TMSCl-mediated catalytic carbocupration of alkynoates: An unprecedented and remarkable effect of catalyst loading on highly selective stereochemical induction via a TMS-allenoate intermediate

Article abstract of DOI:10.1002/ejoc.200400314

The TMSCl-mediated catalytic carbocupration of alkynoates has been investigated. It has been shown that catalyst loadings as low as 30 mol% readily allow for high yields and diastereoselectivities for a series of Grignard reagents. In addition, an unprecedented and remarkable effect of catalyst loading on stereochemical induction has been observed. ( Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim, Germany, 2004).

Full text of DOI:10.1002/ejoc.200400314

SHORT COMMUNICATION
TMSCl-Mediated Catalytic Carbocupration of Alkynoates: An Unprecedented
and Remarkable Effect of Catalyst Loading on Highly Selective
Stereochemical Induction via a TMS-Allenoate Intermediate
Michael P. Jennings*[a] and Kailas B. Sawant[a]
Keywords: Carbocupration / Cuprates / Grignard reaction / Propiolates / TMS-allenoate
The TMSCl-mediated catalytic carbocupration of alkynoates
has been investigated. It has been shown that catalyst load-
ings as low as 30 mol% readily allow for high yields and dia-
stereoselectivities for a series of Grignard reagents. In addi-
tion, an unprecedented and remarkable effect of catalyst
loading on stereochemical induction has been observed.
( Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2004)
Introduction
the crude products reveals only 30% of the desired products
in addition to decomposition of the starting material. After
column chromatography, a combined yield of 28% was iso-
lated. Performing the carbocupration at Ϫ35, Ϫ50, and
Ϫ78 °C in THF had little or no effect on the cuprate turn-
over, diastereoselectivity, and yield of 2a and 2b. The yields
remained under 30% and the selectivity for the E-isomer
remained at 9:1 and slightly better in some instances. An
interesting trend was observed when these conditions were
examined in DME as the solvent. The Z-isomer selectivity
gradually increased as the reaction temperature was de-
creased, as shown in Table 1.
Over the past 35 years, considerable attention has been
given to the stoichiometric carbocupration of α,β-acetylenic
esters with organocopper reagents resulting from lithium-
derived organometallics for the synthesis of substituted ole-
fins.[1] The stereoselectivity of these net 1,4-additions are
dependent on a variety of conditions such as temperature,
copper reagent, solvent effects, and external chelating li-
gands.[1] Interestingly, very little is reported on the utili-
zation of Grignard reagents for these types of reactions.[2]
Based on the report on modified Kharash reaction con-
ditions utilizing CuI·2LiCl by Reetz and Kindler,[3] we de-
cided to investigate the possibility of using catalytic
amounts of a combination copper()-lithium halide salt in
conjunction with Grignard reagents for the carbocupration
of α,β-acetylenic esters. Herein we wish to report, to the
best of our knowledge, the first highly diastereoselective
carbocupration of any propiolate ester that uses only cata-
lytic amounts of a copper() salt. Additionally, an unpre-
cedented and remarkable effect of catalyst loading on
stereochemical induction has been observed.
Table 1. CuI·2LiCl Promoted carbocupration of 1a with PhMgBr
Entry[a]
Temperature [°C]
Solvent
Yield [%]
E[b]
Z
1
2
0
Ϫ35
Ϫ50
Ϫ50
Ϫ78
0
THF
THF
THF
THF
THF
DME
DME
DME
28
29
27
85
26
27
29
24
91
92
93
78
91
80
53
46
9
8
7
22
9
20
47
54
3
4[c]
5
6
7
8
Ϫ35
Ϫ78
[a]
Results and Discussion
All reactions ran with 60% LiCl, 30% CuI, and 1.2 equiv. of
PhMgBr added in reverse, unless otherwise noted. [b] E/Z ratio de-
termined by H NMR spectroscopy (360 MHz) of the crude reac-
tion mixture. 1.3 equiv. TMSCl added.
1
As shown in Equation (1), reaction of 1a at 0 °C in THF
with 30 mol% of CuI·2LiCl and 1.2 equivalents of PhMgBr
furnished, after quenching with NH4Cl and a standard
workup, a crude product mixture of 2a and 2b as a dia-
stereomeric ratio of 91:9 in favor of the E-isomer. In spite
[c]
1
of the high level of selectivity, the H NMR spectrum of
(1)
[a]
Department of Chemistry, The University of Alabama,
Tuscaloosa, AL 35487Ϫ0336, USA
Fax (internat.): ϩ1-205-348-9104
E-mail: jenningm@bama.ua.edu
Eur. J. Org. Chem. 2004, 3201Ϫ3204
DOI: 10.1002/ejoc.200400314
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3201

M. P. Jennings, K. B. Sawant
SHORT COMMUNICATION
Based on the catalytic activities of CuI salts toward pro- Table 2), a slightly higher yield and a dramatic increase in
piolate esters reported by Munch-Petersen, Klein, and Ja- diastereoselectivity from 3:1 to 9:1 for the E-isomer 2a was
lander, these preliminary results were discouraging.[2b,2d,2e] observed. It is also worth noting that if the reaction was
It appeared that the catalyst turnover was not proceeding carried out at Ϫ78 °C and then quenched at 0 °C, the iso-
due to the stability of the intermediate vinylcuprate 3 meric ratio decreased to 3:1, indicating that both the reac-
even at 0 °C.[2b] Thus, no isomerization to the magnesium tion and protic quench must be executed at a low tempera-
allenoate 4 had occurred, therefore not allowing for the ture for maximum diastereoselectivity.
release of the cuprate reagent into a catalytic cycle (Fig-
ure 1). The stability of 3 helped to explain the high levels
Table 2. TMSCl-CuI·2LiCl catalyzed carbocupration of 1a with
of stereoselectivity even at 0 °C, and contrasts that of lith-
PhMgBr in THF
ium-based organocuprates.[1]
Entry Temperature [°C] Catalyst
mol% Yield [%] E[a]
Z
1
2
0
Ϫ35
Ϫ50
Ϫ78
Ϫ78
Ϫ78
Ϫ78
Ϫ78
Ϫ78
CuI·2LiCl 30
CuI·2LiCl 30
CuI·2LiCl 30
CuI·2LiCl 30
CuI·2LiCl 10
CuI·2LiCl 20
CuI·2LiCl 30
CuI·2LiCl 40
CuI·2LiCl 50
28
72
85
75
44
85
87
90
91
91
9
64 36
78 22
76 24
43 57
70 30
90 10
3
4[b]
5
6
7
8
9
92
95
8
5
[a]
E/Z ratio determined by 1H NMR spectroscopy (360 MHz) of
Figure 1. TMS-mediated isomerization of 3Ϫ5
[b]
the crude reaction mixture.
quenched with NH4Cl.
Reaction warmed to 0 °C, then
Based on previous observations that Lewis acid additives
such as TMSCl and BF3·Et2O accelerate conjugate addition
Secondly, the activity of the CuI·2LiCl catalyst was
of cuprates to Michael acceptors,[4] it was reasoned that examined and the results are shown in Table 2, Entries 5Ϫ9.
TMSCl might induce an isomerization of the vinylcuprate From the data, it can be seen that the catalytic system
3 to form the TMS-allenoate (5) and subsequently permit turned over approximately four times. Thus, when the reac-
the release of the bound cuprate.[5] Thus, the addition of tion sequence was performed with only 10 mol% of catalyst,
TMSCl might allow for a carbocupration that utilizes only a 44% combined yield was isolated, and similarly with 20
a catalytic amount of CuI catalyst, as shown in Figure 1.
mol%, a yield of 85% was observed. Interestingly, isomeric
With this accelerated effect in mind, addition of 1.3 ratios rapidly decreased from 9:1 for 2a at 30 mol% loading
equivalents of TMSCl to 1a (30 mol% CuI·2LiCl and 1.2 to 1.5:1 for 2b (the opposite isomer) at 10 mol%. Even the
equivalents of PhMgBr in THF) at Ϫ50 °C led to a 90% difference in the E/Z ratio between 20 and 30 mol%
consumption of the starting material, as shown by 1H CuI·2LiCl was unexpected (Entries 6 and 7). At higher
NMR spectroscopy of the crude mixture. A diastereomeric catalyst loadings (40 and 50 mol%), better yields and excel-
ratio of 78:22 was observed in favor of the E-isomer. Iso- lent diastereoselectivities were observed (12:1 and 20:1,
lation of 2a and 2b by column chromatography provided an respectively) than those for 30 mol%.
85% combined yield. As anticipated, addition of TMSCl to
Expansion of the scope of the reaction with 1a to include
the reaction mixture presumably leads to 5,which allows for other Grignard reagents and a substituted alkynoate (1b) is
the turnover of the copper() salt, thus rendering the se- shown in Equation (2) and summarized in Table 3. All of
quence catalytic. However, the degree of diastereoselective the reactions, including that with MeMgBr, readily pro-
induction decreased from 9:1 to slightly more than 3:1. ceeded with 1a, except for the reaction with the sterically
Nonetheless, it was gratifying to note that: a) the reaction hindered 2-mesityl organocuprate, and furnished selectivit-
was catalytic upon addition of TMSCl, thus providing ies ranging from 8:1 to 40:1 for the E-isomer. Interestingly,
proof-beyond-principle of the initial hypothesis, and b) pro- Et2O does not allow for carbocupration under these con-
tic quench furnished selectivity for 2a over 2b (3:1).
ditions, while coordinating solvents such as THF or DME
To further probe the nature of this TMSCl-mediated car- provide excellent yields and nearly stereochemically pure
bocupration, the reaction conditions were altered with re- products. Much to our delight, the substituted alkynoate 1b
spect to temperature and catalyst loading. The results are showed a propensity for a catalytic and selective carbocupr-
summarized in Table 2 and led to the following two con- ation. Selectivities of about 9:1 for the E-isomer (with both
clusions. Firstly, it was observed that diastereoselectivity PhMgBr and PhCH2MgCl) have been observed for a 50
was influenced by reaction temperature at 30 mol% catalyst mol% catalyst loading with 72 and 74% isolated yields. A
loading. Warming the temperature from Ϫ50 °C to Ϫ35 °C lower catalyst loading of 30 mol% provided a 7:1 selectivity;
produced a lower E/Z isomeric ratio with a slightly lower however the isolated yield was only 55%. These results
yield, and moreover no catalysis was observed at 0 °C. greatly expand the scope of the TMS-promoted catalytic
However, by performing the reaction at Ϫ78 °C (Entry 7, carbocupration to include substituted alkynoates.
3202
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 3201Ϫ3204

TMSCl-Mediated Catalytic Carbocupration of Alkynoates
SHORT COMMUNICATION
the allenoate 5 and primarily ‘‘direct’’ the protonation of
the copper-allenoate complex or an aggregate thereof.
Table 3. TMSCl-CuI·2LiCl catalyzed carbocupration (50 mol%) of
1a and 1b with various Grignard reagents
Entry Ester Solvent Grignard reagent Yield [%] E[a]
Z
Conclusion
1
2
3
4
5
6
7
8
9
10
11[g]
12[g]
13
14
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1b
1b
1b
THF
Et2O
DME
THF
THF
THF
THF
THF
THF
THF
THF
THF
THF
THF
PhMgBr[b]
PhMgBr
91
Ͻ 5
92
87
91
95
Ϫ
5
Ϫ
7
13
12
12
5
3
2
3
7
13
11
9
PhMgBr
93
87
88
88
95
97
98
97
93
87
89
91
In summary, we have accomplished the first highly dia-
stereoselective TMSCl-promoted catalytic carbocupration
of α,β-acetylenic esters with the use of sub-stoichiometric
amounts (as low as 30 mol%) of a copper() salt in conjunc-
tion with a series of Grignard reagents. In addition, an un-
precedented and remarkable effect of catalyst loading on
stereochemical induction was observed. Future work will
be geared toward searching for more robust and selective
catalysts, and deciphering the role of the copper catalyst
during the selective protic quench.
MeMgBr[b]
hexylMgBr[b]
tert-butylMgCl[c] 77
sec-butylMgCl[d] 88
o-tolylMgCl[e]
MesitylMgBr[f]
PhCH2MgCl[b]
PhCH2MgCl
PhMgBr[b]
71
23
92
89
55
72
74
PhMgBr
PhCH2MgCl[h]
[a]
E/Z ratio determined by 1H NMR spectroscopy (360 MHz) of
the crude reaction mixture. [b] Resulting products are commercially
available. [c] Ref. 8. [d] Ref. 9. [e] Ref. 10. [f] Ref. 11. Reactions ran
[g]
Experimental Section
[h]
with 30 mol% CuI·2LiCl. Ref. 12.
All of the reactions were performed under Ar in flame-dried glass-
ware. Anhydrous tetrahydrofuran (THF) and dimethoxyethane
(DME) was obtained from commercial sources and used without
purification. Deuterated chloroform (CDCl3) was stored over
˚
molecular sieves (4 A). Copper() iodide (98% purity) and lithium
chloride (LiCl, 99%ϩ, ACS) were obtained from Aldrich and used
without any further purification. The NMR spectra were recorded
1
with a 360 MHz Bruker spectrometer. H NMR spectra were ob-
(2)
tained using CDCl3 as the solvent with either tetramethylsilane
(TMS: δ ϭ 0 ppm) or chloroform (CHCl3: δ ϭ 7.24 ppm) as the
internal standard. Column chromatography was performed using
60Ϫ200 µm silica gel. Analytical thin layer chromatography was
performed on silica coated glass plates with F-254 indicator.
Visualization was accomplished by UV light (254 nm) and KMnO4.
Attempts to rationalize the observed selectivities are
speculative at this point due to the complex nature of or-
ganocuprate chemistry.[1] However, a couple of key points
merit discussion. First, the structure of the soluble inter-
mediate magnesio-cuprate is unclear; however, the presence
of LiCl is required to catalyze the carbocupration of 1a.
Secondly, it is believed that the organocuprate facilitates
carbocupration of 1a, thus producing an intermediate vinyl-
cuprate species 3, which undergoes TMSCl-mediated iso-
merization to the TMS-allenoate 5 with the liberation of
the cuprate catalyst (Figure 1).[5]
Unfortunately, all attempts to isolate 5 have been unsuc-
cessful. However, we have detected intermediate 5 by 13C
NMR spectroscopy prior to protic quench;[6] this provides
proof that the TMS-allenoate is a true intermediate. There-
fore, it was surprising to observe a direct correlation be-
tween catalyst loading and stereoselective induction from
the protic quench of 5.[1c] These results suggest that the
amount of catalyst plays a significant and unprecedented role
in determining the diastereoselectivity during the proton
quench of the TMS-allenoate. Towards this end, Corey has
suggested the formation of copper-allenolate π complexes
to account for the high stereoselectivity of SN2Ј displace-
ment reactions of chiral 1,3-disubstituted bromoallenes.[7]
Based on this premise, we believe that in our case the alkyl-
General Experimental Procedure for the TMSCl-Promoted Cata-
lytic Carbocupration of 1a and 1b: CuI (0.291 g, 1.50 mmol) and
LiCl (0.130 g, 3.0 mmol) was placed in a 100 mL round bottom
flask (flame dried under vacuum) under Ar. Dry THF (20 mL) was
added to these salts, and the mixture was stirred at room tempera-
ture for a period of 0.5 h until complete dissolution had occurred.
The clear, light yellow homogeneous solution was cooled to Ϫ78
°C, and 1a (0.294 g, 3.0 mmol) was added, followed by TMSCl (1.3
eq., 0.51 mL, 3.9 mmol). After 5 minutes at Ϫ78 °C, benzylmag-
nesium chloride (1.2 equiv., 3.60 mL, 3.6 mmol) was added drop-
wise with a syringe, and the solution was stirred at Ϫ78 °C for 1 h.
Saturated ammonium chloride solution was added to quench the
reaction at Ϫ78 °C, and the mixture was allowed to warm to room
temperature and stir for 30 min. The product was extracted with
Et2O (3 ϫ 25 mL) and washed with deionized H2O followed by
brine. The organic layer was separated, dried with MgSO4, and
concentrated in vacuo to give the crude product, which was then
analyzed by 1H NMR spectroscopy to determine regioselectivity.
Column chromatography of the crude material (10% ethyl acetate
in hexane) afforded an 89% yield of the diastereoselective pure
(E)-olefin.
Acknowledgments
cuprate might (from Figure 1) undergo a π complexation of This work was supported by The University of Alabama.
Eur. J. Org. Chem. 2004, 3201Ϫ3204
www.eurjoc.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3203

M. P. Jennings, K. B. Sawant
SHORT COMMUNICATION
[1] [1a]
lithium organocuprate additions to ynoates see: K. Nilsson, T.
Andersson, C. Ullenius, J. Organomet. Chem. 1997, 545/546,
591Ϫ595.
E. Nakamura; S. Mori, Angew. Chem. Int. Ed. 2000, 39,
[1b]
3750Ϫ3771.
S. Woodward, Chem. Soc. Rev. 2000, 29,
393Ϫ401. [1c] K. Nilsson, T. Andersson, C. Ullenius, A. Gerold,
13C NMR shifts for the allenic portion of intermediate 5 (R ϭ
Bz): δ ϭ 182.5, 163.6, 118.0 ppm. Compound 5 is not thermally
stable at room temperature after 12 h. The chemical shifts are
[6]
[7]
[1d]
N. Krause, Chem. Eur. J. 1998, 4, 2051Ϫ2058.
B. H. Lip-
shutz; S. Sengupta, Organic Reactions. 1992, 41, 135Ϫ631. and
references cited therein.
similar to those reported in ref.[5]
.
[2] [2a]
J. W. J. Kennedy, D. G. Hall, J. Am. Chem. Soc. 2002, 124,
[2b]
For a proposed copper-allenoate π complex that leads to a SN2Ј
selective quench see: E. J. Corey, N. W. Boaz, Tetrahedron Lett.
1984, 25, 3063Ϫ3066.
898Ϫ899.
L. Jalander, K. Iambolieva, V. Sundstrom, Acta
[2c]
Chem. Scand., Ser. B 1980, 34, 715Ϫ720.
R. J. Anderson,
V. L. Corbin, G. Cotterrell, G. R. Cox, C. A. Henrick, F.
Schaub, J. B. Siddall, J. Am. Chem. Soc. 1975, 97, 1197Ϫ1204.
[8]
[9]
´
`
C. Jimeno, M. Pasto, A. Riera, M. A. Pericas, J. Org. Chem.
2003, 68, 3130Ϫ3138.
[2d]
J. Klein, N. Aminadav, J. Chem. Soc., (C) 1970,
[2e]
A. A. VasilЈev, A. L. Vlasyuk, G. D. Gamalevich, E. P. Sereb-
ryakov, Bioorg. Med. Chem. 1996, 4, 389Ϫ400.
Y. Chen, L. Huang, M. A. Ranade, X. P. Zhang, J. Org. Chem.
2003, 68, 3714Ϫ3717.
A. K. Chatterjee, T.-L. Choi, D. P. Sanders, R. H. Grubbs, J.
Am. Chem. Soc. 2003, 125, 11360Ϫ11370.
Y. Yamamoto, S. Hatsuya, J. Yamada, J. Org. Chem. 1990,
55, 3118Ϫ3128.
1380Ϫ1385.
C. Bretting, J. Munch-Petersen, P. M. Jorg-
ensen, S. Refin, Acta Chem. Scand. 1960, 14, 151Ϫ156.
M. T. Reetz, A. Kindler, J. Organomet. Chem. 1995, 502,
C5ϪC7.
[10]
[11]
[12]
[3]
[4] [4a]
C. Chuit, J. P. Foulon, J. F. Normant, Tetrahedron. 1980,
[4b]
36, 2305Ϫ2310.
E. J. Corey, N. W. Boaz, Tetrahedron Lett.
[4c]
1985, 26, 6015Ϫ6018.
Tetrahedron Lett. 1986, 27, 1047Ϫ1050.
For an NMR mechanistic study of the effect of TMSCl on
A. Alexakis, J. Berlan, Y. Besace,
[5]
Received May 5, 2004
3204
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 3201Ϫ3204