Recyclable palladium(0)-catalysed silylstannation of terminal alkynes in ionic liquids

Article abstract of DOI:10.1016/j.molcata.2003.09.032

The palladium(0)-catalysed addition of silylstannanes Bu ₃SnSiMe₃ and Bu₃SnSiMe₂Ph to terminal alkynes has been shown to proceed in the ionic liquids 1-n-butyl-3-methylimidazolium hexafluorophosphate ([bmim]PF₆) and 1-n-butyl-3-methylimidazolium tetrafluoroborate ([bmim]BF₄). These reactions generally proceed in excellent yields to give 1-trialkylsilyl-2- tributylstannyl-1Z-alkenes regio- and stereoselectively as a single product with reaction times comparable to those reported for reactions performed in tetrahydrofuran (THF). The reaction has been shown to tolerate remote functionality on the alkyne. Each of the ionic liquids containing the immobilised palladium catalyst has been recycled up to 10 times without loss of activity, allowing extensive reuse of the expensive solvent/catalyst system. A catalyst activation stage involving mild heating of the palladium catalyst in the ionic liquid/diethyl ether solvent system decreased reaction time in the first cycle once all reagents were added, indicating the formation of a catalytically active palladium species other than tetrakis(triphenylphosphine) palladium(0).

Full text of DOI:10.1016/j.molcata.2003.09.032

Journal of Molecular Catalysis A: Chemical 214 (2004) 33–44
Recyclable palladium(0)-catalysed silylstannation of
terminal alkynes in ionic liquids
Ivan Hemeon, Robert D. Singer∗
Department of Chemistry, Saint Mary’s University, Halifax, Nova Scotia, Canada B3H 3C3
Accepted 11 September 2003
Abstract
The palladium(0)-catalysed addition of silylstannanes Bu3SnSiMe3 and Bu3SnSiMe2Ph to terminal alkynes has been shown to proceed
in the ionic liquids 1-n-butyl-3-methylimidazolium hexafluorophosphate ([bmim]PF6) and 1-n-butyl-3-methylimidazolium tetrafluoroborate
([bmim]BF4). These reactions generally proceed in excellent yields to give 1-trialkylsilyl-2-tributylstannyl-1Z-alkenes regio- and stereoselec-
tively as a single product with reaction times comparable to those reported for reactions performed in tetrahydrofuran (THF). The reaction has
been shown to tolerate remote functionality on the alkyne. Each of the ionic liquids containing the immobilised palladium catalyst has been
recycled up to 10 times without loss of activity, allowing extensive reuse of the expensive solvent/catalyst system. A catalyst activation stage
involving mild heating of the palladium catalyst in the ionic liquid/diethyl ether solvent system decreased reaction time in the first cycle once all
reagents were added, indicating the formation of a catalytically active palladium species other than tetrakis(triphenylphosphine)palladium(0).
© 2004 Elsevier B.V. All rights reserved.
Keywords: Ionic liquids; Palladium; Silylstannanes; Alkynes; Recyclable
1. Introduction
and rate enhancement. While these reactions have been
both catalytic and non-catalytic [4–7], ionic liquids are be-
ing used increasingly in transition metal-catalysed reactions
as they often allow easy recycling of the catalyst and of-
fer greater catalyst stability [8–11]. Several metal catalysts
have been immobilised in ionic liquids, using the ionic liq-
uid both as a solvent and as a support for the catalyst. The
polar, ionic nature of the solvent allows easy dissolution
of metal complexes and its non-co-ordinating nature allows
the catalyst to remain active. While the exact nature of the
interactions between ionic liquids and their solutes remains
unclear, ionic liquids are capable of supporting catalysts and
retaining them after reaction products have been extracted
from the solvent. This allows catalyst to be reused due to
the non-volatile nature of ionic liquids.
Trialkyl(trialkylstannyl)silanes are versatile organometal-
lic reagents that contain a silicon–tin bond. These com-
pounds are easily synthesised by generating a trialkyltin
anion through various methods (e.g. deprotonation of tri-
alkyltin hydride with lithium diisopropylamide, LDA) and
quenching this with a trialkylsilyl chloride [12]. As well,
some are commercially available. Trialkyl(trialkylstannyl)-
silanes are particularly interesting in that there is poten-
tial to incorporate both the trialkylstannyl and trialkylsilyl
Ionic liquids have recently attracted much attention from
many areas of chemistry. The increasing level of interest
in these compounds arises from the fact that they are ionic
compounds that are liquids at or near room temperature.
As such, they have many unique properties that are cur-
rently being utilised in a number of varied applications.
There are many different types of known ionic liquids based
on various cations and anions. The most commonly used
ionic liquid cations are based on N,Nꢀ-dialkylimidazolium
species. These liquids have remarkably different properties
that depend on their associated anion [1]. Ionic liquids typ-
ically have wide liquid temperature ranges, are immiscible
with a number of organic solvents, and exhibit no measur-
able vapour pressure. Their inertness toward many reagents
makes them ideal for use as a new class of polar [2,3],
non-co-ordinating, ionic solvents.
Many different reactions have been performed in ionic
liquids, often affording greater yields, product selectivity
∗
Corresponding author. Tel.: +1-902-496-8189;
fax: +1-902-420-5261.
E-mail address: robert.singer@smu.ca (R.D. Singer).
1381-1169/$ – see front matter © 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcata.2003.09.032

34
I. Hemeon, R.D. Singer / Journal of Molecular Catalysis A: Chemical 214 (2004) 33–44
2. Experimental
R
H
Pd cat.
R C CH
R'3SnSiR"3
R'3Sn
SiR"3
2.1. General details
Fig. 1. Palladium-catalysed silylstannation of terminal alkynes.
Synthesis and reactions of trialkyl(trialkylstannyl)silanes
were performed in oven-dried glassware that was cooled
under vacuum and released to a nitrogen atmosphere. THF
and Et2O that were used in the synthesis and reactions
of trialkyl(trialkylstannyl)silanes were dried and distilled
over potassium/benzophenone immediately prior to use.
Dichloromethane was dried and distilled over calcium hy-
dride. Pyridine and diisopropylamine were dried over and
distilled onto potassium hydroxide pellets. Thionyl chlo-
ride was distilled from quinoline. Ionic liquids [20,21],
Bu3SnSiMe2Ph [12,15], 6-chlorohex-1-yne [22], and 2-
(5-hexynyloxy)tetrahydro-2H-pyran [23] were synthesised
and purified before use according to literature procedures.
All other reagents were used as received from Aldrich.
Column chromatography was performed using Kieselgel
60, 230–400 mesh silica gel. Thin-layer chromatography
was performed using aluminium-backed plates coated with
Kieselgel 60 F254; plates were visualised under ultraviolet
lamp and/or using 5% phosphomolybdic acid in ethanol.
Silylstannation reactions were heated on a Mirak Ther-
molyne temperature controlled hotplate. Low temperatures
for trialkyl(trialkylstannyl)silane syntheses were maintained
by an FTS Systems cooling apparatus. Gas chromatography
(GC)–mass spectrometry was performed using a Varian
3800 gas chromatograph equipped with a flame ionisation
detector and a Varian Saturn 2000 electron impact/ion trap
spectrometer. Nuclear magnetic resonance spectra were
recorded on a Bruker 250 MHz spectrometer at the Atlantic
Regional Magnetic Resonance Centre. These spectra were
recorded using CDCl3 solutions and chemical shifts are
reported in parts per million referenced to tetramethylsilane
unless otherwise stated. The melting point of 1-n-butyl-
3-methylimidazolium chloride ([bmim]Cl) was measured
using a Mettler FP85 differential scanning calorimetry cell
in conjunction with a Mettler FP80 central processing unit
at 2 ◦C/min between 50 and 100 ◦C.
moieties into organic products providing two new reac-
tive centres for further manipulation. Perhaps the most
useful reaction involving trialkyl(trialkylstannyl)silanes is
their palladium-catalysed addition across unsaturated sys-
tems such as terminal alkynes (Fig. 1) [13–17]. These
reactions afford a bifunctional organic dianion equiva-
lent containing both carbon–tin and carbon–silicon bonds.
These functionalised products can then undergo a number
of organic transformations including selective protodesi-
lylation or protodestannylation, tin–lithium exchange via
treatment with alkyllithium reagents, substitution of the tri-
alkyltin and/or trialkylsilyl groups, and palladium-mediated
Stille couplings [13–17]. Drawbacks to the addition of tri-
alkyl(trialkylstannyl)silanes to alkynes include the expense
of the palladium catalyst that is necessary for the reaction
to proceed and the inability to recycle the catalyst.
We have recently reported that the ionic liquid 1-n-butyl-3-
methylimidazolium hexafluorophosphate ([bmim]PF6)
(Fig. 2) can be used to immobilise a palladium(0) catalyst
and act as a co-solvent with diethyl ether in the palladium-
catalysed silylstannation of selected terminal alkynes using
trimethyl(tributylstannyl)silane Bu3SnSiMe3 [18]. This re-
action proceeds in a similar fashion as in tetrahydrofuran
(THF) with excellent yields to afford 1-trimethylsilyl-2-
tributylstannyl-1Z-alkenes as single isomers (Fig. 1) [13,14].
Use of the ionic liquid allowed recycling of the costly pal-
ladium(0) catalyst without loss of activity. This reaction
might be considered a “greener” reaction as it exhibits total
atom economy, incorporating all atoms of the starting ma-
terials into the product with no by-product formation [19].
Ionic liquids have been proposed as ideal reaction media for
green chemistry due to their non-volatile nature and ease
of reuse. The use of ionic liquids as solvents for the silyl-
stannation of alkynes thus increases the overall “greenness”
of the process. Also, the use of a silylstannane expands the
scope of organometallic reagents used in ionic liquids since
to date most organometallic reagents used in ionic liquids
have been part of a transition metal complex.
2.2. Synthesis of ionic liquids
2.2.1. 1-n-Butyl-3-methylimidazolium chloride
We report herein a full account of our investigations into
the palladium(0)-catalysed silylstannation reactions of ter-
minal alkynes in ionic liquids. Reactions between a variety
of terminal alkynes with two silylstannane reagents in two
different ionic liquids have been demonstrated. The robust
nature of this system regarding the recyclability of the im-
mobilised catalyst has also been shown.
1-Methylimidazole (40 ml, 0.50 mol) and 1-chlorobutane
(90 ml, 0.86 mol) were refluxed for 18 h, after which time
excess 1-chlorobutane was decanted off of the newly-formed
viscous tan-coloured lower layer. This lower layer was
washed with ethyl acetate (EtOAc, 3 × 100 ml) then heated
under vacuum at 80 ◦C for 6 h. An off-white solid formed
upon cooling, mp 68.6 ◦C (81.43 g, 0.466 mol, 93%). 1H
3
NMR: δ 0.75 (t, 3H, N–CH2CH2CH2CH3, J = 7.3 Hz),
N
N
3
-X
1.17 (sextet, 2H, N–CH2CH2CH2CH3, J = 7.3 Hz), 1.70
(quintet, 2H, N–CH2CH2CH2CH3, 3J = 7.3 Hz), 3.93
-X = -PF6, -BF4
3
(s, 3H, N–CH3), 4.14 (t, 2H, N–CH2CH2CH2CH3, J =
Fig. 2. Ionic liquids used in this study [bmim]PF6 and [bmim]BF4.
7.3 Hz), 7.41 (br s, 1H, H4/H5), 7.58 (br s, 1H, H4/H5),

I. Hemeon, R.D. Singer / Journal of Molecular Catalysis A: Chemical 214 (2004) 33–44
35
product coming over at 160–165 ◦C as a clear colourless
oil (5.93 g, 13.9 mmol, 70%). 1H NMR: δ 0.54 (s, 6H,
Si(CH3)2), 0.85–0.92 (m, 15H, Sn(CH2CH2CH2CH3)3),
1.23–1.49 (m, 12H, Sn(CH2CH2CH2CH3)3), 7.34–7.51 (m,
5H, ArH). 13C NMR: δ −0.2, 8.3, 13.8, 27.7, 30.3, 128.0,
128.5, 133.7, 141.5. MS (EI) m/z (rel. int. %): 369 (11),
311 (19), 255 (100), 177 (15), 135 (68), 75 (9), 43 (14).
10.36 (s, 1H, H2). 13C NMR: δ 13.3, 19.3, 32.0, 36.4, 49.5,
122.0, 123.7, 137.5.
2.2.2. 1-n-Butyl-3-methylimidazolium hexafluorophosphate
[bmim]Cl (87.95 g, 0.504 mol) was dissolved in water
(150 ml) and cooled in an ice bath. A 60% aqueous solu-
tion of HPF6 (74.5 ml, 0.506 mol) was added dropwise over
20 min as a lower tan-coloured layer formed exothermically.
The mixture was stirred for 24 h, after which time the up-
per aqueous phase was decanted. The lower ionic liquid
phase was washed with water (10 × 100 ml) after which the
washings were pH neutral. After heating under vacuum at
70 ◦C for 6 h the ionic liquid was obtained as a clear tan
oil (96.26 g, 0.339 mol, 67%). 1H NMR (acetone-d6): δ 0.94
(t, 3H, N–CH2CH2CH2CH3, 3J = 7.3 Hz), 1.38 (sextet,
2.4. Synthesis of 6-chlorohex-1-yne
5-Hexyn-1-ol (1.00 ml, 9.1 mmol) was injected into a
flask containing dry Et2O (15 ml) followed by dry pyri-
dine (0.74 ml, 9.1 mmol) and stirred in an ice bath for
30 min. Thionyl chloride (0.80 ml, 11.0 mmol) was injected
dropwise; the flask was fitted with a dry condensor and
the cloudy white mixture heated to reflux under nitrogen
overnight. GC–FID analysis of a reaction aliquot confirmed
complete reaction, at which time the brown solution was
poured into water (50 ml). The mixture was extracted with
Et2O (30 ml). The Et2O layer was washed with water
(2 × 50 ml), dried over anhydrous MgSO4, filtered and con-
centrated to afford a dark oil. This was purified via flash
chromatography (10:1 hexanes/EtOAc) to afford the prod-
uct as a volatile clear light brown oil (0.436 g, 3.7 mmol,
41%). 1H NMR: δ 1.68 (quintet, 2H, H4/H5, 3J = 6.5 Hz),
1.91 (quintet, 2H, H4/H5, 3J= 6.5 Hz), 1.97 (t, 1H, H1,
4J = 2.6 Hz), 2.25 (dt, 2H, H3, 3J = 6.5 Hz, 4J = 2.6 Hz),
3
2H, N–CH2CH2CH2CH3, J = 7.3 Hz), 1.92 (quintet, 2H,
3
N–CH2CH2CH2CH3, J = 7.3 Hz), 4.03 (s, 3H, N–CH3),
3
4.34 (t, 2H, N–CH2CH2CH2CH3, J = 7.3 Hz), 7.67 (br
s, 1H, H4/H5), 7.72 (br s, 1H, H4/H5), 8.91 (s, 1H, H2).
13C NMR (acetone-d6): δ 13.7, 19.9, 32.7, 36.6, 50.2, 123.4,
124.8, 137.3.
2.2.3. 1-n-Butyl-3-methylimidazolium tetrafluoroborate
([bmim]BF4)
[bmim]Cl (30.76 g, 0.176 mol) was dissolved in wa-
ter (150 ml) to which a 48% aqueous solution of HBF4
(23.0 ml, 0.176 mol) was added. The solution stirred for
72 h after which time the water was evaporated under
reduced pressure. Heating under vacuum at 70 ◦C for
6 h afforded the ionic liquid as a clear tan oil (38.87 g,
0.172 mol, 98%). 1H NMR (acetone-d6): δ 0.93 (t, 3H,
N–CH2CH2CH2CH3, 3J = 7.3 Hz), 1.37 (sextet, 2H,
N–CH2CH2CH2CH3, 3J = 7.3 Hz), 1.91 (quintet, 2H,
N–CH2CH2CH2CH3, 3J = 7.32 Hz), 4.03 (s, 3H, N–CH3),
4.34 (t, 2H, N–CH2CH2CH2CH3, 3J = 7.3 Hz), 7.69 (br s,
1H, H4/H5), 7.75 (br s, 1H, H4/H5), 9.01 (s, 1H, H2). 13C
NMR (acetone-d6): δ 13.7, 19.9, 32.8, 36.5, 50.1, 123.4,
124.7, 137.6.
3
3.57 (t, 2H, H6, J = 6.5 Hz). 13C NMR: δ 17.7, 25.6,
31.4, 44.5, 68.9, 83.7. MS (EI) m/z (rel. int. %): 115 (2), 90
(33), 88 (100), 81 (59), 79 (97), 65 (33), 53 (65), 41 (50).
2.5. Synthesis of ( )-2-(5-hexynyloxy)tetrahydro-
2H-pyran
p-Toluenesulphonic acid (0.13 g, 0.68 mmol) was dis-
solved in dry dichloromethane (10 ml). 5-Hexyn-1-ol
(1.50 ml, 13.6 mmol) was injected to the solution, followed
by 3,4-dihydro-2H-pyran (2.50 ml, 27.4 mmol). The pink
solution was stirred at room temperature and turned dark
green after 10 min. When the reaction was complete by
TLC analysis, the solvent was evaporated and the resulting
dark green oil was purified via flash chromatography (10:1
hexanes/EtOAc), affording the product as a clear colourless
oil (2.10 g, 11.5 mmol, 85%). 1H NMR: δ 1.48–1.80 (m,
10H, Hb, Hc, Hd, Hg, Hh), 1.91 (s, 1H, Hj), 2.23 (t, 2H, Hi),
2.3. Synthesis of dimethylphenyl(tributylstannyl)silane
Dry THF (50 ml) was injected into a dry flask followed
by freshly distilled diisopropylamine (3.08 ml, 22.0 mmol).
This was cooled to −10 ◦C and a 1.6 M hexanes solution of
butyllithium (13.75 ml, 22.0 mmol) was injected dropwise.
After stirring for 10 min, Bu3SnH (5.38 ml, 20.0 mmol) was
injected dropwise. This stirred at 0 ◦C for 45 min, after which
time chlorodimethylphenylsilane (3.69 ml, 22.0 mmol) was
injected dropwise. The solution was allowed to warm to
room temperature and stir for 1 h, after which time GC-FID
analysis of a water-quenched aliquot showed almost com-
plete reaction. The solvent was evaporated under reduced
pressure and the resulting cloudy white oil was filtered
through celite and washed with Et2O. After evaporating the
Et2O under reduced pressure, the yellow oil was purified
via short-path distillation under vacuum (∼0.1 mmHg), the
3.37–3.51, 3.71–3.87 (m, 4H, Ha, Hf), 4.56 (t, 1H, He). 13
C
NMR: δ 18.4, 19.8, 25.6, 25.7, 29.0, 30.9, 62.5, 67.1, 68.6,
84.6, 99.0. MS (EI) m/z (rel. int. %): 181 (1), 125 (3), 111
(4), 101 (12), 85 (100), 79 (29), 67 (29), 55 (20), 41 (27).
2.6. Silylstannation of terminal alkynes
2.6.1. (Z)-1-Phenyl-1-(tributylstannyl)-2-(trimethylsilyl)
ethene (representative procedure)
Ionic liquid ([bmim]PF6 or [bmim]BF4, 1.0 ml) was
heated in a dry flask under vacuum at 70 ◦C for 4 h. After

36
I. Hemeon, R.D. Singer / Journal of Molecular Catalysis A: Chemical 214 (2004) 33–44
sat
releasing the vacuum to nitrogen, the flask was transferred
to an argon glove box where Pd(PPh3)4 (0.058 g, 0.05 mmol
for 5%, 0.012 g, 0.01 mmol for 1%) was weighed and added.
After stirring to suspend the palladium, the reaction vessel
was removed from the glove box. Bu3SnSiMe3 (0.42 ml,
1.2 mmol) was injected via needle and syringe, followed by
phenylacetylene (0.11 ml, 1.0 mmol) and dry Et2O (5.0 ml).
A dry reflux condensor was connected and the reaction was
heated under nitrogen in a 70 ◦C oil bath for 36 h. Once the
reaction was complete by GC–FID analysis of a reaction
aliquot, the reaction was cooled and the upper Et2O phase
was removed via needle and syringe. The ionic liquid was
washed with dry Et2O (8 × 8 ml) under nitrogen using
needles and syringes. The Et2O extracts were combined
and solvent was evaporated under reduced pressure. The
resulting oil was purified via flash chromatography (hex-
anes) to afford the product as a clear colourless oil (0.46 g,
0.99 mmol, 99%). 1H NMR: δ 0.32 (s, 9H, Si(CH3)3),
0.98–1.10 (m, 15H, Sn(CH2CH2CH2CH3)3), 1.34–1.65
=
(s, 1H, C C–H,
J
SnH
= 175, 183 Hz). 13C NMR: δ 0.22,
11.2, 13.6, 19.6, 25.5, 26.6, 27.5, 29.2, 29.3, 30.8, 47.5,
62.3, 67.6, 98.9, 143.7, 165.5.
2.6.5. (Z)-6-Chloro-2-(tributylstannyl)-1-(trimethylsilyl)
hex-1-ene
Prepared as above, clear colourless oil (0.327 g,
0.68 mmol, 68%). 1H NMR: δ 0.09 (s, 9H, Si(CH3)3),
0.86–0.96 (m, 15H, Sn(CH2CH2CH2CH3)3), 1.24–1.50,
1.71–1.77 (m, 16H, Sn(CH2CH2CH2CH3)3, ClCH2CH2CH2
=
=
CH2–C C), 2.28 (t, 2H, ClCH2CH2CH2CH2–C C,
3
3
=
J= 6.7 Hz), 3.52 (t, 2H, ClCH2CH2CH2CH2–C C, J =
sat
= 172, 180 Hz). 13C
=
6.7 Hz), 6.33 (s, 1H, C C–H,
J
SnH
NMR: δ 0.2, 11.2, 13.7, 27.0, 27.5, 29.2, 32.1, 45.0, 46.6,
144.1, 164.8. MS (EI) m/z (rel. int. %): 465 (1), 423 (37),
387 (67), 331 (12), 269 (100), 233 (24), 177 (27), 121 (13),
73 (73), 45 (23).
2.6.6. (Z)-2-(Dimethylphenylsilyl)-1-phenyl-1-
(tributylstannyl)ethene
=
(m, 12H, Sn(CH2CH2CH2CH3)3), 6.71 (s, 1H, C C–H,
sat
J
= 160, 168 Hz), 7.13–7.16 (m, 2H, ArH), 7.26–7.32
SnH
Prepared as above using Bu3SnSiMe2Ph (0.510 g,
1.2 mmol) instead of Bu3SnSiMe3, product obtained as
a clear colourless oil (0.516 g, 0.98 mmol, 98%). 1H
NMR: δ 0.11 (s, 6H, Si(CH3)2), 0.33–0.50 (m, 15H,
Sn(CH2CH2CH2CH3)3), 0.78–0.99 (m, 12H, Sn(CH2CH2
(m, 1H, ArH), 7.38–7.41 (m, 2H, ArH). 13C NMR: δ 0.3,
12.1, 13.8, 27.5, 29.2, 125.6, 126.0, 128.0, 148.5, 152.0,
166.2. MS (EI) m/z (rel. int. %): 451 (6), 409 (96), 353 (49),
295 (15), 235 (31), 175 (43), 121 (13), 73 (100), 45 (23).
sat
=
CH2CH3)3), 6.41 (s, 1H, C C–H,
J
= 156, 164 Hz),
SnH
6.70–7.04, 7.24–7.28 (m, 10H, ArH). 13C NMR: δ −0.5,
12.0, 13.7, 27.4, 29.1, 125.8, 126.0, 127.9, 128.1, 129.1,
134.2, 139.6, 146.0, 152.0, 169.0. MS (EI) m/z (rel. int. %):
513 (1), 471 (100), 414 (7), 357 (3), 291 (2), 233 (6), 177
(15), 135 (40).
2.6.2. (Z)-2-(Tributylstannyl)-1-(trimethylsilyl)dec-1-ene
Prepared as above, clear yellow oil (0.484 g, 0.97 mmol,
97%). 1H NMR: δ 0.09 (s, 9H, Si(CH3)3), 0.79–1.58
=
(m, 42H, SnBu3, C7H15CH2–C C), 2.20–2.33 (m, 2H,
sat
=
=
C7H15CH2–C C), 6.32 (s, 1H, C C–H,
J
= 176,
SnH
184 Hz). 13C NMR: δ 0.4, 8.0, 11.4, 13.8, 14.3, 22.8, 27.7,
29.4, 29.7, 30.2, 30.5, 32.1, 47.8, 143.3, 166.0. MS (EI) m/z
(rel. int. %): 488 (7), 444 (86), 390 (57), 334 (17), 291 (3),
235 (28), 179 (36), 121 (9), 73 (100).
2.6.7. (Z)-1-(Dimethylphenylsilyl)-2-(tributylstannyl)
dec-1-ene
Prepared as above, clear colourless oil (0.439 g,
0.78 mmol, 78%). 1H NMR: δ 0.34 (s, 6H, Si(CH3)2),
2.6.3. (Z)-5-(Tributylstannyl)-6-(trimethylsilyl)hex-5-
en-1-ol
Prepared as above, cloudy white oil (0.412 g, 0.89 mmol,
89%). 1H NMR: δ 0.09 (s, 9H, Si(CH3)3), 0.87–0.99
(m, 15H, Sn(CH2CH2CH2CH3)3), 1.25–1.68 (m, 17H,
=
0.73–0.88, 1.18–1.41 (m, 42H, SnBu3, C7H15CH2–C C),
=
=
2.29–2.35 (m, 2H, C7H15CH2–C C), 6.50 (s, 1H, C C–H,
sat
J
= 170, 176 Hz), 7.30–7.35 (m, 3H, ArH), 7.49–7.54
SnH
(m, 2H, ArH). 13C NMR: δ −0.6, 11.1, 13.7, 14.2, 22.7,
27.5, 28.0, 29.2, 29.4, 29.6, 31.0, 31.9, 47.8, 127.7, 128.8,
134.0, 140.1, 140.7, 168.9. MS (EI) m/z (rel. int. %):
550 (1), 508 (100), 452 (8), 394 (3), 233 (7), 177 (17),
135 (37).
=
Sn(CH2CH2CH2CH3)3, HOCH2CH2CH2CH2–C ), 2.30 (t,
=
2H, HOCH2CH2CH2CH2–C C), 3.64 (m, 2H, HOCH2CH2
sat
=
=
CH2CH2–C C), 6.34 (s, 1H, C C–H,
J
= 172,
SnH
180 Hz). 13C NMR: δ 0.2, 11.2, 13.6, 25.9, 27.5, 29.2, 32.3,
47.1, 62.9, 143.7, 165.1. MS (EI) m/z (rel. int. %): 447 (2),
405 (100), 347 (5), 291 (4), 235 (14), 193 (39), 135 (10),
73 (56), 44 (24).
2.6.8. (Z)-6-(Dimethylphenylsilyl)-5-(tributylstannyl)hex-
5-en-1-ol
Prepared as above, cloudy oil (0.325 g, 0.62 mmol,
62%). 1H NMR: δ 0.36 (s, 6H, Si(CH3)2), 0.81–0.90
(m, 15H, Sn(CH2CH2CH2CH3)3), 1.19–1.64 (m, 17H,
2.6.4. ( )-(Z)-2-(5-(Tributylstannyl)-6-(trimethylsilyl)-
5-hexenyloxy)tetrahydro-2H-pyran
=
Sn(CH2CH2CH2CH3)3),
2.37 (t, 2H, HOCH2CH2CH2CH2–C C), 3.65–3.67 (m,
HOCH2CH2CH2CH2–C C),
=
Prepared as above, clear colourless oil (0.538 g,
0.99 mmol, 99%). 1H NMR: δ 0.08 (s, 9H, Si(CH3)3),
0.86–0.95 (m, 15H, Sn(CH2CH2CH2CH3)3), 1.27–1.65 (m,
22H, Sn(CH2CH2CH2CH3)3, Hb, Hc, Hd, Hg, Hh), 2.29 (t,
2H, Hi), 3.30–3.90 (m, 4H, Ha, Hf), 4.58 (t, 1H, He), 6.33
=
=
2H, HOCH2CH2CH2CH2–C C), 6.53 (s, 1H, C C–H,
sat
J
= 170, 178 Hz), 7.31–7.55 (m, 5H, ArH). 13C
SnH
NMR: δ −0.4, 11.4, 13.8, 26.2, 27.6, 29.5, 32.6, 47.4, 63.0,
127.9, 129.0, 134.4, 139.6, 141.5, 168.4.

I. Hemeon, R.D. Singer / Journal of Molecular Catalysis A: Chemical 214 (2004) 33–44
37
2.6.9. ( )-(Z)-2-(6-(Dimethylphenylsilyl)-5-(tributyl-
stannyl)-5-hexenyloxy)tetrahydro-2H-pyran
LDA
THF, -10 o
PhMe2SiCl
oC 25 o
Bu3SnSiMe2Ph
LiCl
Bu3SnH
Bu3Sn
C
0
C
Prepared as above, clear colourless oil (0.492 g,
0.83 mmol, 83%). 1H NMR: δ 0.36 (s, 6H, Si(CH3)2),
0.77–0.93 (m, 15H, Sn(CH2CH2CH2CH3)3), 1.18–1.72 (m,
22H, Sn(CH2CH2CH2CH3)3, Hb, Hc, Hd, Hg, Hh), 2.37 (t,
2H, Hi), 3.38–3.90 (m, 4H, Ha, Hf), 4.58 (t, 1H, He), 6.53
Fig. 3. Synthesis of dimethylphenyl(tributylstannyl)silane [12].
commercial availability and its general use in the literature.
To increase the scope of our methodology, further inves-
tigations using an additional trialkyl(trialkylstannyl)silane
were conducted. Dimethylphenyl(tributylstannyl)silane
(Bu3SnSiMe2Ph) was easily synthesised via generation of a
tributyltin anion followed by a quench with chlorodimethyl-
phenylsilane (Fig. 3). This trialkyl(trialkylstannyl)silane has
been used in addition reactions with alkynes to generate
products more easily protodesilylated than their trimethylsi-
lyl counterparts [15].
sat
=
(s, 1H, C C–H,
J
= 177, 185 Hz), 7.26–7.55 (m, 5H,
SnH
ArH). 13C NMR: δ −0.5, 11.2, 13.8, 19.8, 25.7, 26.7, 27.6,
29.3, 29.5, 30.9, 47.6, 62.3, 67.5, 98.9, 127.8, 128.9, 134.1,
140.0, 141.2, 168.6.
3. Results and discussion
3.1. Addition of Bu3SnSiMe3 to terminal alkynes
in ionic liquids
Our initial attempts at coupling phenylacetylene with ex-
cess Bu3SnSiMe3 were performed in neat [bmim]PF6 at
70 ◦C without the addition of a co-solvent. While gas chro-
matographic analysis showed no phenylacetylene present af-
ter 18 h, isolated yields were less than quantitative (i.e. 66%).
It was observed that material was condensing on the walls of
the flask during the reaction and was not coming in contact
with the palladium/ionic liquid phase. For subsequent reac-
tions the flask was fitted with a dry reflux condensor and di-
ethyl ether was used as a co-solvent to “wash” the reagents
down into the ionic liquid phase as the ether refluxed in
the 70 ◦C oil bath. Diethyl ether was chosen due to its im-
miscibility with the ionic liquid and its ability to dissolve
the starting materials and products. This resulted in more
favourable isolated yields. Bu3SnSiMe3 was added to five
terminal alkynes of differing functionality in the presence of
5 or 1 mol% Pd(PPh3)4 in [bmim]PF6 and [bmim]BF4 ionic
liquids with diethyl ether as a co-solvent in a 70 ◦C oil bath,
the results of which are presented in Table 1. All reactions
were monitored by gas chromatography and were deemed
complete after the disappearance of the alkyne starting ma-
terial peak.
The ionic liquid 1-n-butyl-3-methylimidazolium hex-
afluorophosphate was investigated as a solvent during the
initial stages of our study of addition reactions of tri-
alkyl(trialkylstannyl)silanes to terminal alkynes due to its
prevalent use in a variety of palladium-catalysed reactions.
1-n-Butyl-3-methylimidazolium tetrafluoroborate was also
used in the course of our studies of this reaction. These ionic
liquids are perhaps the two most widely used in current
literature. They are liquids below room temperature, they
have a wide liquid temperature range, they are moisture-
and heat-stable, and they are immiscible with a number
of organic solvents. They are commercially available but
can be easily synthesised for a small fraction of the cost
[20,21]. For example, 1 l of 1-methylimidazole cost $251,
1 l of 1-chlorobutane cost $88, and 500 g of 60% HPF6 cost
$70 from Aldrich in the 2003–2004 catalogue, allowing
[bmim]PF6 to be synthesised for less than 5% of the cost of
its purchase ($93 for 5 g), based on raw material price. Sev-
eral research groups, including our own, have investigated
various methods of ionic liquid preparation for product pu-
rity, especially with regard to halide contamination [24,25].
The methods used in this paper appear to provide some of
the cleanest ionic liquids.
Table 1
≡
Results of Bu3SnSiMe3 addition to terminal alkynes R–C CH in
[bmim]PF6/Et2O and [bmim]BF4/Et2O in the presence of Pd(PPh3)4 heat-
ing at 70 ◦C
The catalyst tetrakis(triphenylphosphine)palladium(0)
Pd(PPh3)4 was chosen based on previous studies by Chenard
and van Zyl [13] demonstrating its activity in the addition
of trialkyl(trialkylstannyl)silanes to terminal alkynes. This
group attempted reactions using different palladium(0) and
palladium(II) catalysts as well as molybdenum-, rhodium-
and platinum-containing catalysts and concluded that
Pd(PPh3)4 was by far the superior catalyst. These results on
catalyst activity with terminal alkynes were supported by
the work of other groups [14,15] as well by a reaction we
performed using a palladium(II) catalyst with no success as
will be discussed further on.
Entry R-group
of alkyne
Ionic
liquid
Pd
(mol%)
Reaction
time (h)
Isolated
yield (%)
1
2
Ph
Ph
[bmim]PF6
[bmim]PF6
[bmim]BF4
[bmim]PF6
[bmim]PF6
[bmim]BF4
[bmim]PF6
[bmim]PF6
[bmim]BF4
5
1
5
5
1
5
5
1
5
5
5
1
36
17
36
15
72
120
18
24
24
84
100
99
100
97
100
82
89
87
61
99
68
3
Ph
4
5
6
7
8
9
(CH2)7CH3
(CH2)7CH3
(CH2)7CH3
(CH2)4OH
(CH2)4OH
(CH2)4OH
10
11
12
(CH2)4OTHP [bmim]PF6
(CH2)4Cl
(CH2)4Cl
[bmim]PF6
[bmim]PF6
120
144
22
The trialkyl(trialkylstannyl)silane used in our original
investigation of these reactions in [bmim]PF6, trimethyl-
(tributylstannyl)silane (Bu3SnSiMe3) was chosen due to its
Reaction conditions: 1 mmol alkyne, 1.2 mmol Bu3SnSiMe3, 1.0 ml ionic
liquid, 5.0 ml Et2O, 70 ◦C oil bath; THP: tetrahydropyranyl.

38
I. Hemeon, R.D. Singer / Journal of Molecular Catalysis A: Chemical 214 (2004) 33–44
As can be seen from the data presented, the addition
THF, indicating that these chloroalkynes are less reactive
[13].
of Bu3SnSiMe3 to the five terminal alkynes proceeded in
good to excellent yields under all conditions tested. All
reactions proceeded in excellent yield (i.e. >95%) when
reaction aliquots were assessed by gas chromatography as
evidenced by the complete disappearance of the alkyne sub-
strate peak. The GC analyses were supported in most cases
by equally high isolated yields after column chromatogra-
phy, separating the product from excess Bu3SnSiMe3. All
reactions proceeded to give only a single product isomer,
the 1,2-bismetallated Z-alkene containing the trimethylsilyl
group on the terminal carbon and the tributylstannyl group
on the internal carbon (Fig. 2). This was confirmed by the
value of the coupling constant between the tin atom and the
terminal vinylic proton. This three-bond coupling constant
(3J) occurs from coupling between the vinylic proton and
the two abundant NMR-active isotopes of tin, 117Sn and
119Sn, both of which have a spin of 1/2 and are 7.68 and
8.59% abundant, respectively [26]. 115Sn also has a spin
of 1/2 but is not abundant enough to be easily detected
(0.34%). The coupling constant is known to be around
180 Hz for trans-coupling and around 100 Hz for cis- and
geminal-coupling [13]. The products from the reactions de-
As can be seen from the results in Table 1, the reactions
proceed better with 5 mol% Pd than they do with 1 mol%
Pd. Although this may be the intuitive result of increasing
the amount of catalyst, it demonstrates that in general the
reaction proceeds smoothly with 1 mol% Pd to give compa-
rable yields to those obtained from reactions using 5 mol%
Pd. The reactions with less catalyst simply require longer
reaction times. It may be that the observed differences in
yields between the two catalyst loadings are results of cata-
lyst poisoning or deactivation.
As mentioned earlier, Pd(PPh3)4 was chosen as the cat-
alyst as a result of previous studies by other groups show-
ing it to display optimal activity [13,14]. We tried using
bis(acetonitrile)palladium(II) chloride ((CH3CN)2PdCl2) as
a palladium(II) catalyst for the reaction with phenylacety-
lene. After 24 h at room temperature in a [bmim]PF6/Et2O
system with 5 mol% catalyst very little reaction had oc-
curred. After heating for 48 h in a 70 ◦C oil bath very
little product had formed. A large amount of Bu3SnSiMe3
remained in addition to an appreciable amount of hex-
abutylditin, Bu3SnSnBu3, having been formed after ex-
tended reaction times. This by-product is likely the result
of a palladium-catalysed disproportionation reaction known
to occur with trimethyl(trimethylstannyl)silane during reac-
tion with unreactive alkynes resulting in the formation of
hexamethyldisilane and hexamethyldistannane [13].
The reaction between phenylacetylene and Bu3SnSiMe3
catalysed by Pd(PPh3)4 has previously been shown to pro-
ceed at room temperature in THF [13]. When we performed
the same reaction in [bmim]PF6 at room temperature with-
out Et2O as a co-solvent, no reaction was observed after
18 h (our initial results showed that these reactions pro-
ceed smoothly in neat [bmim]PF6 at 70 ◦C and that Et2O
is not necessary to promote the reaction [18]). After the
addition of Et2O and heating in a 70 ◦C oil bath, com-
plete reaction was observed overnight. The ionic liquid
phase was a light yellow colour, showing no deposition
of palladium black. The reactions in THF started with the
light orange colour of the starting palladium catalyst and
quickly turned black, suggesting that palladium black de-
posited. These reactions are thus very slow in ionic liquids
alone at room temperature and require heating to increase
the rate; however, the catalyst appears to be more stable
demonstrating potential recyclability. A kinetic study was
undertaken to measure the initial reaction rate between
phenylacetylene and Bu3SnSiMe3 in [bmim]PF6 in the
presence of 5 mol% Pd(PPh3)4 heating in a 70 ◦C oil bath.
After 8 h, however, the reaction had only proceeded to
4.5% completion. It was evident from this that there was
no rate enhancement of the reaction resulting from the use
of ionic liquids as the same reaction was complete in 3 h
in refluxing THF; hence, no further kinetic measurements
were performed on the ionic liquid systems used in our
study.
3
scribed in Table 1 all have J coupling constants between
160 and 184 Hz.
The isolated yields of these reactions compare well
with literature values, being slightly higher in most cases
[13–15]. In many cases the reaction times are somewhat
lengthy, being more than 24 h; however, this is typical of
these reactions. As well, these reactions were allowed to
proceed to completion, based on absence of starting mate-
rial, as monitored by gas chromatography before products
were isolated. Phenylacetylene afforded quantitative yields
of product after 17 h with 1% Pd (Table 1, Entry 2), some-
what longer than the 3 h required to obtain a 91% yield
in refluxing THF [13]. 1-Decyne gave a 97% yield after
15 h with 5% Pd (Table 1, Entry 4), while a 52% yield
had previously been obtained with the analogous alkyne
1-hexyne after 20 h under solventless conditions at 80 ◦C
[14]. The 5-hexyn-1-ol addition product was isolated in
89% yield after 18 h (Table 1, Entry 7) with 5% Pd while
the adduct from the analogous alkyne 4-pentyn-1-ol was
previously obtained in 88% yield after refluxing in THF
for 72 h [13]. The tetrahydropyranyl (THP) ether pre-
pared from 5-hexyn-1-ol and 2,3-dihydro-2H-pyran [23]
(2-(5-hexynyloxy)tetrahydro-2H-pyran) was coupled with
Bu3SnSiMe3 to afford the product in 99% yield after 84 h
(Table 1, Entry 10). For comparison, a 92% yield had
previously been obtained with the THP ether of the anal-
ogous alkyne 3-butyn-1-ol after 48 h in refluxing THF
[13]. 5-Hexyn-1-ol was also used as a starting material
to generate 6-chlorohex-1-yne using thionyl chloride and
pyridine [22]. 6-Chlorohex-1-yne gave the addition product
in somewhat lower yield, 68% after 120 h (Table 1, Entry
11). This is comparable to the 80% isolated yield using the
analogous alkyne 5-chloropent-1-yne after 72 h in refluxing

I. Hemeon, R.D. Singer / Journal of Molecular Catalysis A: Chemical 214 (2004) 33–44
39
Table 2
observed using THF as a solvent [13–15]. Lower yields us-
ing Bu3SnSiMe2Ph may be due to steric factors influencing
reactions with the larger dimethylphenylsilyl moiety.
≡
Results of Bu3SnSiMe2Ph addition to terminal alkynes R–C CH in
[bmim]PF6/Et2O and [bmim]BF4/Et2O in the presence of 5% Pd(PPh3)4
heating at 70 ◦C
When studying the results in terms of which ionic liquid is
a better solvent for the addition of trialkyl(trialkylstannyl)-
silanes to alkynes, the data shows that [bmim]PF6 af-
fords somewhat better yields of products in less time than
[bmim]BF4. This trend is observed regardless of which tri-
alkyl(trialkylstannyl)silane or alkyne is employed, with the
exception of phenylacetylene which has been shown to be
exceptionally reactive in these reactions [13]. It thus appears
as though the reaction in [bmim]BF4 is slower than that
in [bmim]PF6. The reason for this does not immediately
present itself, as these two ionic liquids are very similar
in terms of their physical properties. Viscosity differences
may be proposed as a factor in the rates of these reac-
tions. If [bmim]BF4 were more viscous than [bmim]PF6,
this may explain the longer reaction times as molecules
would take longer to move through the solvent to reach one
another to react. However, several groups have measured
the viscosities of these ionic liquids (albeit with each ob-
taining slightly different results), showing them to range
from both ionic liquids having almost the same viscosity to
[bmim]PF6 being about twice as viscous as [bmim]BF4 [1].
In addition, ionic liquids generally become much less vis-
cous once organic materials such as reagents or co-solvents
are added. Chloride impurities have been shown to increase
the viscosity of ionic liquids [24]. These impurities r−esult
from incomplete reaction of [bmim]Cl with the PF6 or
Entry
R-group
on alkyne
Ionic
liquid
Reaction
time (h)
Isolated
yield (%)
1
2
3
4
5
6
Ph
Ph
[bmim]PF6
[bmim]BF4
[bmim]PF6
[bmim]BF4
[bmim]PF6
[bmim]PF6
96
96
108
144
19
98
97
78
62
62
83
(CH2)7CH3
(CH2)7CH3
(CH2)4OH
(CH2)4OTHP
168
Reaction conditions: 1.0 mmol alkyne, 1.2 mmol Bu3SnSiMe2Ph, 1.0 ml
ionic liquid, 5.0 ml Et2O, 70 ◦C oil bath; THP: tetrahydropyranyl.
3.2. Addition of Bu3SnSiMe2Ph to terminal alkynes in
ionic liquids
The slightly larger silylstannane Bu3SnSiMe2Ph was also
used in these reactions in both ionic liquids. The results
of the addition of Bu3SnSiMe2Ph to four alkynes in the
presence of 5% Pd are presented in Table 2. As with the
less bulky trimethyl(tributylstannyl)silane, the majority of
the reactions proceeded in good to excellent yields. These
reactions were also monitored by gas chromatography for
the disappearance of the alkyne substrate peak. As well the
reaction proceeded regio- and stereoselectively, giving the
1,2-bismetallated Z-alkene containing the trialkylsilyl group
on the terminal carbon and the tributylstannyl group on the
internal carbon. The isomeric natures of the products were
3
−
again confirmed by the values of the tin–vinylic proton J
BF4 source (e.g. NaBF4 versus HBF4). Studies have been
coupling constants corresponding with trans-coupling, the
values being between 156 and185 Hz.
performed to determine which method of ionic liquid prepa-
ration results in the lowest amount of chloride contamina-
tion. The acid/base neutralisation method used in this study
involving HBF4 and HPF6 provides ionic liquid with some
of the lowest levels of chloride contamination [24,25]. It is
thus unlikely that viscosity differences are responsible for
the differences in efficiency of each ionic liquid for these
reactions.
Since the [bmim]PF6 ionic liquid is “hydrophobic” and
[bmim]BF4 is water-soluble, it follows that [bmim]PF6
would have a lower water content. Although the ionic liq-
uids were dried by heating under vacuum for several hours
before use, they still contain a certain amount of water as ev-
idenced by Rogers and co-workers [1]. This group showed
that after heating under vacuum for 4 h at 70 ◦C [bmim]BF4
contained 4530 ppm water, about 7.5 times the amount held
by [bmim]PF6 with 590 ppm, showing that [bmim]PF6 is
not strictly “hydrophobic” as is commonly claimed. The
presence of water in both ionic liquids could account for
some of the longer reaction times when compared to THF
and would also account for the longer reaction time re-
quired for [bmim]BF4 since it has substantially more bound
water than [bmim]PF6. Indeed, a silylstannation reaction
was performed in [bmim]BF4 that had only been dried
for 3 h, which after 24 h showed mainly decomposition of
Bu3SnSiMe3 (which is moisture-sensitive).
As with the products from the coupling of Bu3SnSiMe3,
the isolated yields of products from reactions with
Bu3SnSiMe2Ph were generally similar to those obtained by
other researchers who used THF as a solvent [15]. A 98%
yield of the addition product from phenylacetylene was ob-
tained in the ionic liquid after 96 h (Table 2, Entry 1), while
a 95% yield had previously been obtained after only 3 h in
refluxing THF [15]. A 78% yield of the addition product
from 1-decyne was obtained from the ionic liquid after 108 h
(Table 2, Entry 3), which is comparable to the 80% yield ob-
tained previously with 1-hexyne, an analogous alkyne, after
120 h [15]. The 5-hexyn-1-ol addition product was obtained
in 62% yield after only 19 h in the ionic liquid (Table 2,
Entry 5) while an 81% yield of product was reported with
the analogous alkyne 3-butyn-1-ol after 120 h in refluxing
THF [15]. 2-(5-Hexynyloxy)tetrahydro-2H-pyran reacted
to produce an 83% yield of addition product after 168 h in
the ionic liquid (Table 2, Entry 6), which was similar to
the 82% yield reported with the analogous THP ether of
3-butyn-1-ol after 120 h in refluxing THF [15]. It appears as
though the reactions using Bu3SnSiMe2Ph are slower than
those using Bu3SnSiMe3 as the product yields are slightly
lower and the reaction times are longer. This is likely not
due to the use of an ionic liquid since the same trend is

40
I. Hemeon, R.D. Singer / Journal of Molecular Catalysis A: Chemical 214 (2004) 33–44
SnR'
SiR"
Chenard and van Zyl [13] have stated that less reactive
3
3
R C CH
Et O
2
alkynes give poorer yields of addition product because they
are poor ligands for palladium due to steric bulk. This would
explain the lower yields obtained with Bu3SnSiMe2Ph as
this silylstannane imposes even more steric bulk around
the palladium atom. However, relatively high yields were
obtained from reactions using 2-(5-hexynyloxy)tetrahydro-
2H-pyran and 1-decyne while lower yields were obtained
from reactions using 5-hexyn-1-ol and 6-chlorohex-1-yne.
1-Decyne and 2-(5-hexynyloxy)tetrahydro-2H-pyran have
the largest substituents but are also the least volatile, giv-
ing the highest yields. In fact, 6-chlorohex-1-yne was ob-
served to have a relatively high volatility that may have
contributed to low isolated yields of its addition products.
It may also be that the addition products of 5-hexyn-1-ol
and 6-chlorohex-1-yne had greater affinity for the silica
gel used during chromatography, complicating purification.
The fact that alkynes with variable functionality can be
coupled with trialkyl(trialkylstannyl)silanes shows some
tolerance of this reaction for different functional groups.
Aromatic alkynes, saturated alkynes, and hydroxy-, ether-
and chloride-bearing alkynes all coupled with two different
trialkyl(trialkylstannyl)silanes in two different ionic liquids
in the presence of Pd(PPh3)4.
SnR'
SiR"
SiR"
L Pd SnR'
L
3
[bmim]PF
3
6
R C CH
3
3
L
-2L
SiR"
L Pd SnR'
HC CR
3
PdL
3
4
+2L
SiR"
3
H
L
R
H
L Pd
R
L
R' Sn SiR"
3
3
R' Sn
3
Fig. 4. Proposed reaction sequence and mechanism for palladium-catalysed
trialkyl(trialkylstannyl)silane addition to terminal alkynes in ionic liquid/
Et2O biphasic systems.
3.3. Catalytic cycle
Under biphasic conditions, the trialkyl(trialkylstannyl)-
silane and alkyne are dissolved in the upper Et2O phase
from which they partition into the lower ionic liquid phase
containing the palladium catalyst. The reaction proceeds
in the ionic liquid, after which the products partition back
into the Et2O phase. The starting materials and products
are very soluble in Et2O and are extracted quite easily from
the ionic liquid. The final washings of the ionic liquids are
always devoid of products or starting materials by gas chro-
matography. As well, the palladium catalyst resides in the
ionic liquid phase preferentially and is likely not extracted
during product isolation with Et2O, as evidenced by several
researchers [27–29]. It may be that the reaction products are
more soluble in [bmim]BF4 than in [bmim]PF6 and reside
in the [bmim]BF4 phase preferentially. If the reactions were
inhibited by the build-up of products in the catalytic phase,
impeding further reaction and lowering the rate, the reac-
tions would take longer to proceed using the ionic liquid in
which the products were more soluble. It is likely that the
catalytic reaction proceeds in the ionic liquid phase since
the palladium catalyst resides there (Fig. 4).
The mechanism of the catalytic cycle depicted in Fig. 4
has been developed based on experimental evidence, on the
proposed mechanism of the closely related bis-silylation of
unsaturated molecules, and on a theoretical study using ab
initio Hartree–Fock calculations to determine the energies
of steps during the reaction of the hypothetical silylstannane
H3SnSiH3 catalysed by the hypothetical catalyst Pd(PH3)2
[17,30]. After the silylstannane and alkyne partition into
the ionic liquid phase, the first step in the cycle is oxidative
addition of the silylstannane to the palladium(0) species,
followed by co-ordination of the alkyne. The next step,
regioselective insertion of the alkyne into the palladium–tin
bond, is presumed to be the rate-determining step based on
calculations. The final step is then reductive elimination of
the product to regenerate the palladium(0) species and give
the product as a single isomer, which would then dissolve
in the upper Et2O phase.
3.4. Recyclability studies
The ionic liquid/catalyst system was extensively investi-
gated regarding recyclability. In studies performed by other
groups using palladium(0) catalysts for allylation reactions,
their [bmim]PF6/catalyst systems were processed between
successive reactions using a complex purification procedure
[31,32]. This process involved washing the ionic liquid with
water, dissolving it in ethyl acetate, drying it over sodium
sulphate, and evaporating the ethyl acetate to recover the
ionic liquid that still contained palladium. The ionic liq-
uid/catalyst system recovered in this manner continuously
lost activity over successive cycles. When working with
a palladacycle palladium(II) source for the Heck reaction
in ionic liquids, Bohm and Herrmann [28] noted that ex-
tremely careful drying procedures had to be undertaken in
order to obtain optimal activity. They also saw that addition
of small amounts of water lowered the catalyst’s turnover
number and lifetime, indicating that washing the ionic liq-
uid/catalyst system with water as a purification method be-
tween successive reactions may be responsible for loss of
catalyst activity. Mathews et al. [27], however, found that
washing their ionic liquid/Pd(PPh3)4 catalyst system had no
effect on its activity in Suzuki couplings. In fact, Hagiwara
et al. [33] found that a water wash was necessary to remove
ionic impurities that built up from Heck reactions catalysed
by Pd/C that was suspended in the ionic liquid. When they
performed a water wash after the fifth cycle, the yield that
had been steadily decreasing increased back to a value com-
parable to that obtained after the first cycle. Many other
groups have also found that palladium catalysts can be re-

I. Hemeon, R.D. Singer / Journal of Molecular Catalysis A: Chemical 214 (2004) 33–44
41
after 48 h as evidenced by the complete disappearance of
the alkyne substrate peak. In addition, the reaction between
Bu3SnSiMe2Ph and phenylacetylene was performed four
times in a [bmim]PF6/Pd(PPh3)4 system and five times in
a [bmim]BF4/Pd(PPh3)4 system, also without loss of activ-
ity (Fig. 5b). These reactions were also monitored by gas
chromatography and were generally complete in 72 h.
As can be seen from the results presented in Fig. 5,
the ionic liquid/palladium catalyst systems are highly re-
cyclable. The results presented in Fig. 5 represent isolated
yields, obtained after the reaction was deemed complete by
gas chromatography. The reaction conditions for the recy-
clability studies were the same as before, employing 5 mol%
Pd(PPh3)4 in a biphasic ionic liquid/Et2O system heated in
a 70 ◦C oil bath under nitrogen. Products were isolated by
washing the ionic liquid/catalyst under nitrogen with dry
Et2O, which was then removed via needle and syringe. Fresh
trialkyl(trialkylstannyl)silane and alkyne were then injected
along with dry Et2O and the reaction repeated. Fig. 6 depicts
the process used to recycle the ionic liquid/catalyst systems.
The ionic liquid/catalyst system is very robust and
amenable to continuous recycling. Recycling of the system
using 1-decyne and 5-hexyn-1-ol was possible, also with-
out activity loss. On one occasion, the ionic liquid/catalyst
system previously used for 5-hexyn-1-ol was recycled us-
ing 1-decyne and vice versa. These systems provided their
expected addition products, showing that the ionic liq-
uid/catalyst system can be recycled with any alkyne regard-
less of which alkyne was used previously. Most of the ionic
liquid/catalyst systems investigated for recyclability con-
tained 5 mol% palladium, but recycling was also performed
on occasion with systems containing 1 mol% palladium,
which also showed no loss of activity. In general, the next
Fig. 5. Recyclability studies on [bmim]PF6 and [bmim]BF4 contain-
ing 5 mol% Pd(PPh3)4 in the coupling of phenylacetylene with (a)
Bu3SnSiMe3 and (b) Bu3SnSiMe2Ph, both heating in a 70 ◦C oil bath.
cycled in ionic liquids with little loss in activity simply by
extracting the reaction products from the ionic liquid with
an organic solvent such as Et2O. Generally, this does not
result in leaching of palladium species from the ionic liquid
[20,29,34,35].
We have previously reported that Pd(PPh3)4 could be
recycled in [bmim]PF6 four times in the coupling of
Bu3SnSiMe3 and phenylacetylene without loss of activity
[18]. Here, we report that the same reaction can be per-
formed at least 10 times with the same ionic liquid/catalyst
system without loss of activity; the same has been shown for
[bmim]BF4, as well (Fig. 5a). These reactions were moni-
tored by gas chromatography and were generally complete
Bu3SnSiMe3
CH
ionic liquid
+ catalyst
Et2O +
products
R
C
Et2O
Et2O
(8x8 mL)
Et2O +
products
Et2O +
starting
materials
ionic liquid
+ catalyst
ionic liquid
+ catalyst
reaction
proceeds
Fig. 6. Ionic liquid/catalyst recycling process.

42
I. Hemeon, R.D. Singer / Journal of Molecular Catalysis A: Chemical 214 (2004) 33–44
reaction during recycling was performed immediately after
the product was isolated from the previous reaction. How-
ever, the ionic liquid/catalyst system has been left for up to
10 days without special precautions, after which the catalyst
displayed no loss of activity in subsequent reactions.
R
N
R
N
Y
Pd
Y
N
R
N
R
Y = Br, BF4
3.5. Catalyst activation
Fig. 7. Palladium(0)–dialkylimidazolylidene carbene complex isolated
from Heck reactions performed in [bmim]BF4 and [bmim]Br ionic liquids
[41].
Interestingly, the first reaction performed in both ionic
liquids using both trialkyl(trialkylstannyl)silanes required
longer reaction times than subsequent reactions that went to
completion much faster. In fact, after 144 h the reaction of
Bu3SnSiMe3 with phenylacetylene in [bmim]BF4 was only
∼85% complete by GC and an isolated yield of 58% was ob-
tained. In the second cycle for this system, however, 100%
GC and isolated yields were obtained after only 36 h. This
seems to indicate that some time may be required for “cat-
alyst activation” as has been observed for other palladium-
catalysed reactions in ionic liquids that require pre-heating
of the ionic liquid/catalyst system to obtain optimal activity
[27,36–38].
We investigated the necessity of catalyst activation by
performing the coupling of Bu3SnSiMe3 and phenylacety-
lene under various conditions, the results of which are
presented in Table 3. In the case of Suzuki reactions, other
groups heated their palladium catalyst in an ionic liquid
along with the aryl halide substrate at high temperatures,
after which they added the boronic acid substrate and car-
bonate base to promote the cross-coupling reaction [27,36].
We heated 5 mol% Pd(PPh3)4 in [bmim]PF6 at 70 ◦C which
was black after 20 h. The silylstannation reaction performed
in this system proceeded to give a 97% isolated yield but
required 120 h to reach completion (Table 3, Entry 1). We
obtained more favourable results heating the catalyst in the
biphasic ionic liquid/Et2O systems. Pd(PPh3)4 (5 mol%)
was heated in [bmim]PF6/Et2O and in [bmim]BF4/Et2O
in a 70 ◦C oil bath along with Bu3SnSiMe3. After 20 h,
phenylacetylene was added to the bright yellow reagent mix
and coupled to provide 100 and 92% yields after 19 and
20 h in [bmim]PF6/Et2O and [bmim]BF4/Et2O, respectively
(Table 3, Entries 4 and 5). In fact, GC analyses of these
reactions indicated they had proceeded ∼65 and ∼90% to
completion after only 3 h, respectively. The catalyst was
also heated in a 70 ◦C oil bath in a biphasic system con-
sisting of [bmim]PF6/Et2O without any additional reagents.
After 20 h, Bu3SnSiMe3 and phenylacetylene were added
to the bright orange catalytic mixture, coupling to give
100% yield after 20 h (Table 3, Entry 2). This result shows
that it is not necessary to have silylstannane present at the
activation stage and that the formation of any other catalytic
species in the ionic liquid is not a result of reagent addition
to the palladium(0) centre. It seems that mild heating of the
catalyst in the ionic liquid is sufficient to activate the cata-
lyst. Heating Pd(PPh3)4 in neat ionic liquid at 70 ◦C does
not provide active catalyst species, as evidenced by the long
time required to achieve complete addition of Bu3SnSiMe3
to phenylacetylene. This may be a result of the high temper-
ature. Temperature measurements of the reaction mixture
rather than the oil bath made while heating 1.0 ml of ionic
liquid with 5.0 ml Et2O in a 70 ◦C oil bath show that the
temperature of the ionic liquid does not exceed 40 ◦C.
Presumably the temperature of the ionic liquid is being
regulated by the boiling point of Et2O (35 ◦C). It there-
fore seems that the catalyst is activated simply with gentle
heating in the ionic liquid. Furthermore, these pre-activated
systems were recyclable, giving comparable results on their
second cycles (Table 3, Entries 3 and 6).
Some researchers ascribe these catalyst activation effects
to the possible formation of a palladium–imidazolylidene
carbene complex (Fig. 7). These types of complexes, pre-
pared independently of ionic liquid research, have been
shown to be very active in Heck and Suzuki couplings in
conventional solvents [39], showing excellent stabilities and
reaction rates [40].
Several groups have been able to observe and in some
cases isolate palladium–imidazolylidene carbene com-
plexes from Suzuki and Heck reactions in ionic liquids.
The proton on carbon 2 of the imidazolium ring is acidic
and can be removed with a base to generate a carbene.
Xu et al. [41] observed the formation of such a complex
during a Heck reaction in [bmim]Br as the reaction colour
changed from dark brown to red to yellow, similar to the
colour of reactions performed in our studies. The palla-
dium complex was isolated from a reaction performed in
a mixture of [bmim]Br and THF, then characterised. The
isolated carbene complex was then used in a subsequent
Table 3
Catalyst activation studies performed on the addition of Bu3SnSiMe3
to phenylacetylene in the presence of 5 mol% Pd(PPh3)4 under various
conditions, heating in a 70 ◦C oil bath
Entry
Ionic liquid
Reagents present
during activation
Reaction
time (h)
Isolated
yield (%)
1
2
3a
4
5
6a
[bmim]PF6
[bmim]PF6
[bmim]PF6
[bmim]PF6
[bmim]BF4
[bmim]BF4
None
Et2O
Et2O
Bu3SnSiMe3, Et2O
Bu3SnSiMe3, Et2O
Bu3SnSiMe3, Et2O
120
20
16
19
20
16
97
100
100
100
92
97
Reaction conditions: 1.0 mmol phenylacetylene, 1.2 mmol Bu3SnSiMe3,
1.0 ml ionic liquid, 5.0 ml Et2O, 70 ◦C oil bath, 20 h activation period.
a
Second cycle with same ionic liquid/catalyst system.

I. Hemeon, R.D. Singer / Journal of Molecular Catalysis A: Chemical 214 (2004) 33–44
43
Heck reaction and shown to be highly active. The carbene
complex was not observed in [bmim]BF4, indicating that a
halide was required to deprotonate imidazolium. Mathews
et al. [42] also observed and isolated a palladium–carbene
complex in a Suzuki reaction in [bmim]BF4 but added a
halide source such as NaCl to induce its formation. Indeed,
it has been shown that in Heck reactions performed in
non-imidazolium-based ionic liquids palladium black depo-
sition slowly decreases catalyst activity, inhibiting its recy-
clability [28]. In fact, density functional analysis has shown
that oxidative C–H addition of the proton–carbon 2 bond of
an imidazolium cation across closely related platinum(0) is
exothermic, supported experimentally by the reaction of the
1,3-dimethylimidazolium cation with Pt(PPh3)4 to produce
the carbene complex [43]. Hamill et al. [44] have per-
formed investigations using X-ray absorption fine structure
(XAFS) to determine the speciation of the palladium cata-
lyst during a Heck reaction in several ionic liquids. These
studies showed the formation of a bis-carbene–palladium
complex in N,N-dialkylimidazolium chloride ionic liquids
while in N,N-dialkylimidazolium tetrafluoroborate and hex-
afluorophosphate ionic liquids palladium nanoclusters were
observed. The addition of triphenylphosphine to the sys-
tems increased the induction time required for nanocluster
formation but showed that they were more stable and resis-
tant to palladium black deposition in ionic liquids. Desh-
mukh et al. [45] claimed to have isolated a Pd–bis-carbene
complex from both 1,3-di-n-butylimidazolium bromide,
[bmim]Br and [bmim]BF4. Through transmission electron
microscopy they observed the conversion of this complex
to palladium nanoparticles using sonication. Isolated pal-
ladium nanoclusters have been shown to be active in the
Heck reaction. Reactions performed in our studies also
required an induction time and were similar in colour to
reactions in which palladium-imidazolylidene complexes
and/or palladium nanoparticles were observed. It therefore
seems likely that the active palladium species in the silyl-
stannation of alkynes in ionic liquids is not Pd(PPh3)4, a
possibility requiring further study.
alkyl(trialkylstannyl)silanes and five alkynes containing
differing functionalities were successfully coupled in good
to excellent isolated yields. Although acceptable yields
were obtained from both ionic liquids, it appears as though
the reactions proceed slightly faster in [bmim]PF6 than
they do in [bmim]BF4, perhaps due to catalyst inactivation
resulting from higher water content of [bmim]BF4 or better
stabilisation of the active catalytic palladium species by
[bmim]PF6. All reactions proceeded with complete stereo-
and regioselectivity. These results add to the growing reper-
toire of organometallic reactions that can be performed in
ionic liquids in which the organometallic reagent is not part
of a transition metal complex.
Perhaps more significantly, we have also shown that the
ionic liquid/catalyst system is extensively recyclable, being
reused 10 times without loss of activity in both ionic liquids.
The requirement of an induction time for optimal catalyst
activity, coupled with the colours of the reactions, seems to
suggest that Pd(PPh3)4 is not the active catalytic species.
Based on our results and previous studies, it seems likely
that either a palladium–imidazolylidene carbene complex or
palladium metal nanoclusters are the active species. This re-
quires further study through NMR and/or XAFS analysis
of the reaction during its course. We also intend to inves-
tigate tandem reactions in our systems, performing further
reactions with the silylstannylated alkenes in the same re-
action vessel after their formation. Such reactions include
palladium-catalysed Stille couplings with aryl halides and
protodesilylations.
Acknowledgements
We wish to thank the Natural Sciences and Engineering
Research Council of Canada (PGS-A to I. Hemeon and Dis-
covery Grant to R.D. Singer) and Saint Mary’s University,
Faculty of Graduate Studies and Research. NMR spectra
were recorded at the Atlantic Regional Magnetic Resonance
Centre.
The exact nature of the active catalytic species in the
silylstannation of alkynes in imidazolium based ionic liquids
is topic of a current study in our laboratory and will reported
in due course. An aspect of this study involves analysis of
catalytic leaching into the organic extract using ICP-MS.
Furthermore, we have used ICP-MS to determine residual
halide content in various ionic liquids, including [bmim]BF4
and [bmim]PF6, prepared via various methods [25]. This
study will give insight into the effect of halide ions (residual
or added) to the systems described here.
References
[1] J.G. Huddleston, A.E. Visser, W.M. Reichert, H.D. Willauer, G.A.
Broker, R.D. Rogers, Green Chem. 3 (2001) 156.
[2] S.N. Baker, G.A. Baker, F.V. Bright, Green Chem. 4 (2002) 165.
[3] S.V. Dzyuba, R.A. Bartsch, Tetrahedron Lett. 43 (2002) 4657.
[4] H. Zhao, S.V. Malhotra, Aldrichim. Acta 35 (2002) 75.
[5] H. Olivier-Bourbigou, L. Magna, J. Mol. Catal. A: Chem. (2002)
419.
[6] T. Welton, Chem. Rev. 99 (1999) 2071.
[7] I. Hemeon, S. Dickson, R.D. Singer, Recent Res. Dev. Org. Chem.
6 (2002) 499.
[8] J. Dupont, R.F. de Souza, P.A.Z. Suarez, Chem. Rev. 102 (2002)
3667.
4. Conclusions
[9] D. Zhao, M. Wu, Y. Kou, E. Min, Catal. Today 74 (2002) 157.
[10] R. Sheldon, Chem. Commun. (2001) 2399.
[11] P. Wasserscheid, W. Keim, Angew. Chem. Int. Ed. Engl. 39 (2000)
3772.
We have shown that it is possible to perform addi-
tion reactions between trialkyl(trialkylstannyl)silanes and
terminal alkynes in two different ionic liquids. Two tri-

44
I. Hemeon, R.D. Singer / Journal of Molecular Catalysis A: Chemical 214 (2004) 33–44
[12] I. Hemeon, R.D. Singer, in: I. Fleming (Ed.), Houben-Weyl Methods
of Molecular Transformations, Science of Synthesis, vol. 4, Georg
Theime Verlag, New York, 2002, Chapter 4.4.6, p. 205.
[13] B.L. Chenard, C.M. van Zyl, J. Org. Chem. 51 (1986) 3561.
[14] T.N. Mitchell, R. Wickenkamp, A. Amamria, R. Dicke, U. Schneider,
J. Org. Chem. 52 (1987) 4868.
[31] I. Kmentova, B. Gotov, E. Solcaniova, S. Toma, Green Chem. 4
(2002) 103.
[32] S. Toma, B. Gotov, I. Kmentova, E. Solcaniova, Green Chem. 2
(2000) 149.
[33] H. Hagiwara, Y. Shimizu, T. Hoshi, T. Suzuki, M. Ando, K. Ohkubo,
C. Yokoyama, Tetrahedron Lett. 42 (2001) 4349.
[34] R. Rajagopal, D.V. Jarikote, K.V. Srinivasan, Chem. Commun. (2002)
616.
[35] A.J. Carmichael, M.J. Earle, J.D. Holbrey, P.B. McCormac, K.R.
Seddon, Org. Lett. 1 (1999) 997.
[36] J.D. Revell, A. Ganesan, Org. Lett. 4 (2002) 3071.
[37] K. Okubo, M. Shirai, C. Yokoyama, Tetrahedron Lett. 43 (2002)
7115.
[38] W. Chen, L. Xu, C. Chatterton, J. Xiao, Chem. Commun. (1999)
1247.
[39] W.A. Herrmann, Angew. Chem. Int. Ed. Engl. 41 (2002) 1291.
[40] I.P. Beletskaya, A.V. Cheprakov, Chem. Rev. 100 (2000) 3009.
[41] L. Xu, W. Chen, J. Xiao, Organometallics 19 (2000) 1123.
[42] C.J. Mathews, P.J. Smith, T. Welton, A.J.P. White, D.J. Williams,
Organometallics 20 (2001) 3848.
[15] K. Ritter, Synthesis (1989) 218.
[16] I. Beletskaya, C. Moberg, Chem. Rev. 99 (1999) 3435.
[17] M. Suginome, Y. Ito, Chem. Rev. 100 (2000) 3221.
[18] I. Hemeon, R.D. Singer, Chem. Commun. (2002) 1884.
[19] B.M. Trost, Science 254 (1991) 1471.
[20] S.T. Handy, X. Zhang, Org. Lett. 3 (2001) 233.
[21] J.G. Huddleston, H.D. Willauer, R.P. Swatloski, A.E. Visser, R.D.
Rogers, Chem. Commun. (1998) 1765.
[22] J.B. Armitage, M.C. Whiting, J. Chem. Soc. (1952) 2005.
[23] K.F. Bernade, M.B. Floyd, J.F. Poletto, M.J. Weiss, J. Org. Chem.
44 (1979) 1438.
[24] K.R. Seddon, A. Stark, M. Torres, Pure Appl. Chem. 72 (2000) 2275.
[25] K. McCamley, N.A. Warner, M. Lamoureux, P.J. Scammells, R.D.
Singer, unpublished data.
[26] A.G. Davies, Organotin Chemistry, VCH Verlagsgesellschaft, Wein-
heim, 1997, pp. 1–27.
[43] D.S. McGuinness, K.J. Cavell, B.F. Yates, Chem. Commun. (2001)
355.
[27] C.J. Mathews, P.J. Smith, T. Welton, Chem. Commun. (2000) 1249.
[28] V.P.W. Bohm, W.A. Herrmann, Chem. Eur. J. 6 (2000) 1017.
[29] J. Howarth, A. Dallas, Molecules 5 (2000) 851.
[30] M. Hada, Y. Tanaka, M. Ito, M. Murakami, H. Amii, Y. Ito, H.
Nakatsuji, J. Am. Chem. Soc. 116 (1994) 8754.
[44] N.A. Hamill, C. Hardacre, S.E.J. McMath, Green Chem. 4 (2002)
139.
[45] R.R. Deshmukh, R. Rajagopal, K.V. Srinivasan, Chem. Commun.
(2001) 1544.