scale synthesis, as well as in automated synthesis.
In tr od u ction
tial for facilitating purifications in organic synthesis.8,9
For extraction purposes, the fluorous liquid phase con-
One of the current driving forces in organic synthesis
is efficiency. In the small scale area, parallel synthesis
is helping to make new compounds available more
quickly, so the discovery of interesting new molecules
sists of low-boiling, perfluorinated or highly fluorinated
10
hydrocarbons, ethers, and amines.
Many of these
liquids are commercially available and are routinely used
8
in areas outside synthetic chemistry. In the simple view,
1
-3
becomes more efficient with respect to time.
On the
the fluorous liquid phase constitutes a third extraction
phase because its members are all immiscible in water,
and they are immiscible in many common organic
solvents as well. An assortment of two- and three-phase
extractions can be conducted with fluorous, organic, and
water (acid, base, or neutral) phases.
Organic compounds will not generally partition into a
fluorous phase, so the use of a fluorous phase in an
extraction requires that at least one of the reaction
components be neither organic nor inorganic but fluorous.
Our research is focusing methods to convert organic
compounds to fluorous ones for use in synthesis. This
4
large-scale front, atom economy and associated needs
to recycle reagents and to minimize waste provide
different kinds of driving forces for chemical efficiency.
Much of the effort to increase efficiency has focused on
the purification stage of organic reactions, where there
has been renewed interest in using simple phase separa-
tion techniques to accomplish substantive separations.
For example, diverse solid-phase methods exploit the
power of the simple technique of filtration.1
,2,5
Purifications based on extractions are usually limited
to the separation of neutral organic compounds from
inorganic compounds or organic salts. Organic-aqueous
extractions are used routinely, yet they lack the power
to effect substantive separations of most types of organic
was accomplished in a “permanent” fashion by Horv a´ th
and R a´ bai,7 who prepared and used a fluorous phos-
phine: (C
a new tin hydride [(C
6
F
13CH
2
11
CH
2
)
3
P. We have recently introduced
CH SnH] as example of
6
compounds. Inspired by the work of Zhu and Horv a´ th
6
F
13CH
2
2 3
)
7
and R a´ bai, we have suggested that the “fluorous liquid
a fluorous reagent (and catalyst), and we have com-
phase” provides a third liquid phase of untapped poten-
11
municated the use of fluorous aryltin reactants for Stille
couplings.1
2a
In these embodiments of the fluorous
X
Abstract published in Advance ACS Abstracts, November 1, 1997.
strategy, the substrate and the product are organic and
one of the other reaction components and its byproduct
are fluorous. More recently, we have used organic and
inorganic reagents in combination with “temporarily”
fluorous substrates and products.13 In this paper, we
report the full details of our study of the Stille coupling
(1) Solid phase: a) Gallop, M. A.; Barrett, R. W.; Dower, W. J .; Fodor,
S. P. A.; Gordon, E. M. J . Med. Chem. 1994, 37, 1233 and 1386. b)
Terrett, N. K.; Gardner, M.; Gordon, D. W.; Kobylecki, R. J .; Steele, J .
Tetrahedron 1995, 51, 8135. c) Thompson, L. A.; Ellman, J . A. Chem.
Rev. (Washington, D.C.) 1996, 96, 555. d) Gordon, E. M.; Gallop, M.
A.; Patel, D. V. Acc. Chem. Res 1996, 29, 144. e) Rinnova, M.; Lebl, M.
Collect. Czech. Chem. Commun. 1996, 61, 171. f) Balkenhohl, F.; von
dem Bussche-H u¨ nn efeld, C.; Lansky, A.; Zechel, C. Angew. Chem.,
Int. Ed. Engl. 1996, 35, 2288.
(
2) Soluble polymers: a) J ung, K. W.; Zhao, X. Y.; J anda, K. D.
(7) a) Horv a´ th, I. T.; R a´ bai, J . Science 1994, 266, 72. b) Horv a´ th, I.
T.; R a´ bai, J ., U.S. Patent no. 5,463,082, 1995.
(8) Curran, D. P. Chemtracts-Org. Chem. 1996, 9, 75.
Tetrahedron Lett. 1996, 37, 6491. b) Han, H. S.; J anda, K. D. J . Am.
Chem. Soc. 1996, 118, 7632.
(
3) Solution phase: a) Carell, T.; Wintner, E. A.; Sutherland, A. J .;
Rebek, J ., J r.; Dunayevskiy, Y. M.; Vouros, P. Chemistry and Biology
995, 2, 171. b) Boger, D. L.; Tarby, C. M.; Myers, P. L.; Caporale, L.
(9) For new fluorous applications, see: a) Hughes, R. P.; Trujillo,
H. A. Organomet. 1996, 15, 286. b) Di Magno, S. G.; Dussault, P. H.;
Schultz, J . A. J . Am. Chem. Soc. 1996, 118, 5312. c) Pozzi, G.; Banfi,
S.; Manfredi, A.; Montanari, F.; Quici, S. Tetrahedron 1996, 52, 11879.
(10) a) Scott, R. L. J . Am. Chem. Soc. 1948, 70, 4090. b) Scott, R. L.
J . Phys. Chem. 1958, 62, 136. c) Hildebrand, J . H.; Cochran, D. R. F.
A. J . Am. Chem. Soc. 1949, 71, 22. See also: d) Hudlicky, M. Chemistry
of Organic Fluorine Compounds; Ellis Horwood: Chichester, UK, 1992.
(11) Curran, D. P.; Hadida, S. J . Am. Chem. Soc. 1996, 118, 2531.
(12) a) Curran, D. P.; Hoshino, M. J . Org. Chem. 1996, 61, 6480. b)
Larhed, M.; Hoshino, M.; Hadida, S.; Curran, D. P.; Hallberg, A. J .
Org. Chem., in press.
1
H. J . Am. Chem. Soc. 1996, 118, 2109. c) Cheng, S.; Comer, D. D.;
Williams, J . P.; Myers, P. L.; Boger, D. L. J . Am. Chem. Soc. 1996,
18, 2567.
1
(
(
4) Trost, B. M. Angew. Chem., Int. Ed. Engl. 1995, 34, 259.
5) a) Brown, S. D.; Armstrong, R. W. J . Am. Chem. Soc. 1996, 118,
6
331-32. b) Dressman, B. A.; Spangle, L. A.; Kaldor, S. W. Tetrahedron
Lett. 1996, 37, 937. c) Booth, R. J .; Hodges, J . C.; submitted for
publication.
(
6) Zhu, D.-W. Synthesis 1993, 953.
S0022-3263(97)00941-9 CCC: $14.00 © 1997 American Chemical Society

8
342 J . Org. Chem., Vol. 62, No. 24, 1997
Hoshino et al.
with fluorous aryltin reactants.12a In a parallel paper,
we describe a microwave version of this reaction that is
Stille couplings and (2) whether the expected purification
features of the fluorous method were manifested.
1
2b
both convenient and of broad scope.
The Stille,14 Suzuki, and related cross couplings have
made accessible, even routine, the formation of many
kinds of carbon-carbon bonds that were previously
difficult to make under mild conditions. Solid-phase
variants have recently been introduced.16 The defining
feature of the Stille coupling is the use of a trialkyltin
reagent in a palladium-catalyzed coupling with a halide
or triflate. The alkyl groups, usually methyl or butyl,17
are designed as “non-transferable” ligands, and the fourth
ligand on tin is coupled to a halide or triflate.
15
Resu lts a n d Discu ssion
Equation 2 shows the reaction and purification strat-
egy for Stille couplings with fluorous tin reactants. We
(2)
(1)
decided to synthesize and study the reactions of tris(2-
(
perfluorohexyl)ethyl]aryltin [1, (C
6 2 2 3
F13CH CH ) SnAr] as
The Stille reaction is popular because of its scope and
functional group tolerance and because tin reagents are
easy to make and are very stable compared to most
organometallic derivatives. Like most areas of tin chem-
istry, the detractions come not in the reaction stage but
in the purification stage. Low molecular weight trialkyl-
tin compounds are toxic, can be difficult to separate, and
are rarely recovered and recycled. Comparing the tin
derivatives used in the Stille couplings, trimethyltin
halides are easy to remove because of their water
solubility and volatility, but they are toxic. Also, methyl
transfer sometimes occurs as a significant side reaction
with trimethyltin reagents. Tributyltin halides are less
toxic but much more difficult to remove because they are
nonvolatile and insoluble in water but soluble in most
a prototypical example of a fluorous tin reagent. Com-
pound 1 has perfluorohexylethyl groups in place of the
usual butyl groups as nontransferable ligands on tin. The
perfluorohexyl group supplies fluorous solubility while
7
the ethylene spacer serves to insulate the tin from the
powerful inductive effect of the perfluorohexyl group.
The procedure is designed to facilitate the purification
of the target organic product and to allow ready recovery
of fluorous tin compounds for reuse. In the reaction stage
of the Stille coupling, a substrate 2 is treated under
typical Stille conditions with 1, which can be used either
in stoichiometric quantities or in excess. The purification
stage begins with three-phase extraction between an
organic solvent, a fluorous solvent, and water. The
organic phase of the extraction contains the target cross-
coupled product 3 along with a small amount of homo-
coupled product 4 derived from 1. (Homocoupling is a
common side reaction in Stille couplings.)14 The fluorous
phase contains excess 1, if any, and the fluorous tin
halide 5, which can readily be recycled to the same or a
different fluorous tin reactant. The water phase contains
inorganic salts, and this is discarded. If necessary, the
organic components of the reaction can be further sepa-
rated by conventional means.
1
8
common organic solvents.
These features make the Stille coupling a prime target
as a test for the strategy of using fluorous reactants. We
therefore set out to see if aryltin reagents could be
rendered “fluorous” by suitable substitutions and to learn
(1) whether the resulting reactants were suitable for
(13) a) Studer, A.; Hadida, S.; Ferrito, R.; Kim, S.-Y.; J ager, P.; Wipf,
P.; Curran, D. P. Science 1997, 275, 823. b) Studer, A.; Curran, D. P.
Tetrahedron, in press. c) Curran, D. P.; Studer, A.; J eger, P.; Wipf, P.
J . Org. Chem., in press.
Equation 3 shows the syntheses of the main fluorous
tin reactants used in this study. Preparation of the
Grignard reagent from commercially available 2-(per-
fluorohexyl)-1-iodoethane and quenching with phenyl-
(
14) a) Stille, J . K. Pure Appl. Chem. 1985, 57, 1771. b) Stille, J . K.
Angew. Chem., Int. Ed. Engl. 1986, 25, 508. c) Mitchell, T. N. Synthesis
992, 803. d) Farina, V.; Roth, G. P. In Advances in Metal-Organic
1
Chemistry; Liebeskind, L. S., ed.; J AI: Greenwich, 1995, vol. 5. e)
Farina, V. Pure Appl. Chem. 1996, 68, 73. Early reports: f) Kosugi,
M.; Sasazawa, K.; Shimizu, Y.; Migita, T. Chem. Lett. 1977, 301. g)
Kosugi, M.; Shimizu, Y; Migita, T. Chem. Lett. 1977, 1423. h) Kosugi,
M.; Shimizu, Y.; Migita, T. J . Organomet. Chem. 1977, 129, C-36. i)
Milstein, D.; Stille, J . K. J . Am. Chem. Soc. 1978, 100, 3636. j) Milstein,
D.; Stille, J . K. J . Am. Chem. Soc. 1979, 101, 4992. k) Labadie,
J . W.; Stille, J . K. J . Am. Chem. Soc. 1983, 105, 6129. l) Scott, W. J .;
Stille, J . K. J . Am. Chem. Soc. 1986, 108, 3033. m) Echavarren, A. M.;
Stille, J . K. J . Am. Chem. Soc. 1987, 109, 5478. n) Chen, Q.-Y.; He,
Y.-B. Chin. J . Chem. 1990, 451. o) An authoritative review has just
appeared. Farina, V.; Krshnamurthy, V.; Scott, W. J . Org. React. 1997,
trichlorotin provided fluorous phenyltin reactant 1a
11
[tris[2-(perfluorohexyl)ethyl]phenyl tin]. Brominolysis
5
0, 1.
(15) Miyaura, N.; Suzuki, A. Chem. Rev. (Washington, D.C.) 1995,
9
5, 2457.
(3)
(16) a) Deshpande, M. S. Tetrahedron Lett. 1994, 35, 5613. b)
Forman, F. W.; Sucholeiki, I. J . Org. Chem. 1995, 60, 523. c) Plunkett,
M. J .; Ellman, J . A. J . Am. Chem. Soc. 1995, 117, 3306. d) Guiles, J .
W.; J ohnson, S. G.; Murray, W. V. J . Org. Chem. 1996, 61, 5169.
(17) Interesting alternatives to the standard methyl and butyl
reagents are available. For example, see: a) Vedejs, E.; Haight, A. R.;
Moss, W. O. J . Am. Chem. Soc. 1992, 114, 6556. b) Brown, J . M.;
Pearson, M.; J astrzebski, J . T. B. H.; van Koten, G. J . Chem. Soc.,
Chem. Commun. 1992, 1440.
(
18) For a new separation and recyling procedure and references to
other procedures, see: Crich, D.; Sun, S. R. J . Org. Chem. 1996, 61,
200.
7

Stille Couplings with Fluorous Tin Reactants
J . Org. Chem., Vol. 62, No. 24, 1997 8343
Ta ble 1. Effect of Va r yin g Rea ction Con d ition s on th e Cou p lin g betw een 1a a n d Ar yl Ha lid e 2a a
entry
catalyst
LiCl
solvent
T (°C)
% yield of 3a b
% yield of 4a c
1
2
3
4
5
6
7
8
9
PdCl2(PPh3)2
PdCl2(PPh3)2
Pd(PPh3)4
Pd2dba3+AsPh3
PdCl2(PPh3)2
PdCl2(PPh3)2
Pd(PPh3)4
-
+
+
+
-
+
+
+
+
-
+
THF
THF
THF
THF
DMF
DMF
DMF
DMF
NMP
NMP
BTF
67
67
67
67
80
80
80
80
80
80
80
incomplete (both 2a /3a d)
90
6
d
incomplete (mostly 2a )
incomplete (both 2a /3a )
incomplete (both 2a /3a )
e
d
d
82
88
5
5
e
e
d
Pd2dba3+AsPh3
incomplete (mostly 2a )
incomplete (both 2a /3a )
incomplete (mostly 2a )
incomplete (both 2a /3a )
d
Pd2dba3+AsPh3
f
d
1
1
0
1
Pd/C+AsPh3+CuI
d
PdCl2(PPh3)2
a
Reactions were performed for 46 h under N2, 48 mM 1a , 40 mM 2a , 0.8 mM catalyst unless otherwise stated, 0.12 M LiCl if it was
b
c
d
added. Isolated yield based on 2a . Isolated yield based on 1a . Estimated by TLC with UV visualization; mostly 2a ) TLC spot for
e
2
a clearly larger; mostly 3a ) TLC spot for 3a clearly larger; both 2a /3a ) TLC spots roughly comparable. 0.4 mM Pd2dba3, 3.2 mM
f
AsPh3. 0.24 mM Pd/C, 8 mM AsPh3, 4 mM CuI.
of 1a provided the tin bromide 5a , which served as the
precursor for preparing the 4-methoxyphenyl (1b), 2-furyl
proceeded for some time prior to catalyst deactivation.
This differentiation was not deemed crucial at this stage
since neither type of reaction is suitable for preparative
purposes.
(1c), and 2-pyridyl (1d ) fluorous tin reactants by standard
reactions with either Grignard or aryllithium reagents.
As an aside, the tin bromide 5a also serves as the
THF was the first solvent that we tried because it has
good dissolving power for both organic and fluorous
compounds. Indeed, the reactions in THF were homo-
geneous from start to finish as judged by the naked eye
(Table 1, entries 1-4). While the reaction with PdCl2-
(PPh ) at reflux in the absence of LiCl was slow and
1
1
precursor for a fluorous tin hydride and several other
fluorous tin reagents.14
A number of efficient reaction conditions for Stille
1
5a-e
couplings now exist,
and we initially surveyed some
of these to identify any characteristic behavior of fluorous
tin compounds. At the outset, we were seriously con-
cerned that fluorous tin reactants might be less reactive
than standard alkyltin reagents. While little is known
about the effects of electron-withdrawing groups on the
3
2
incomplete even after 46 h (Table 1, entry 1), the same
reaction in the presence of 3 equiv of LiCl was complete
and gave 90% yield of desired cross-coupled product 3a
and 5% yield of homocoupled product 4a derived from
1a (Table 1, entry 2). These results are interesting
because LiCl is not usually beneficial in reactions of
normal aryltin reactants with aryl halides. The effect
of LiCl and other additives was later studied in more
detail as described below. The reaction with Pd(PPh3)4
and LiCl in THF proceeded very slowly and was incom-
plete after 16 h (Table 1, entry 3), while Pd dba + AsPh
“dummy ligands” of tin, it is well-known that the R
3
SnX
(X ) halide) products of Stille couplings are very resis-
tant to a second Stille coupling. This raises the concern
that the perfluoroalkyl groups might retard Stille cou-
plings if the ethylene spacer were an insufficient insula-
tor. In addition to the effects on transmetalation, we
viewed with special importance the task of identifying
suitable reaction solvents because the fluorous tin re-
agents are not “organic” (as measured by fluorous/organic
extraction) and because Stille couplings are often affected
by the combination of reaction solvent, Pd catalyst, and
additive.
2
3
3
+ LiCl19 was an even less effective combination (Table
1, entry 4).
Next, the use of DMF as reaction solvent was studied
(Table 1, entries 5-8). At the standard reaction concen-
tration (0.05 M in 1a ), DMF partially dissolves fluorous
phenyltin 1a at 80 °C, and it dissolves the fluorous tin
chloride 5b even better. The reaction mixture in DMF
was two phases at the beginning, and then it became
homogeneous as the reaction proceeded. In DMF at 80
An initial series of experiments was conducted with
fluorous phenyltin reactant 1a and 4′-bromoacetophenone
(2a ), and the results of some of these experiments are
shown in Table 1. Reactions were conducted in THF,
DMF, and NMP (N-methylpyrrolidinone), with an as-
sortment of catalysts in the presence and absence of
lithium chloride. A 20% excess of fluorous phenyl tin
reagent 1a was used relative to 2a . Reactions judged
complete by TLC were worked up by three-phase extrac-
tion, and the isolated yields of the cross-coupled and
homocoupled products were determined by preparative
TLC. Many of the reactions were still incomplete after
°
2 3 2 3 4
C with LiCl present, PdCl (PPh ) and Pd(PPh ) both
catalyzed the reaction well (Table 1, entries 6 and 7).
However, in the absence of LiCl the reaction was again
slow and incomplete (Table 1, entry 5). Pd
did not give complete conversion in either DMF or NMP
Table 1, entries 8 and 9), though the NMP reaction
mixture was homogeneous from the beginning. The
2
dba
3
+ AsPh
3
(
20
reaction with Pd/C + AsPh
proceed well (Table 1, entry 10). The partially fluori-
nated solvent BTF (benzotrifluoride, C CF ) dissolves
both 1a and 2a even at room temperature, but the
3
+ CuI in NMP also did not
4
6 h, and these were not worked up or purified. For these
we provide a very rough estimate of the progress of the
reaction. This estimate is based on TLC spot intensity,
so it has no quantitative meaning. Furthermore, reaction
rates were not measured, so we cannot differentiate
reactions that are inherently slow from those that
6
H
5
3
(
19) Farina, V.; Krishnan, B. J . Am. Chem. Soc. 1991, 113, 9585.
(20) Roth, G. P.; Farina, V. Tetrahedron Lett. 1995, 36, 2191.

8
344 J . Org. Chem., Vol. 62, No. 24, 1997
Hoshino et al.
Ta ble 2. Effect of Va r yin g Rea ction Con d ition s on th e Cou p lin g betw een 1a a n d Ar yl Ha lid e 2ba
entry
catalyst
LiCl
solvent
T (°C)
% yield of 3b b
incomplete (trace 3bd)
no reaction
% yield of 4a c
1
2
3
4
5
6
PdCl2(PPh3)2
Pd(PPh3)4
Pd2dba3+AsPh3
PdCl2(PPh3)2
Pd(PPh3)4
+
+
+
+
+
+
+
THF
THF
THF
dioxane
dioxane
DMF
67
67
67
100
100
80
e
d
incomplete (trace 3b )
no reaction
no reaction
PdCl2(PPh3)2
Pd(PPh3)4
82
70
9
6
f
g
7
DMF
80
a
Reactions were performed under N2 for 25 h except entry 7, 48 mM 1a , 40 mM 2b, 0.8 mM catalyst unless otherwise stated, 0.12 M
b
c
d
e
LiCl if it was added. Isolated yield based on 2b. Isolated yield based on 1a . Observed by TLC with UV visualization. 0.4 mM
Pd2dba3, 3.2 mM AsPh3. Reaction was performed for 46 h. 17% of p-nitrophenol was also isolated.
f
g
Ta ble 3. Use of Mixtu r es of Rea ction Solven ts for th e
Cou p lin g betw een 1a a n d Ar yl Ha lid e 2a Ca ta lyzed by
fluorocarbon fluid that is part of the 3M Fluorinert
Liquids family. It consists mostly of isomers of perfluo-
rohexane C
14 with the linear isomer being the major
a
P d Cl2(P P h 3)2 in th e P r esen ce of LiCl
6
F
yield (%)
reaction
one. FC-72 has a convenient boiling point (56 °C), and
it provides favorable partition coefficients for the fluorous
tin reagents when paired with many organic solvents;
however, it is rather inexpensive (1 L ≈ $200). FC-77
entry
solvent
DMF
DMF/THF (1:1)
DMF/BTF (1:1)
T (°C)
time (h)
3a b
4a c
1
2
3
80
67
67
46
22
22
82
90
92
5
7
9
(
perfluorooctanes) can be used interchangeably with
a
FC-72, and although the price of FC-77 is about the same
as FC-72, its higher boiling point (90 °C) makes it easier
to recover for reuse. Simple extractions demonstrated
that fluorous tin reactants 1a and fluorous tin byproduct
(C F CH CH ) SnCl (5b ) readily partition into the
FC-72 phase. We estimate that the partition coefficients
of 1a and 5b between FC-72 and either dichloromethane
or toluene are g97/3. However, DMF does have a
detrimental effect on the fluorous tin halide partitioning
(see below).
Reactions were performed under N2, 0.24 M 1a , 0.20 M 2a , 4
b
mM PdCl2(PPh3)2, 0.6 M LiCl.
Isolated yield based on 1a .
Isolated yield based on 2a .
c
2 3 2
reaction in BTF with PdCl (PPh ) was only about half
complete after 46 h (Table 1, entry 11).
Next, we studied the coupling of 4-nitrophenyl triflate
b because triflates often require different reaction
6
13
2
2 3
2
conditions from halides (Table 2). It is expected from
past studies1
5l,n
that LiCl will be required for the tri-
flate coupling, so this was included in all reactions. In
THF, none of the catalysts tried gave significant amounts
of the Stille product 3b after reflux for 25 h (Table 2,
entries 1-3). Reactions in dioxane at 100 °C were
homogeneous but gave no product 3b at all (Table 2,
entries 4 and 5). On the other hand, in DMF at 80 °C,
A preparative reaction was conducted with phenyltin
reactant 1a (2.4 mmol) and 1-bromo-4-nitrobenzene 2c
-3
(
2 mmol) in the presence of PdCl
2
(PPh
3
)
2
(4 × 10 mmol)
and LiCl (6 mmol) in 10 mL of 1/1 DMF/THF at 67 °C
for 22 h (eq 4). Both the substrate 2c and tin reactant
both of PdCl
tion well, but PdCl
82%) of 3b in a shorter reaction time (Table 2, entries 6
and 7).
3 2
These results suggested that the use of PdCl (PPh )
2
(PPh
3
)
2
2
and Pd(PPh
3 4
) catalyzed the reac-
(PPh gave a better isolated yield
3 2
)
(
2
(4)
in the presence of LiCl in DMF provided a single set of
reaction conditions suitable for couplings with both
halides and triflates. That common reaction conditions
can be used for different types of substrates is advanta-
geous for parallel synthesis. However, these reactions
were still rather slow in some cases. To provide better
solubility for the fluorous tin reagent, we used mixtures
of DMF and another solvent. Equal parts of DMF/THF
6 5 3
and DMF/C H CF both gave homogeneous reaction
mixtures from the start, and the reactions occurred at
reasonable rates (<22 h) at 67 °C to provide a high yield
2 3 2
of desired product 3a (Table 3). The use of PdCl (PPh )
in the presence of LiCl in DMF/THF mixture (1/1) was
selected for the standard experiments. As we describe
below, the reaction time in DMF/THF can be reduced to
1a were consumed according to TLC analysis. The
reaction mixture was evaporated with toluene at 75 °C
(to remove the THF and some of the DMF) and then
partitioned in a three-phase extraction between water
(top), dichloromethane (middle), and FC-72 (bottom).
After this first extraction, only 81% of fluorous tin
chloride 5b was recovered from the FC-72 phase. The
dichloromethane phase was then washed three more
times with water and FC-72 (together) to remove the
remaining fluorous products and DMF. Evaporation of
1
0 h by heating at 80 °C in a sealed tube.
We selected FC-72 as the fluorous solvent for the three-
phase extraction that is the crucial feature of the
purification stage. This is a commercially available

Stille Couplings with Fluorous Tin Reactants
J . Org. Chem., Vol. 62, No. 24, 1997 8345
Ta ble 4. Yield sa of Cr oss-Cou p led P r od u cts 3 F or m ed
fr om F lu or ou s Tin Rea cta n ts 1 a n d Su bstr a tes 2
Ta ble 5. Com p a r ison Betw een F lu or ou s Tin Rea cta n ts
1e,f a n d Bu tyltin Rea cta n ts 6a ,b
b,c
tin reactant 1
substrate 2
C6H5I
1a (% yield) 1b (% yield) 1c (% yield)
90d,e
97 (10)
87 (6)
98 (8)
45f
72
93
4
4
4
-CH3COC6H4Br, 2a
-NO2C6H4Br, 2c
-NO2C6H4OTf, 2b
90 (7)
94 (9)
82 (9)
77 (5)
e
g
h
86 (8)
98 (7)
83
C6H5CH2Br
32f
a
Isolated yield based on 2. b Unless otherwise stated, reactions
were performed in sealed tubes at 80 °C for 22 h under N2, 1 (0.24
-
3
mmol), 2 (0.2 mmol), PdCl2(PPh3)2 (4 × 10 mmol), LiCl (0.6
c
mmol), mixture (1:1) of DMF/THF (1 mL). Isolated yields (%) of
d
symmetrical biaryl 4 based on 1 in parentheses. 3 and 4 are
e
the same (biphenyl). Solvent was THF (5 mL), in unsealed tube
f
g
at 67 °C for 9 h. Volatile product. Solvent was DMF (5 mL),
h
in unsealed tube at 80 °C for 25 h. About 10% of 4-nitrophenol
entry LiCl R1
R2 ratio 7/8 yield (%) reactn time (n)
was also isolated.
1
2
3
4
+
-
+
-
Et
Et
Et
Me
Me
Me
1/2.3
1.8/1
1.8/1
1/2.1
91
84
80
85
1
8
1
these combined fluorous phases provided 18% more of
the tin chloride 5b, raising its total recovered yield to
b
Me Et
20
9
9%.
The crude organic product from the dichloromethane
phase was purified by flash chromatography to provide
-nitrobiphenyl (3b) (85% yield based on 2c) and biphenyl
4a ) (5% yield based on 1a ). The crude fluorous tin
a
Reactions were performed in sealed tubes under Ar with 0.4
b
M 1e,f 0.2 M 2a , 0.004 M Pd, and 0.6 M LiCl (if used). At 75%
conversion.
4
(
duced with 4-nitrophenyl triflate and bromide and with
iodobenzene; however, yields of pure products were not
determined. These poor results parallel observations
under similar conditions with the related pyridyltribu-
tyltin reagent.23
chloride 5b from the FC-72 phase was treated with
phenylmagnesium bromide to provide the original tin
reactant 1a (96% overall yield from original 1a ) after
purification by passing through a short column of neutral
alumina.
To survey the scope of the reaction, we conducted 20
small-scale (0.2 mmol) Stille reactions. Couplings be-
tween various fluorous tin reactants (1a -d ) and halides
or triflates (2a -e) were performed under the standard
reaction conditions in sealed tubes at 80 °C for 22 h.
Reactions were conducted in individual vessels in groups
of five (one tin reagent was reacted simultaneously in
different flasks with all five partners). After the reaction,
each mixture was evaporated to remove some of the
solvent, and a three-phase extraction was conducted as
above. Evaporation of the FC-72 phase provided the
To compare the reactivities between fluorous tin
reactants and normal tin reactants, we conducted a series
of simple competition experiments, as summarized in
Table 5. The new fluorous tin reactants 1e (p-meth-
ylphenyl) and 1f (p-ethylphenyl) were prepared by the
standard route (see eq 1). Each of these was then coupled
in competition with the corresponding tributyltin deriva-
tive 6a or 6b under a fixed protocol where 1 equiv of
p-bromoacetophenone 2a was reacted with 2 equiv each
of a pair of tin reagents in the presence and absence of
lithium chloride.
2
1
byproduct fluorous tin chloride 5b (80-90% yield),
In the presence of lithium chloride, the fluorous
reactants 1e and 1f were more reactive than 6a or 6b by
a factor of about 2 (Table 5, entries 1 and 3). As expected,
the reactions were fast (only 1 h) and high yielding. In
contrast, the normal trialkyltin reagents won out by a
factor of about 2 when reactions were conducted in the
absence of lithium chloride (Table 5, entries 2 and 4).
These reactions were much slower; indeed, the coupling
of 1e with 6b appears to stop due to catalyst deactivation.
This reaction was terminated after 20 h, at which time
about 25% of p-bromoacetophenone still remained. These
results show that the coupling of the fluorous tin reac-
tants is promoted by LiCl, as anticipated from the results
in Table 1.
which was routinely recycled (see above). Evaporation
of the organic phase provided a crude organic product
that was further purified by preparative TLC to provide
the major cross-coupled biaryl or diarylmethane 3 along
with small amounts of the symmetrical biaryl 4 derived
from the tin reactant 1. The results of the 15 most
successful couplings are shown in Table 4.
Tin reactants 1a -c gave very clean crude products
with all partners. Isolated yields of 3 were generally high
(
1
>80%), except for a few cases with the furyl tin reactant
c where the products are somewhat volatile and un-
2
2
stable. For 1a ,b the isolated yields of the symmetrical
biaryl 4 (biphenyl from 1a and 4-methoxybiphenyl from
1
b) ranged from 5 to 10%. We believe that comparable
In the two experiments containing lithium chloride in
Table 5, we noticed resonances for a small impurity that
was not present in the experiments without lithium
chloride. Especially characteristic was a multiplet of
about δ 3.0 that we attributed to the benzylic methylene
protons of 9a , the product of cross-coupling between 2a
and a fluoroalkyl group of 1e or 1f (eq 5). The analogous
amounts of bifuryl were formed from 1c, but isolated
yields (not shown) were much lower because this com-
pound is volatile and unstable.
The five crude organic products from the pyridyl tin
reactant 1d were not very clean, so these reaction
mixtures were not fully purified. Significant amounts
of cross-coupled products (estimated 25-50%) were pro-
(
21) Most of the residual 10-20% remained in the organic phase. If
(23) See, Gronowitz, S.; Bj o´ rk, P.; Malm, J .; H o´ rnfeldt, A.-B. J .
Organomet. Chem. 1993, 460, 127. The yield problem can be solved
by adding copper salts. This technique is likewise successful in the
microwave variant of the fluorous Stille coupling (ref 12b).
desired, this can be removed by washing with FC-72, as shown in the
preparative procedure.
(
22) Pelter, A.; Rowlands, M.; Clements, G. Synthesis 1987, 51.

8
346 J . Org. Chem., Vol. 62, No. 24, 1997
Hoshino et al.
product of butyl transfer is a well-known side product in
the Stille coupling.
Ta ble 6. Yield s of Cr oss-Cou p led P r od u cts 10 F or m ed
fr om F lu or ou s Tin Rea cta n ts 1e a n d Su bstr a tes 2 w ith
Cu Ia
1
4,24
(5)
substrate 2
time (h)
10 (% yield)b,c
4e (% yield)d
4
4
4
-MeCOC6H5Br
-NO2C6H4Br
-NO2C6H4OTf
3
6
6
1
2
98
98
91
95
72
6
11
C6H5I
C6H5CH2Br
An attempt to prepare an authentic sample of 9a by
reaction of (C CH Sn with 2a was not successful,
16
6
F
13CH
2
2 4
)
a
Reactions were performed in sealed tubes under argon, 0.24
nor was it possible to separate this product from the
major cross-coupled product. However, GCMS analysis
of preparative reactions provided clear evidence for the
structure of 9a . Coupling of 1e and 1f with 2a provided
different major products but the same side product in
M 1e, 0.2 M 2, 4 mM PdCl2(PPh3)2, 0.6 M LiCl, 16 mM CuI, DMF/
THF (1:1). Isolated yields based on 2. No fluoroalkyl transfer
was observed. Isolated yields based on 1e.
b
c
d
Ta ble 7. Effect of Ad d itives on th e Cou p lin g of F lu or ou s
Tin Rea cta n t 1a a n d Ar yl Ha lid es 2a a
1
ratios of 9/1 and 15/1, respectively, according to H NMR
analysis. A GCMS analysis of the minor peak showed a
molecular ion peak for 9a at m/e 466 with a large
+
fragmentation peak of m/e 451 (M - Me). In contrast
to the failure to prepare pure 9a , we were able to prepare
and characterize the related product p-NO
CH 13 9b by cross coupling of (C CH
p-NO Br, although the yield was poor (15%, see the
2
C
6
H
4
CH
2
-
2
C
6
F
6
F
13CH
2
2 4
) Sn with
2
6 4
C H
1
Experimental Section). Reviews of the crude H NMR
spectra for some of the other products in Table 4 also
showed resonances consistent with the formation of this
side product in low yield (0-10%).
Farina and Liebeskind21 have shown that CuI can be
used to suppress butyl transfer in normal Stille couplings
with aryltributylstannanes. We therefore conducted a
series of couplings of 1e with several halides as shown
in Table 6. Indeed, the production of fluoroalkyl-coupled
products was suppressed, and yields were generally good.
The only exception was with p-methoxybromobenzene
reaction
time (h)
% yield
% yield
of 4a
additiveb
of 3a c
d
entry
1
2
3
4
5
6
7
8
9
LiCl (3 equiv)
no additive
10
46
46
89
77
77
84
72
86
80
86
90
6
7
6
10
12
13
5
e
e
LiCl (0.4 equiv)
LiCl (1.2 equiv)
LiBr (3 equiv)
LiI (3 equiv)
Bu4NCl (3 equiv)
Bu4NI (3 equiv)
LiCl4O
24
10
46
e
(
(
not shown). This reaction appeared to stop after a 1 h
Pd black observed), and a 1/1 mixture of product and
10
46
e
4
8
46
starting material was obtained. As before, LiCl was
essential; a control reaction in its absence rapidly depos-
ited palladium black and the starting materials were not
consumed.
a
Reactions were performed in sealed tubes at 80 °C under N2,
b
0.24 M 1a , 0.20 M 2a , 4 mM PdCl2(PPh3)2. Equiv was based on
c
d
2
a . Isolated yield basded on 2a . Isolated yield based on 1a .
Even after 46 h, some amount of 2a remained.
e
With normal trialkyltin reactants, LiCl generally does
not affect the couplings when the partners are aryl
halides, but it is necessary to induce couplings when
partners are triflates. In the later case, the chloride ion
is thought to convert the organopalladium triflate com-
plex to the more reactive organopalladium chloride
reactions were run under the standard conditions; only
the nature and amount of additive were varied.
At 80 °C in a sealed tube, the coupling in the presence
of 3 equiv of LiCl was complete (both 1a and 2a were
consumed according to TLC analysis) within 10 h (Table
complex before transmetalation.1
4l,n,25
To provide more
7
, entry 1). On the other hand, the coupling in absence
insight into the unusual observations with the fluorous
tin reagents, we studied effects of LiCl and other addi-
tives on the coupling between fluorous tin reactant 1a
and aryl halide 2a in detail as shown in Table 7. All
of LiCl proceeded very slowly and was unfinished even
after 46 h (Table 7, entry 2). The addition of 0.5 equiv
of LiCl promoted the coupling at first (as judged by TLC),
but the effect tapered off and in the end the reaction
required almost as much time as that without LiCl (Table
7, entry 3). These observations are consistent with the
(24) Farina, V.; Krishnan, B.; Marshall, D. R.; Roth, G. P. J . Org.
Chem. 1993, 58, 5434.

Stille Couplings with Fluorous Tin Reactants
J . Org. Chem., Vol. 62, No. 24, 1997 8347
isolation of the tin chloride 5b from all the reactions with
bromides and triflates. As the reaction progresses, the
lithium chloride is consumed on a stoichiometric basis.
With the use of 1.2 equiv of LiCl, the reaction rate was
not as high as with the use of 3 equiv of LiCl, but the
coupling proceeded to completion in 24 h and provided a
good yield of 3a (Table 7, entry 4).
disposal and recycling are lesser concerns, but there is a
premium placed on simple and effective separations.
Assuming that simple two- and three-phase extractions
could be automated, it would be possible to use excess
fluorous reactants or reagents and to separate all the
organic products both from the excess reagents them-
selves and from any byproducts derived from their
reactions (provided that the byproducts retain the fluo-
rous group).
The replacement of LiCl with other salts was studied.
LiBr (Table 7, entry 5) and Bu
enhanced the reaction rate at a level roughly comparable
to LiCl, but LiI (Table 7, entry 6), Bu NI (Table 7, entry
), and LiClO (Table 7, entry 9) had modest rate
4
NCl (Table 7, entry 7)
4
Exp er im en ta l Section
8
4
enhancement at best. All additives in Table 7 except
LiBr eventually gave the desired product 3a in more than
Gen er a l Meth od s. Reactions were carried out under
nitrogen using dried glassware. Anhydrous DMF, NMP, BTF,
and dioxane were obtained from Aldrich. THF and diethyl
ether were obtained by distillation from sodium/benzophenone
8
0% yield. The reaction with LiBr was not as clean as
the others, and unidentified byproducts were formed. The
use of Bu NCl and Bu NI caused a problem in the three-
1
19
prior to use. FC-72 was obtained from 3M. In the Sn NMR
spectra, chemical shifts are relative to tetramethyltin.
Tr is(3,3,4,4,5,5,6,6,7,7,8,8,8-Tr id eca flu or ooctyl)p h en yl-
4
4
phase extraction because the first dichloromethane phase
contained all the tetrabutylammonium salts and a con-
siderable amount (about 70%) of fluorous tin halides.
Considering the results in every respect (reaction time,
yield of desired product 3a , and workup), LiCl was the
best of the additives that were studied here.
It is not currently clear why the chloride ion improves
the Stille couplings with fluorous tin reactants. One
possibility is that an interaction between the chloride ion
and the fluoroalkyltin reagent helps to promote trans-
metalation. However, it is also clear from our results
that the lithium chloride helps preserve the active
catalyst, so it is possible that there is no specific interac-
tion of the chloride ion with the fluorous tin reagent.
However, if the chloride ion is operating only with the
catalyst, it is not clear why the relative reactivity of
tributyltin and fluoroalkyltin reagents is changed by the
presence or absence of LiCl.
11
t in (Tr is[2-(P er flu or oh exyl)et h yl]p h en ylt in ) (1a ). To
the Grignard reagent, prepared from 2-(perfluorohexyl)-1-
iodoethane (100.0 g, 211 mmol) and magnesium (6.53 g, 269
mmol) in dry ether (150 mL), was added phenyltin trichloride
(
15.9 g, 52.7 mmol) dissolved in dry benzene (100 mL). After
being refluxed for 4 h, the reaction was stirred for 16 h at 25
C. The reaction mixture was quenched with NH Cl solution,
and the organic phase was washed with 5% Na solution
and deionized water and then dried over anhydrous MgSO
°
4
2
2 3
S O
4
.
The mixture was concentrated under reduced pressure. After
removal of the major byproduct, 1,4-bis(perfluorohexyl)butane,
by vacuum distillation (87-92 °C, 0.2 mmHg), the resulting
residue was purified by column chromatography on neutral
alumina with hexane to give pure compound 1a (56.1 g, 86%)
as a colorless oil: 1H NMR (CDCl
) δ 7.41 (s, 5 H), 2.31 (m, 6
119
3
2
119
H), 1.31 (t, J ) 8.3 Hz, J ( Sn-H) ) 53.4 Hz, 6 H); Sn
NMR (CDCl ) -11.7 ppm; IR (thin film) 3100, 2950, 1238,
190, 1144, 655 cm ; MS m/z 1161 (M - Ph), 891 (M
CH CH C F13).
3
-1
+
+
1
-
2
2
6
Br om oTr is(3,3,4,4,5,5,6,6,7,7,8,8,8-Tr id e ca flu or ooc-
tyl)tin (Br om otr is[2-(p er flu or oh exyl)eth yl]tin ) (5a ). Bro-
mine (5.83 g, 36.5 mmol) in ether (10 mL) was added dropwise
to an ice-cold solution of 1a (43.0 g, 34.8 mmol) in dry ether
Con clu sion s
Although the scope and generality of fluorous Stille
couplings will only be revealed by more extensive studies,
the results reported herein provide an encouraging
outlook. Despite the alteration of solubility that attends
the use of the fluorous tin reagents, Stille couplings still
occur smoothly in a number of organic solvents that can
solubilize both organic and fluorous compounds. While
the beneficial effect of lithium chloride was unantici-
pated, the reaction conditions and results with the
fluorous aryltin reactions otherwise seem to parallel
those of standard alkyltin reactants.
(80 mL). The mixture was warmed to 25 °C over 2 h with
stirring. Removal of the ether, bromobenzene, and excess
bromine by evaporation under reduced pressure resulted in
an orange oil. Purification by vacuum distillation (150-152
°C, 0.5 mmHg) yielded compound 5a (42.4 g, 98%) as a
1
colorless oil: H NMR (CDCl
3
) δ 2.42 (m, 6 H), 1.56 (t, J ) 8.3
2
119
119
Hz, J ( Sn-H) ) 53.4 Hz, 6 H); Sn NMR (hexane) 259.2
-
1
ppm (m); IR (thin film) 3600, 1250, 1227, 1145, 534 cm ; MS
+
+
2 2 6
m/z 1161 (M - Br), 893 (M - CH CH C F13).
Tr is(3,3,4,4,5,5,6,6,7,7,8,8,8-Tr idecaflu or ooctyl)(4′-m eth -
oxyp h en yl)t in (Tr is[2-(P er flu or oh exyl)et h yl](4′-m et h -
oxyp h en yl)tin ) (1b). To the Grignard reagent prepared from
In the purification stage, the fluorous reagents exhib-
ited the anticipated advantages. From the standpoint
of traditional organic synthesis, the simple extractive
separation not only allows the convenient separation of
the fluorous tin products from the organic products but
also facilitates the recycling of the resulting fluorous tin
chloride to the same or another fluorous tin reagent.
Indeed, we routinely recycle our fluorous reagents, and
some of the couplings reported in this paper were
doubtless conducted with the same individual tin group
or with a tin group that had been used previously in a
different kind of reaction. The material balance with
respect to the fluorous component is consistently excel-
lent, often nearly quantitative, adding an additional
attraction to the recycling strategy. We are reporting
separately on additional extensions of the fluorous tin
4
-bromoanisole (681 mg, 3.64 mmol) and magnesium (102 mg,
4.20 mmol) in dry ether (20 mL) was added a solution of 5a
(3.47 g, 2.80 mmol) in dry ether (10 mL). After being refluxed
for 1 h, the reaction was allowed to stand 16 h at 25 °C. The
reaction mixture was quenched with NH
diluted with ether, and the organic phase was washed with
deionized water and then dried over anhydrous MgSO . The
4
Cl solution and
4
mixture was concentrated under reduced pressure. Purifica-
tion by vacuum distillation (166 °C, 0.25 mmHg) and then
column chromatography on neutral alumina with hexane
1
yielded pure compound 1b (5.20 g, 74%) as a colorless oil:
NMR (CDCl
H), 3.82 (s, 3 H), 2.29 (m, 6 H), 1.27 (t, J ) 8.3 Hz, J (
H
3
) δ 7.30 (d, J ) 8.3 Hz, 2 H), 6.98 (d, J ) 8.3 Hz,
2 119
2
-
1
19
Sn-H) ) 54.0 Hz, 6 H); Sn NMR (CDCl
3
) -26.3 ppm; IR
(
thin film) 1500, 1375, 1240, 1205, 1145, 1065, 745, 700
-1
+
+
+
cm ; MS m/z 1267 (M ), 1161 (M - PhOMe), 921 (M
CH CH
-
2
2 6
C F13).
1
4
and silicon chemistry.
Tr is(3,3,4,4,5,5,6,6,7,7,8,8,8-Tr idecaflu or ooctyl)(2′-fu r yl)-
tin (Tr is[2-(P er flu or oh exyl)eth yl](2′-fu r yl)tin ) (1c). To
the solution of furan (667 mg, 9.80 mmol) in dry THF (25 mL)
The chemistry also exhibits favorable features for
small-scale parallel synthesis. In this kind of chemistry,

8
348 J . Org. Chem., Vol. 62, No. 24, 1997
Hoshino et al.
at 0 °C was added 1.5 M LDA/cyclohexane (6.53 mL, 9.80
mmol). After 1 h at 0 °C, the resulting mixture was treated
with a solution of 5a (8.67 g, 7.00 mmol) in dry THF (15 mL).
The reaction mixture was warmed to 25 °C over 1 h and then
stirred 16 h at 25 °C. The reaction mixture was quenched with
evaporated under reduced pressure and the residue was
partitioned between water (10 mL), dichloromethane (15
mL), and FC-72 (15 mL). The three phases were separated,
and the dichloromethane phase was dried over anhydrous
MgSO
4
. Evaporation of the FC-72 phase provided chloro-
NH
was washed with deionized water and then dried over anhy-
drous MgSO . The mixture was concentrated under reduced
pressure. Column chromatography on neutral alumina with
hexane yielded pure compound 1c (2.44 g, 28%) as a colorless
4
Cl solution and diluted with ether, and the organic phase
tris(3,3,4,4,5,5,6,6,7,7,8,8,8-tridecafluorooctyl)tin (5b) (chloro-
tris[2-(perfluorohexyl)ethyl]tin), which was routinely reused.
Evaporation of the dichloromethane phase provided the crude
organic product, which was further purified by preparative
TLC (silica gel) to provide the major cross-coupled product 3
and a small amount of the symmetrical biaryl 4 derived from
the tin reactant.
4
1
oil: H NMR (CDCl
3
) δ 7.76 (s, 1 H), 6.63 (s, 1 H), 6.47 (s, 1
2
119
H), 2.35 (m, 6 H), 1.29 (t, J ) 9.7 Hz, J ( Sn-H) ) 56.8 Hz,
H); Sn NMR (CDCl
240, 1205, 1145, 1065, 745, 700 cm ; MS m/z 1228 (M ), 1161
- furyl), 881 (M - CH
Tr is(3,3,4,4,5,5,6,6,7,7,8,8,8-Tr id eca flu or ooct yl)(2′-p y-
r id yl)t in (Tr is[2-(P er flu or oh exyl)et h yl](2′-p yr id yl)t in )
1
19
6
1
3
) -49.3 ppm; IR (thin film) 1445, 1355,
Ch lor ot r is(3,3,4,4,5,5,6,6,7,7,8,8,8-t r id eca flu or ooct yl-
-1
+
1
)t in (5b) (Ch lor ot r is[2-(p er flu or oh exyl)et h yl)t in ]:
H
+
+
2
119
(M
2
CH
2
C
6
F
13).
NMR (CDCl
H) ) 47.6 Hz, 6 H); 1 Sn NMR (CDCl
1450, 1360, 1240, 1205, 1145, 1065, 735, 700 cm ; MS m/z
3
) δ 2.46 (m, 6 H), 1.53 (t, J ) 7.9 Hz, J ( Sn-
19
3
) 273 ppm; IR (thin film)
-
1
+
+
(
1d ). To the Grignard reagent prepared from 2-bromopyridine
2 2 6
1161 (M - Cl), 849 (M - CH CH C F13).
(822 mg, 5.20 mmol) and magnesium (146 mg, 6.20 mmol) in
Exa m p le of a P r ep a r a tive Stille Cou p lin g (Eq 4). A
tube under nitrogen was charged with tin reactant 1a (2.97 g,
2.40 mmol), 1-bromo-4-nitrobenzene (2c) (404.0 mg, 2.00
mmol), lithium chloride (254.3 mg, 6.00 mmol), dichlorobis-
(triphenylphosphine)palladium(II) (28.1 mg, 0.04 mmol), dry
DMF (5 mL), and dry THF (5 mL). The tube was sealed, the
mixture was heated to 80 °C, and a homogeneous solution
resulted. The mixture was stirred at 80 °C for 22 h. After
azeotropic evaporation under reduced pressure with toluene
at 75 °C (to remove some of the DMF), the resulting residue
was partitioned between water (40 mL), dichloromethane (60
mL), and FC-72 (40 mL). The three phases were separated.
Evaporation of the FC-72 phase provided 2.31 g (80.6% from
1a ) of tin chloride 5b as a colorless oil. The dichloromethane
phase was washed three more times with water (40 mL) and
FC-72 (40 mL) to remove DMF and fluorous products (mostly
chlorotris[2-(perfluorohexyl)ethyl]tin). Evaporation of all the
FC-72 phase (including the first phase) provided 2.85 g (99.4%
from 1a ) of tin chloride 5b. The final dichloromethane phase
dry ether (30 mL) was added a solution of 5a (2.48 g, 2.00
mmol) in dry ether (5 mL). After refluxing for 1 min, the
reaction was allowed to stand for 16 h at 25 °C. The reaction
mixture was quenched with NH
ether, and the organic phase was washed with deionized water
and then dried over anhydrous MgSO The solvent was
4
Cl solution and diluted with
4
.
evaporated to dryness, and the resulting residure was dis-
solved in toluene and FC-72. The two phases were separated.
The FC-72 phase was washed with toluene and concentrated
to afford pure compound 1d (2.18 g, 88%) as a pale yellow oil:
1
H NMR (CDCl ) δ 8.71 (d, J ) 4.3 Hz, 1 H), 7.58 (t, J ) 7.7
3
Hz, 1 H), 7.36 (d, J ) 7.3 Hz, 1 H), 8.20 (m, 1 H), 2.29 (m, 6
2
119
119
H), 1.34 (t, J ) 8.2 Hz, J ( Sn-H) ) 54.3 Hz, 6 H); Sn
NMR (CDCl ) -61.4 ppm; IR (thin film) 1570, 1450, 1360,
240, 1205, 1145, 1060, 735, 700 cm ; MS m/z 1238 (M ), 1161
3
-1
+
1
(
+
+
M
- pyridyl), 892 (M - CH
2 2 6
CH C F13).
Tr is(3,3,4,4,5,5,6,6,7,7,8,8,8-Tr id e ca flu or ooct yl)(4′-
m eth ylp h en yl)tin (1e). To the Grignard reagent prepared
from 1-bromo-4-methylbenzene (533 mg, 3.12 mmol) and
magnesium (89 mg, 3.56 mmol) in dry ether (16 mL) was added
a solution of 5a (2.97 g, 2.39 mmol) in dry ether (8 mL). After
being refluxed for 1 h, the reaction was allowed to stir
overnight at 25 °C. The reaction mixture was quenched with
was dried over anhydrous MgSO
4
and evaporated to give
yellow crystals free of fluorous reactant 1a and fluorous tin
halides. The crude organic product was further purified by
column chromathography on silica gel to provide the cross-
coupled product 4-nitrobiphenyl (3b) (337 mg, 85%) as yellow
crystals and the homocoupled product biphenyl (4a ) (17 mg,
5%) as white crystals.
Exa m p le of a Recycle of th e Tin Rea cta n ts (Eq 4). The
tin chloride 5b (2.85 g) isolated by evaporation of FC-72 phase
after the above Stille coupling was treated with a 3 M solution
of phenylmagnesium bromide (1.04 mL, 3.12 mmol in ether)
in dry ether (25 mL) under stirring at 25 °C for 6 h. The
NH
was washed with deionized water and then dried over anhy-
drous MgSO . The mixture was concentrated under reduced
4
Cl solution and diluted with ether, and the organic phase
4
pressure. The resulting residue was purified by column
chromatography on silica gel with hexane to give pure 1e
1
(
2.71 g, 91%) as a colorless oil: H NMR (CDCl
3
) δ 7.3-7.19
(
m, 4 H), 2.4-2.2 (m, 6 H), 2.39 (s, Me), 1.28 (t, J ) 8.3 Hz,
2
119 119
J ( Sn-H) ) 53.4 Hz, 6 H); Sn NMR (BTF/C
6
D
6
) -36.3
reaction mixture was quenched with NH
diluted with ether, and the organic phase was washed with
deionized water and dried over anhydrous MgSO The
4
Cl solution and
-1
ppm; IR (thin film) 2930, 2430, 1230, 1060 cm ; MS m/z 1161
+
+
M
- C
Tr is(3,3,4,4,5,5,6,6,7,7,8,8,8-Tr id eca flu or ooctyl)(4′-eth -
ylp h en yl)tin (1f). To the Grignard reagent prepared from
-bromo-4-ethylbenzene (682 mg, 3.69 mmol) and magnesium
103 mg, 4.41 mmol) in dry ether (20 mL) was added a solution
6
H
4
Me), 905 (M - CH
2
CH
2
C
6
F
13).
4
.
mixture was concentrated under reduced pressure. Column
chromathography on neutral alumina with hexane yielded
pure compound 1a (2.85 g, 96% overall from 1a in the
preceding section) as a colorless oil.
1
(
of 5a (1.80 g, 1.45 mmol) in dry ether (8 mL). After being
refluxed for 1 h, the reaction was allowed to stir overnight at
Exa m p le of a Stille Cou p lin g w ith Cu I.21 A tube under
argon was charged with 4-bromoacetophenone (2a ) (39.8 mg,
0.2 mmol), lithium chloride (25.4 mg, 0.6 mmol), dichlorobis-
(triphenylphosphine)palladium(II) (2.9 mg, 0.004 mmol), tin
reactant 1d (300.3 mg, 0.24 mmol), copper(I)iodide (3.0 mg,
0.016 mmol), dry DMF (0.5 mL), and dry THF (0.5 mL). The
tube was sealed, and the mixture was heated at 80 °C for 3 h.
The red-brown solution was cooled to room temperature, and
the solvent was evaporated. The resulting residue was
partitioned between water, dichloromethane, and FC-72. The
three phases were separated, and the dichloromethane phase
was washed three more times with water and FC-72. The final
2
4
5 °C. The reaction mixture was quenched with NH Cl
solution and diluted with ether, and the organic phase was
washed with deionized water and then dried over anhydrous
MgSO . The mixture was concentrated under reduced pres-
4
sure. The resulting residue was purified by column chroma-
tography on silica gel with hexane to give the pure compound
1
1
7
1
2
1
f (1.66 g, 91%) as a colorless oil: H NMR (CDCl
3
) δ 7.35-
.25 (m, 4 H), 2.71 (q, J ) 7.4 Hz, 2 H), 2.32 (m, 6 H), 1.35-
1
19
.2 (m, 9 H); Sn NMR (BTF/C
6 6
D ) -34.7 ppm; IR (thin film)
-
1
+
970, 1440, 1235, 1200, 1150, 1060 cm ; MS m/z 1265 (M ),
+
+
161 (M - C
6
H
4
Et), 919 (M - CH
Gen er a l P r oced u r e for th e Stille Cou p lin gs (Ta ble 4).
2
CH
2
C
6
F
13
)
dichloromethane phase was dried over anhydrous MgSO
4
.
After filtration, the solvent was evaporated, and a solid was
obtained. Further purification by flash chromatography (hex-
ane/ether 6:1) over silica gel provided 4-acetyl-4′-methylbi-
phenyl (7) (41.0 mg, 98%) as white crystals.
A tube under nitrogen was charged with tin reactant 1 (0.24
mmol), substrate 2 (0.20 mmol), lithium chloride (25.4 mg, 0.60
mmol), dichlorobis(triphenylphosphine)palladium(II) (2.8 mg,
0
.004 mmol), dry DMF (0.5 mL), and dry THF (0.5 mL). The
1-Nitr o-4-[2-(p er flu or oh exyl)eth yl]ben zen e (9b) was
prepared according to the general procedure for the Stille
couplings. After purification by preparative TLC, pure com-
tube was sealed, the mixture was heated to 80 °C, and a
homogeneous solution resulted. After 22 h, the solvent was

Stille Couplings with Fluorous Tin Reactants
J . Org. Chem., Vol. 62, No. 24, 1997 8349
pound 9b (14.1 mg, 15%) was given as pale yellow crystals:
2-(P h en ylm eth yl)fu r a n (p r od u ct betw een 1c a n d ben -
zyl br om id e): CAS registry no. 37542-92-0 provided by the
author.
Bip h en yl (4a ): CAS registry no. 92-52-4 provided by the
author.
1
mp 39 °C; H NMR (CDCl
J ) 8.6 Hz, 2 H), 3.05(m, 2 H), 2.42 (m, 2 H); IR (thin film)
3
) δ 8.20 (d, J ) 8.6 Hz, 2 H), 7.40 (t,
-
1
3
4
020, 2855, 1525, 1350, 1240, 1215, 1145, 765 cm ; MS m/z
+
+
69 (M ), 423 (M - NO
All of the isolated products 3, 4, and 7 are known.
-Acetylbip h en yl (3a ): CAS registry no. 92-91-1 provided
by the author.
2
).
4
,4’-Dim eth oxybip h en yl (h om ocou p led p r od u ct fr om
b): CAS registry no. 2132-80-1 provided by the author.
-Acetyl-4′-m eth ylbip h en yl (3c): CAS registry no. 7722-
4-1 provided by the author.
′-Eth yl-4-a cetylbip h en yl (3d ): CAS registry no. 5730-
2-7 provided by the author.
′-Met h ylb ip h en yl (p r od u ct b et w een 1e a n d iod o-
4
1
8
9
4
4
-Nitr obip h en yl (3b): CAS registry no. 92-93-3 provided
by the author.
4
Dip h en ylm et h a n e (p r od u ct b et w een 1a a n d b en zyl
br om id e): CAS registry no. 101-81-5 provided by the author.
4
4
-Met h oxyb ip h en yl (p r od u ct b et w een 1b a n d iod o-
ben zen e): CAS registry no. 613-37-6 provided by the author.
-Acetyl-4’-m eth oxybip h en yl (p r od u ct betw een 1b a n d
’-br om oa cetop h en on e): CAS registry no. 13021-18-6 pro-
vided by the author.
-Meth oxy-4’-n itr obip h en yl (p r od u ct betw een 1b a n d
ben zen e): CAS registry no. 188701-34-0 provided by the
author.
4
4
-Meth yl-4′-n itr obip h en yl (p r od u ct betw een 1e a n d
4
eit h er 1-b r om o-4-n it r ob en zen e or 4-n it r op h en yl t r i-
flu or om eth a n esu lfon a te): CAS registry no. 2143-88-6 pro-
vided by the author.
4
eit h er 1-b r om o-4-n it r ob en zen e or 4-n it r op h en yl t r i-
flu or om eth a n esu lfon a te): CAS registry no. 2143-90-0 pro-
vided by the author.
4
-Meth oxy-4′-m eth ylbip h en yl (p r od u ct betw een 1e
a n d 1-br om o-4-m eth oxyben zen e): CAS registry no. 53040-
9
6
2
2-9 provided by the author.
-(Meth ylp h en yl)p h en ylm eth yl (4b): CAS registry no.
13-33-2 provided by the author.
-Acetyl-2’-m eth ylbip h en yl (9): CAS registry no. 76650-
1
-Meth oxy-4-(ph en ym eth yl)ben zen e (pr odu ct between
b a n d ben zyl br om id e): CAS registry no. 13021-18-6 pro-
vided by the author.
-P h en ylfu r a n (p r od u ct betw een 1c a n d iod oben -
zen e): CAS registry no. 17113-33-6 provided by the author.
-(4-Acet ylp h en yl)fu r a n (p r od u ct b et w een 1c a n d
’-br om oa cetop h en on e): CAS registry no. 35216-08-01 pro-
vided by the author.
-(4-Nitr op h en yl)fu r a n (p r od u ct betw een 1c a n d ei-
4
1
4
2
9-8 provided by the author.
2
4
Ack n ow led gm en t. We thank the National Insti-
tutes of Health and the Kao Corporation for support of
this work. Peter Degenkolb thanks the Technical
University of Berlin for an Erwin Stephan Fellowship.
2
th er 1-br om o-4-n itr oben zen e or 4-Nitr op h en yl tr iflu or o-
m eth a n esu lfon a te): CAS registry no. 28123-72-0 provided
by the author.
J O9709413