Trimethylaluminum (TMA)-catalyzed reaction of alkynyllithiums with ethylene oxide: Increased yields and purity of homopropargylic alcohols

Article abstract of DOI:10.1055/s-2005-917078

An efficient protocol was developed to obtain homopropargylic alcohols. Subjecting alkynyllithiums and ethylene oxide to 10-20 mol% of trimethylaluminum provided homopropargylic alcohols in good yields. Georg Thieme Verlag Stuttgart.

Full text of DOI:10.1055/s-2005-917078

LETTER
2457
Trimethylaluminum (TMA)-Catalyzed Reaction of Alkynyllithiums with
Ethylene Oxide: Increased Yields and Purity of Homopropargylic Alcohols
T
K
rimethylaluminum
a
-Catalyze
y
dReactionof
Alky
M
nyllithiums . Brummond,* Jamie M. McCabe
Department of Chemistry, University of Pittsburgh, Pittsburgh, PA, 15260, USA
Fax +1(412)6248611; E-mail: kbrummon@pitt.edu; E-mail: jmm210@pitt.edu
Received 29 June 2005
There are many reports of Lewis acids catalyzing the
Abstract: An efficient protocol was developed to obtain homo-
propargylic alcohols. Subjecting alkynyllithiums and ethylene
oxide to 10–20 mol% of trimethylaluminum provided homopro-
pargylic alcohols in good yields.
addition of acetylides to epoxides, the most common of
2
which was developed by Yamaguchi and Hirao using a
stoichiometric amount of boron trifluoride etherate. In our
hands this protocol was not only technically challenging
upon scale up, but inseparable by-products were formed.
Another reagent used to facilitate the reaction between ep-
oxides and alkynyllithiums is catalytic amounts of tri-
Key words: catalysis, homopropargylic alcohols, alkynyllithiums,
nucleophilic additions, ethylene oxide, trimethylaluminum
3
methylgallium. This methodology appears to be a very
Successful completion of a synthesis of a natural product
possessing moderate complexity typically requires the
preparation of starting materials in gram quantities. The
syntheses of these relatively simple and low molecular
weight precursors are usually the least interesting in a syn-
efficient way to form homopropargylic alcohols; howev-
er, it seems not to be a very well-liked protocol. It is
thought that the cost ($13 per gram) of Me Ga and special
3
handling needs detract from its usefulness.
thetic sequence that may ultimately lead to an interesting Stoichiometric amounts of trimethylaluminium (TMA)
bioactive compound via a novel carbon–carbon bond have been reported to promote reaction with epoxides to
forming reaction. However, low yields in the beginning give good to moderate yields; however, in some cases
stages of a synthesis can foil the best of synthetic strate- TMA was not reactive enough and needed to be used in
4
gies if key step precursors cannot be accessed in sufficient conjunction with stoichiometric amounts of BF ·OEt . In
3
2
quantities.
one publication, catalytic TMA could be used only if the
acetylide was reacted with an epoxy ether, not a simple
epoxide like 5-phenyl-1-pentene oxide. There were also
Recently, while working towards the synthesis of guana-
castepene A, we desired the homopropargylic alcohol 3 in
5
reports of the use of diethylaluminum chloride by Fried,6
which gave good yields for the addition of the alkynyllith-
ium of octyne, but low yields for more complex alkynes.
Herein, we report the use of catalytic TMA for the forma-
tion of a variety of homopropargylic alcohols in good
yields.
gram quantities. We obtained this alcohol by addition of
ethylene oxide (2) to lithiated 3-methyl-1-butyne (1);1
however, the yields of this reaction suddenly went from
9
5% to 10%. After some experimentation, the irreproduc-
ibility of this reaction was traced to the ethylene oxide. It
was observed that an older expired lecture bottle of ethyl-
ene oxide gave 95% yields of alkynol 3. Since the older Addition of catalytic amounts of TMA to lithiumiso-
bottle had been discarded, we could only speculate that a propylacetylide followed by addition of ethylene oxide
7
contaminant was responsible for the higher yields. Based gave a 95% yield of 3 (entry 3, Table 1); 20 mol% of
upon our need for a reliable and cheap protocol for the TMA (Conditions C) gave the best yield based upon en-
preparation of gram quantities of homopropargylic alco- tries 1–4, Table 1. With the success of this first substrate,
hol 3, we initiated a systematic investigation into the the scope and limitations of this new methodology was
addition reaction of acetylides to ethylene oxide explored and the results are displayed in Table 1.
(
Equation 1).
Other alkyl acetylides were investigated. Addition of the
lithium anion of 1-octyne (4) gave 3-decyn-1-ol (5) in
64% or 69% yield depending upon the amount of TMA
used (entries 5 and 6). In this example, a direct compari-
O
+
n-BuLi
OH
1
2
3
son to the Me Ga protocol was possible. Trimethylgalli-
3
um gave a 92% yield of the 3-decyn-1-ol and a 10% yield
in the absence of any Lewis acid. However, the extreme
Equation 1
3
pyrophoric nature of this catalyst renders this reagent less
user friendly than TMA. Addition of the lithium anion of
propyne gave a 38% yield of 3-pentyn-1-ol (7, entry 7).
This low yield was attributed to the in situ formation of
propyne anion from 1-bromopropene, in which one equiva-
lent of LiBr is formed and adventitious base is present.
SYNLETT 2005, No. 16, pp 2457–2460
0
4
.1
0
.2
0
0
5
Advanced online publication: 21.09.2005
DOI: 10.1055/s-2005-917078; Art ID: S06305ST
Georg Thieme Verlag Stuttgart · New York
©

2
458
K. M. Brummond, J. M. McCabe
LETTER
Table 1 (Me) Al-Catalyzed Ring-Opening of Epoxidesa
3
conditions
O
R
R
Li
+
OH
Product
3
Entry
1
Alkynyllithium, R = Substrate
Conditionsb
A
Yield (%)
53
Yield brsmg
1
2
3
4
5
6
7
8
9
1
1
1
3
3
B
C
Cc
B
C
A
A
C
A
B
C
A
C
Ae
A
58
95
3
71
C H
4
4
5
64
6
13
5
69
CH3
6
7
38d
748b
80
TMS
8
9
8
9
1
1
1
1
1
1
1
0
1
2
3
4
5
6
TIPS
10
10
10
12
12
12
14
11
11
11
13
13
13
15
56
71
90
80
77
79
70
Ph
6613
67
59
67
1
1
1
7
8
9
14
14
16
15
15
19
B
C
C
63
68
55f
67
71
EtO
a
b
c
d
e
f
Experiments were performed on a 200 mg scale in THF; reaction temperature 0 °C to r.t.
Conditions A: 10 mol% Me Al, 24 h. Conditions B: 10 mol% Me Al, 48 h. Conditions C: 20 mol% Me Al, 24 h.
3
3
3
Experiment was performed on 5 g scale.
GC yield.
Experiment apparatus included internal reflux condenser (see text).
Yield of TBDPS-protected product.
g
Brsm: based upon recovered starting material.
Lithiated silyl acetylides underwent the catalyzed
addition to ethylene oxide giving high yields of homo-
propargylic alcohols. Trimethylsilylacetylene upon de-
protonation reacted with ethylene oxide to give 4-
trimethylsilyl-3-butyn-1-ol (9) in 74% and 80% yields
O
H+
EtO
OH
OH
EtO
17
18
EtO
imidazole
TBPDSCl
OTBDPS
(
entries 8 and 9). Substrate 9 is typically produced by silyl
protection of 3-butyn-1-ol. While the latter protocol gives
19
8
approximately the same yields (71–94%) as the former it Scheme 1
is not applicable for the formation of other homopropar-
gylic alcohols with more robust silyl groups. For example,
Investigation into the reaction of aryl acetylenes with eth-
ylene oxide also gave good results. Subjection of phenyl
acetylene to the catalytic reaction conditions gave 59–
7% yields of aryl acetylene 13 depending on the reaction
conditions (entries 13–15). Lithiated 1-ethynyl-cyclohex-
triisopropylsilyl-3-butyn-1-ol (11) can be obtained in a
1% yield (entries 10–12) under these TMA catalytic con-
ditions whereas previously it could only be acquired in a
2% yield via an alkynylation with triisopropylsilyl-
7
6
6
acetylenic triflone under a photochemical irradiation
ene (14) underwent smooth conversion to the enyne 15 in
3–68% yields (entries 16–18). This is an alternative to
9
procedure.
6
Synlett 2005, No. 16, 2457–2460 © Thieme Stuttgart · New York

LETTER
Trimethylaluminum-Catalyzed Reaction of Alkynyllithiums
2459
using palladium coupling procedures previously em- graphy eluting with 10% Et
ployed for preparing substrate 15 and opens the door to
O–pentane to afford 312 mg of alcohol
2
1
0
3 in 95% yield.
-Methylhex-3-yn-1-ol (3)
make a variety of homopropargylic enynes.
5
1
This protocol is even compatible with ethyl ethynyl ether,
H NMR (CDCl ): d = 1.16 (d, J = 6.8 Hz, 6 H), 1.75 (br s, 1 H),
3
which gave moderate yields (55%) after subsequent silyl 2.44 (dt, J = 2.2 Hz, J = 6.2 Hz, 2 H), 2.55 (tsep, J = 2.2 Hz, J = 6.8
1
3
Hz, 1 H), 3.68 (t, J = 6.2 Hz, 2 H). C NMR (CDCl ): d = 87.9,
protection of the alcohol (entry 19). The protection was
done for two reasons; first, if the free alcohol was
chromatographed or worked up with dilute acid complete
formation of ester 18 was seen and second, this gave us
the ability to compare our results with past literature find-
ings (Scheme 1). Previously, compound 19 was obtained
in a 28% yield over two steps by using LiNH and NH
3
7
5.4, 61.2, 23.1, 22.8, 20.3. IR (neat): 3346, 2969, 2876, 1465, 1041
–
1
cm .
4
-Trimethylsilyl-3-butyn-1-ol (9)
All spectroscopic data obtained matched literature findings.
4-Triisopropyl-3-butyn-1-ol (11)
2
3
1
1
1
H NMR (CDCl ): d = 1.09–1.06 (m, 21 H), 1.84 (br s, 1 H), 2.54 (t,
3
13
and the same silylation protocol.
A systematic study of the addition of ethylene oxide to
alkynyllithiums was undertaken looking at catalyst load- 2866, 2174, 1464 cm . MS (GCMS): m/e (relative intensity) = 226
J = 6.2 Hz, 2 H), 3.73 (t, J = 6.2 Hz, 2 H). C NMR (CDCl ): d =
3
1
04.9, 83.0, 61.2, 24.3, 18.6, 11.2. IR (neat): 3334, 2943, 2891,
–
1
+
ing, reaction times, and reaction apparatus. Based upon (10) [M] , 183 (98), 59 (100). HRMS: m/z calcd for C10
H19OSi:
+
+
1
83.1205 [M – C H ] ; found: 183.1213 [M – C H ] .
3 7 3 7
the entries in Table 1, conditions C produced the highest
yields for most substrates. However, for many of these
substrates the yield is only marginally higher than the
yields observed when half the amount of TMA was em-
ployed (conditions A). As expected, the increase in cata-
lyst loading caused the reaction times to be shorter,
4
-Phenyl-3-butyn-1-ol (13)
All spectroscopic data obtained matched literature findings.
4
-Cyclohexenyl-3-butyn-1-ol (15)
1
H NMR (CDCl ): d = 1.64–1.42 (m, 4 H), 2.05–1.92 (m, 4 H), 2.45
3
usually from 48 hours to 24 hours (conditions C). Certain (t, J = 6.6 Hz, 2 H), 3.13 (br s, 1 H), 3.59 (t, J = 6.6 Hz, 2 H), 5.97–
1
3
5
2
.90 (m, 1 H). C NMR (CDCl ): d = 133.6, 120.4, 83.6, 83.2, 60.9,
examples were hard to follow due to the volatility or insta-
bility of the starting materials (1, 6, 8, 16); therefore, all
reactions were carried out in 24 hours increments for con-
sistency. For most of the examples where the starting ma-
terials were not volatile (4, 10, 12, 14), the reaction could
not be pushed to completion even with prolonged reaction
3
9.2, 25.3, 23.2, 22.1, 21.3. IR (neat): 3346, 2930, 2858, 1436, 1043
cm . MS (GCMS): m/e (relative intensity) = 150 (50) [M] , 91
–
1
+
(
100). HRMS: m/z calcd for C H O: 150.1045; found: 150.1046.
10 14
3
-Decyn-1-ol (5)
1
H NMR (CDCl ): d = 0.89 (t, J = 7.1 Hz, 3 H), 1.54–1.21 (m, 8 H),
3
times; therefore, starting material was recovered. It was 1.87 (br s, 1 H), 2.20–2.12 (m, 2 H), 2.48–2.39 (m, 2 H), 3.68 (t,
1
3
J = 6.1 Hz, 2 H). C NMR (CDCl ): d = 82.5, 76.3, 61.3, 31.3, 28.9,
thought that maybe the ethylene oxide, having a boiling
point of 10 °C, was evaporating before completion of the
reaction. To prevent this from happening, an internal re-
flux condenser was used with the salt ice water running
3
–
1
28.5, 23.1, 22.5, 18.7, 13.9. IR (neat): 3349, 2930, 2858, 1045 cm .
+
MS (GCMS): m/e (relative intensity) = 154 (1) [M] , 97 (35), 55
100).
(
through it; however, the yields were not affected (entry 3-Pentyn-1-ol (7)
2).
1
GC analysis data matched commercially available material.
In conclusion, we have developed a TMA-catalyzed addi-
tion of simple alkynes to ethylene oxide forming a diverse
variety of homopropargylic alcohols. This protocol elimi-
4
-Ethoxybut-3-ynyloxy-tert-butyl-diphenylsilane (19)
To a solution of alkyne 16 (0.68 mL of a 40% by weight solution in
hexane, 2.85 mmol) in 3 mL of THF at –78 °C was added n-BuLi
nates the undesired by-products that were found when bo- (2.14 mL of a 1.6 M solution in hexane, 3.42 mmol) dropwise. After
ron trifluoride etherate was used and allows for an 45 min at –78 °C the flask was placed in an 0 °C ice bath for 15 min
then TMA (0.29 mL of a 2 M solution in hexane, 0.57 mmol) was
inexpensive alternative to obtaining large amounts of ho-
1
2
added followed by ethylene oxide (0.17 mL, 3.42 mmol). The ice
bath was removed and the reaction flask was kept at r.t. for 24–48 h
after which it was quenched by the addition of phosphate buffer (pH
mopropargylic alcohols.
7
.38, 0.5 mL), Rochelle’s salt sat. solution (2 mL), and Et O (10
2
Preparation of Homopropargylic Alcohols
To a solution of alkyne (1, 0.20 g, 2.94 mmol) in 3 mL of THF at
mL), consecutively. The organic layer was removed and the aque-
ous layer was extracted with Et O. The combined organics were
dried with MgSO , filtered, and concentrated under reduced pres-
sure. The crude mixture was taken up in 4 mL of CH Cl and imi-
dazole (0.39 g, 5.70 mmol), and then TBDPSCl (0.82 mL, 3.14
mmol) was added. After 1.5 h the reaction was diluted with Et
and washed sequentially with H O and NaHCO . The organic layers
were dried with MgSO , filtered, and concentrated under reduced
2
–
3
78 °C was added n-BuLi (2.21 mL of a 1.6 M solution in hexane,
.53 mmol) dropwise. After 45 min at –78 °C the flask was placed
4
2
2
in an 0 °C ice bath for 15 min then TMA (0.29 mL of a 2 M solution
in hexane, 0.59 mmol) was added followed by ethylene oxide (0.18
mL, 3.53 mmol).12 The ice bath was removed and the reaction flask
was left at r.t. for 24–48 h after which it was quenched by the addi-
tion of H O (1 mL) and Et O (10 mL). The bi-phasic mixture was
O
2
2
3
4
pressure. The residue was purified by silica gel chromatography
eluting with 1–5% EtOAc–hexane to afford 553 mg of 19 in a 55%
yield for the two steps.
2
2
transferred to a separatory funnel and 10% HCl was added until it
eliminated the aluminum emulsion. The organic layer was removed
and the aqueous layer was extracted with Et O. The combined
1
2
H NMR (CDCl ): d = 1.10 (s, 9 H), 1.35 (t, J = 7.1 Hz, 3 H, 2.44
3
organics were dried with MgSO , filtered, and concentrated under
4
(t, J = 7.3 Hz, 2 H), 3.76 (t, J = 7.2 Hz, 2 H), 4.02 (q, J = 7.1 Hz, 2
H), 7.50–7.36 (m, 6 H), 7.77–7.70 (m, 4 H). C NMR (CDCl ): d =
reduced pressure. The residue was purified by silica gel chromato-
13
3
Synlett 2005, No. 16, 2457–2460 © Thieme Stuttgart · New York

2
460
K. M. Brummond, J. M. McCabe
LETTER
1
1
35.6, 133.9, 129.5, 127.6, 90.1, 73.9, 63.6, 34.2, 26.8, 21.4, 19.2,
(4) (a) Ooi, T.; Kagoshima, N.; Maruoka, K. J. Am. Chem. Soc.
1997, 119, 5754. (b) Asao, N.; Kasahara, T.; Yamamoto, Y.
Tetrahedron Lett. 2001, 42, 7903. (c) Zhao, H.; Pagenkopf,
B. L. Chem. Commun. 2003, 2592. (d) Trost, B. M.; Yang,
H.; Probst, G. D. J. Am. Chem. Soc. 2004, 126, 48.
–
1
4.3. IR (neat): 3070, 2958, 2858, 2273, 1428, 1112 cm . MS (GC-
+
MS): m/e (relative intensity) = 352 (5) [M] , 295 (98), 199 (100).
HRMS: m/z calcd for C H O Si: 295.1154 [M – C H ] ; found:
2
+
1
8
19
2
4
9
+
95.1155 [M – C H ] .
4 9
(5) Ooi, T.; Kagoshima, N.; Ichikawa, H.; Maruoka, K. J. Am.
Ethyl 4-Hydroxybutanoate (18)
Chem. Soc. 1999, 121, 3328.
1
H NMR (CDCl ): d = 1.26 (t, J = 7.1 Hz, 3 H), 1.89 (quin, J = 6.6
(6) (a) Fried, J.; Lin, C.; Ford, S. H. Tetrahedron Lett. 1969, 10,
1379. (b) Sasaki, M.; Miyazawa, M.; Tanino, K.; Miyashita,
M. Tetrahedron Lett. 1999, 40, 9267.
3
Hz, 2 H), 2.43 (t, J = 7.1 Hz, 2 H), 3.69 (t, J = 6.1 Hz, 2 H), 4.14
1
3
(
3
quin, J = 7.1 Hz, 2 H). C NMR (CDCl ): d = 174.0, 62.1, 60.5,
3
1.1, 27.7, 14.2.
(7) Without the addition of Me Al the yields were
3
approximately 10% and using stoichiometric BF ·OEt gave
3
2
5
4% yield of a mixture of inseparable by-products.
Acknowledgment
(
8) (a) Foote, K. M.; John, M.; Pattenden, G. Synlett 2001, 365.
b) Overman, L. E.; Brown, M. J.; McCann, S. F. Org. Synth.
990, 68, 182. (c) Pattenden, G.; Teague, S. J. J. Chem. Soc.,
(
1
We thank the University of Pittsburgh, the National Institutes of
Health (GM54161), and the Johnson & Johnson Focused Giving
Award for financial support of this research. J.M.M. thanks Pfizer
for the 2004 Pfizer Research Fellowship Supporting Diversity in
Organic Chemistry.
Perkin Trans. 1 1988, 5, 1077.
(
9) Xiang, J.; Fuchs, P. L. Tetrahedron Lett. 1998, 39, 8597.
(
(
(
10) Alami, M.; Ferri, F.; Linstrumelle, G. Tetrahedron Lett.
993, 34, 6403.
11) Overman, L. E.; Thompson, A. S. J. Am. Chem. Soc. 1988,
10, 2248.
1
1
References
12) The ethylene oxide was condensed in a separate two-neck
pear shaped flask equipped with a coldfinger. Then, the
ethylene oxide was removed with a plastic syringe that was
pre-cooled by capping the needle with a large septa and
placing it in a Ziploc bag, which was then placed in the
freezer for 20 min prior to use. If larger amounts (>1 mL) of
ethylene oxide were needed the desired amount was
condensed into an addition funnel with a cooling jacket and
added dropwise to the reaction.
(
(
1) Brummond, K. M.; Gao, D. Org. Lett. 2003, 5, 3491.
2) (a) Yamaguchi, M.; Hirao, I. Tetrahedron Lett. 1983, 24,
391. (b) Yamaguchi, M.; Nobayashi, Y.; Hirao, I.
Tetrahedron 1984, 40, 4261.
(
3) (a) Utimoto, K.; Lambert, C.; Fukuda, Y.; Shiragami, H.;
Nozaki, H. Tetrahedron Lett. 1984, 25, 5423. (b) Fukuda,
Y.; Matsubara, S.; Lambert, C.; Shiragami, H.; Nanko, T.;
Utimoto, K.; Nozaki, H. Bull. Chem. Soc. Jpn. 1991, 64,
1810.
(13) Zhang, H.; Larock, R. C. J. Org. Chem. 2003, 68, 5132.
Synlett 2005, No. 16, 2457–2460 © Thieme Stuttgart · New York