Catalytic redox-initiated glycolate aldol additions of silyl glyoxylates

Article abstract of DOI:10.1021/ol802828d

Lanthanide triisopropoxides catalyze a rapid, tandem MPV reduction/Brook rearrangement/aldol sequence between silyl glyoxylates and aldehydes that achieves catalytic turnover through alkoxide transfer from a strain-release Lewis acidic silacycle.

Full text of DOI:10.1021/ol802828d

ORGANIC
LETTERS
2009
Vol. 11, No. 4
827-830
Catalytic Redox-Initiated Glycolate Aldol
Additions of Silyl Glyoxylates
Stephen N. Greszler and Jeffrey S. Johnson*
Department of Chemistry, UniVersity of North Carolina at Chapel Hill,
Chapel Hill, North Carolina 27599-3290
jsj@unc.edu
Received December 8, 2008
ABSTRACT
Lanthanide triisopropoxides catalyze a rapid, tandem MPV reduction/Brook rearrangement/aldol sequence between silyl glyoxylates and aldehydes
that achieves catalytic turnover through alkoxide transfer from a strain-release Lewis acidic silacycle.
Bimolecular chemical reactions that achieve dual symbiotic
activation of both reaction partners offer unique and attractive
opportunities for efficiency and atom economy. Examples
of this unusual reaction mode include aldol reactions initiated
by redox reaction between silyl glyoxylates and magnesium
alkoxides1 and carbonyl allylations initiated by formal H2
redistribution.2a-c The former reaction demonstrated a new
means for generating active coupling partners in situ, but a
challenge that emerged during subsequent studies was the
regeneration of the metal alkoxide used as the reducing agent.
The absence of a turnover mechanism necessitated the use
of a stoichiometric metal alkoxide species (Figure 1). This
paper delineates reaction parameters that are essential in
achieving catalysis of the title reaction, and we specifically
describe the application of strain-release Lewis acidic silox-
anes in turnover-enabling alkoxide metathesis.3
Key mechanistic features of the projected catalytic reaction
are summarized in Figure 1. Following the established silyl
Figure 1. Catalytic reductive aldol reaction: proposal.
(1) Linghu, X.; Satterfield, A. D.; Johnson, J. S. J. Am. Chem. Soc. 2006,
128, 9302–9303.
glyoxylate reactivity pattern involving a Meerwein-Ponndorf-
Verley (MPV) reduction, Brook rearrangement, and aldol
addition,1 our point of departure from the stoichiometric
reaction would be the transfer of some undefined moiety Ω
from 4 to the terminal metal aldolate 3b. This proposed
alkoxide metathesis would concurrently release the aldolate
product 5 and regenerate the MPV reductant 2.
(2) (a) Kim, I. S.; Ngai, M.-Y.; Krische, M. J. J. Am. Chem. Soc. 2008,
130, 14891–14899. (b) Kim, I. S.; Ngai, M.-Y.; Krische, M. J. J. Am. Chem.
Soc. 2008, 130, 6340–6341. (c) Catalytic functionalization of carbinol C-H
bonds has also been achieved: Shi, L.; Tu, Y.-Q.; Wang, M.; Zhang, F.-
M.; Fan, C.-A.; Zhao, Y.-M.; Xia, W.-J. J. Am. Chem. Soc. 2005, 127,
10836–10837.
(3) For a monograph on the aldol reaction, see: Modern Aldol Reactions;
Mahrwald, R., Ed.; Wiley-VCH: Weinheim, Germany, 2004.
10.1021/ol802828d CCC: $40.75
Published on Web 01/22/2009
2009 American Chemical Society

Preliminary experiments focused on defining two key
reaction components. First, the identity of the metal cation
would likely prove a determining factor, as previous reactions
employing silyl glyoxylates have required the careful selec-
tion of a metal that not only promotes MPV reduction but
also Brook rearrangement of the intermediate C-silyl alkox-
ides.4
Table 1. Preliminary Trials Using Strain-Release Silacyclesa
Additionally, alkoxide donor 4 must be sufficiently labile
in order to facilitate effective catalytic turnover. In an initial
screen of a variety of common metal triisopropoxides and
acylating or silylating agents (Figure 2), we achieved no
(silacycle)
Er(OiPr)3
(mol %)
7 (8)b
(% yield)
time
(min)
entry
R1
R2
1
2
3
4
5
Me
iPr
Me
Me
Me
H (6a)
5
5
5
10
1
50-(32)
50-(28)
62-(3)
84-(6)
trace
5
40
60
30
120
H (6b)
Me (6c)
Me (6c)
Me (6c)
a Conditions: 1.5 equiv of PhCHO, 2.0 equiv of 6, [1]0 ) 0.08 M.
b
1
Yields determined by H NMR versus an internal standard.
but isopropoxysilacycle 6c provided the desired coupling
product in a 62% yield and with only 3% of the byproduct
8c present (entries 2 and 3), corresponding to reaction with
the sacrificial equivalent of acetone generated. Increased
yields could be attained with the use of 10% of the metal
catalyst, while catalyst loadings less than 5% provided only
trace product formation (entries 4 and 5).
Figure 2. Summary of initial catalyst and turnover reagent screens.
greater than 30% conversion to the desired aldol product,
suggesting that alkoxide transfer from the putative turnover
reagents was not occurring.
With a successful means for catalytic turnover, we
screened additional reaction parameters. Toluene proved to
be the optimal solvent choice, providing the desired product
in moderate yields and in shorter reaction times than in ether,
while incomplete reactions were observed in dichlo-
romethane, THF, and 2-methyl-THF.6
Realizing that a more reactive turnover agent may be
necessary, we turned to strain-release silacycles, which have
gained attention through their application in a variety of
transformations.5 The enhanced Lewis acidity of these
silacycles is due to their ring constraints and contributes to
their ability to function as potent allylating agents and enolate
equivalents. We wondered if they might exhibit accelerated
subsitution chemistry relative to unconstrained variants.
Ethoxysilacycle 6a was synthesized using a procedure
modified from Leighton’s published work.5a In the presence
of 6a and 5 mol % of erbium(III) isopropoxide, silyl
glyoxylate 1 and benzaldehyde reacted completely in under
5 min. Product analysis revealed an important pitfall from
this preliminary trial: aldol reaction with the sacrificial
equivalent of acetaldehyde generated from the ethoxide
transfer and MPV reduction sequence proved competitive
with the desired reaction with benzaldehyde (Table 1, entry
1). Isobutoxysilacycle 6b afforded a similar product ratio,
Among the cations screened, aluminum7 and magnesium1
provided only trace product in this catalytic system. Yttrium
Table 2. Screen of Metal Catalystsa
M(OiPr)n
(mol %)
% yield
7 (8c)b
reaction time
(min)
radiusc
(Å)
entry
1
2
3
4
5
6
7
8
Al(OiPr)3 (5)
Dy(OiPr)3 (5)
Zr(OiPr)4 (5)
Mg(OiPr)2 (5)
Y(OiPr)3 (5)
Er(OiPr)3 (5)
Er(OiPr)3 (10)
Gd(OiPr)3 (10)
Yb(OiPr)3 (10)
Sm(OiPr)3 (10)
Pr(OiPr)3 (10)
trace
trace
trace
trace
52 (1)
62 (5)
84 (6)
67 (12)
72 (8)
60 (11)
91 (8)
300
300
300
300
120
120
30
25
15
10
1
(4) For reduction of acyl silanes, see: (a) Bolm, C.; Kasyan, A.; Heider,
P.; Saladin, S.; Drauz, K.; Guenther, K.; Wagner, C. Org. Lett. 2002, 4,
2265–2267. (b) Brook, A. G.; Quigley, M. A.; Peddle, G. J. D.; Schwartz,
N. V.; Warner, C. M. J. Am. Chem. Soc. 1960, 82, 5102–5106. (c) Reich,
H. J.; Holtan, R. C.; Bolm, C. J. Am. Chem. Soc. 1990, 112, 5609–5617.
(5) (a) Kinnaird, J. W. A.; Ng, P. Y.; Kubota, K.; Wang, X.; Leighton,
J. L. J. Am. Chem. Soc. 2002, 124, 7920–7921. (b) Wang, X.; Meng, Q;
Perl, N. R.; Xu, Y.; Leighton, J. L. J. Am. Chem. Soc. 2005, 127, 12806–
12807. (c) Myers, A. G.; Widdowson, K. L.; Kukkola, P. J. J. Am. Chem.
Soc. 1992, 114, 2765–2767. (d) Denmark, S. E.; Griedel, B. D.; Coe, D. M.
J. Org. Chem. 1993, 58, 988–990. (e) Sunderhaus, J. D.; Lam, H.; Dudley,
G. B. Org. Lett. 2003, 5, 4571–4573.
0.900
0.890
0.890
0.938
0.868
0.958
0.997
9
10
11
a Conditions: 1.5 equiv of PhCHO, 2 equiv of 6c, [1]0 ) 0.08 M. b Yields
determined by 1H NMR spectroscopy versus an internal standard. c See ref
8 for coordination number ) 6.
(6) Detailed optimization studies are available in the Supporting
Information.
828
Org. Lett., Vol. 11, No. 4, 2009

and a variety of lanthanides exhibited an inverse relationship
between ionic radius8 and reaction time (Table 2, entries 5-11).
Reactions employing erbium required 30 min to reach
completion, while praseodymium catalyzed the addition to
>98% conversion in approximately 1 min with benzaldehyde
(Table 2, entries 7 and 11).
Based on previous studies in related reactions from our
laboratory and the results of the labeling experiment, the
catalytic cycle in Figure 5 is proposed. Initial MPV reduction
Two reasonable reduction mechanisms could be formu-
lated to account for the observed reductive aldol products
(Figure 3). Initial reduction of benzaldehyde could precede
Figure 3. Possible hydride-transfer pathways.
Figure 5. Proposed catalytic cycle employing silacycle 6c.
reduction of the silyl glyoxylate by the resulting benzyl
alkoxide (path A); alternatively, C-silyl alkoxide intermediate
9 could arise through a direct MPV reduction of the
silylglyoxylate (path B).
The two mechanisms may be distinguished through the
application of the deuterium-labeled isopropylsiloxane 10.
Using 10 mol % of Pr(2-d-OiPr)3 as a catalyst, 99%
deuterium incorporation was observed at CR (Figure 4). This
of silyl glyoxylate 1 results in generation of a transient
alkoxide intermediate that undergoes Brook rearrangement
to afford ester enolate 13. Aldol reaction with the aldehyde
provides terminal alkoxide 14, which then attacks the strained
silacycle 6c and expels another isopropoxide equivalent to
regenerate the lanthanide triisopropoxide catalyst, possibly
facilitated by participation of the ester carbonyl.
A sacrificial equivalent of acetone is generated in the initial
step of this cycle; dissociation of acetone from the metal
catalyst therefore must proceed more rapidly than its aldol
reaction occurs to avoid predominant production of byproduct
8c. This is consistent with crossover experiments previously
conducted in a related system.1 When the MPV reduction
produces a sacrificial aldehyde, however, competitive aldol
reactions afford the product ratios observed in Table 1
(entries 1 and 2).
The inverse relationship between ionic radius and reaction
time potentially reflects steric limitations in the alkoxide
transfer step; the larger coordination sphere of praseodymium
likely facilitates the necessary complex formation as well
as dissociation of the final product from the metal center.
The reaction worked for the aryl, linear and branched alkyl,
and heteroaromatic aldehydes shown in Table 3. In spite of
the modest diastereoselectivities observed, reactions were
generally quite rapid and gave good yields of the glycolate
aldol products. Reaction with acetophenone provided only
17% of the desired addition product, with the remaining silyl
glyoxylate consumed by addition to acetone. The background
Figure 4. Isotopic labeling study. Conditions: 1.5 equiv of PhCHO,
2 equiv of 10, [1]0 ) 0.08 M.
result indicates that hydride scrambling via aldehyde reduc-
tion does not contribute appreciably to product formation
and that path B is active in this system. This is consistent
with the increased electrophilicity of 1 that is generally
observed.9
(7) (a) Campbell, E. J.; Zhou, H.; Nguyen, S. T. Org. Lett. 2001, 3,
2391–2393. (b) Yin, J.; Huffman, M. A.; Conrad, K. M.; Armstrong, J. D.
J. Org. Chem. 2006, 71, 840–843. (c) de Graauw, C. F.; Peters, J. A.; van
Bekkum, H.; Huskens, J. Synthesis 1994, 1007–1017.
(8) For coordination number of 6, M3+: Shannon, R. D. Acta Crysallogr.
A 1976, 32, 751–767.
(10) A competition experiment using equimolar quantities of these
aldehydes and hexanal provided only trace product:
(9) (a) Nicewicz, D. A.; Johnson, J. S. J. Am. Chem. Soc. 2005, 127,
6170–6171. (b) Nicewicz, D. A.; Breteche, G.; Johnson, J. S. Org. Synth.
2008, 85, 278–286.
Org. Lett., Vol. 11, No. 4, 2009
829

Table 3. Electrophiles for the Reductive Aldola
Table 4. Asymmetric Aldolization with (salen)Ln Complexesa
yieldb time
entry product
R1
R2
(%)
(min)
dr
1
2
3
4
5
6
7
8
9d
16a
16b
16c
8b
16d
16e
16f
16g
8c
Ph
H
H
H
H
80c
69
73c
71
1
2
3
1
7
120
120
120
15
1.5:1
1.5:1
1.3:1
1.5:1
1.3:1
n.d.
n.d.
n.d.
n.a.
CH3(CH2)4-
2-furyl
iPr
entry
(ligand) Ln
%
yield
R
R1
R2
erb
dr
Ph
Me 17
1-(17a) Er
2-(17b) Er
3-(17a) Sm -(CH2)4- Me
-(CH2)4- Me
H
18
50:50 1.5:1
-(CH2)4- tBu tBu
trace n.d.
PhCt C-
(E)-PhCHdCH-
PhCH2CH2-
Me
H
H
H
trace
trace
trace
H
63
41
0
61:39 1.4:1
55:45 2.1:1
n.a.
57:43 2.1:1
58:42 1.9:1
63:37
4-(17c) Sm Ph
tBu tBu
H
Me 67c
5-(17d) Sm -(CH2)4-
H
6-(17e) Sm -(CH2)4- tBu OMe 45
a Reaction conditions: 1.5 equiv of aldehyde or ketone, 10% Pr(OiPr)3,
2 equiv of 6c, [1]0 ) 0.08 M. b Unless otherwise noted, yields were
determined by 1H NMR spectroscopy versus an internal standard. c Isolated
yield. d No aldehyde or ketone was added.
7-(17a) Pr
8-(17c) Pr
9-(17f)
-(CH2)4- Me
H
33
80
58
Ph
tBu tBu
tBu tBu
2:1
Pr
naphthyl
53:47 1.7:1
a Reaction conditions: 1.5 equiv of PhCHO, 10 mol % of Ln(OiPr)3, 2
equiv of 6c, 10 mol % of 17a-f, [1]0 ) 0.08 M. b For major diastereomer.
reaction proceeded over 15 min to give the coupling product
with acetone in 67% yield (Table 3, entry 9).
decrease in yield (80%) and with 63:37 er. Ongoing studies
in our group aim to build upon these preliminary results.
In conclusion, we have developed a new method for the
MPV reduction/Brook rearrangement/aldol reaction of silyl
glyoxylates. The reactions are catalyzed by lanthanide
triisopropoxides and feature a unique turnover step, an
alkoxide transfer from a strain-release Lewis acidic silacycle.
Notably poor substrates were R,ꢀ-unsaturated aldehydes
and dihydrocinnamaldehyde (Table 3, entries 6-8); no
greater than trace quantities of desired product were observed.
Although we cannot provide a detailed rationale for the
failure of such substrates, they possibly impede reaction
through either complexation with or degradation of the metal
catalysts.10
Preliminary results also suggest the potential for asym-
metric catalysis in this system (Table 4). Salen ligands have
been shown to be effective in a lanthanide-catalyzed
aldol-Tischenko reduction developed by Morken;11 they
provided, therefore, a reasonable starting point. We have
verified that ligand f product chirality transfer is feasible,
with 17c·Pr(OiPr)3 providing the desired product with no
Acknowledgment. Funding for this work was provided
by the National Institutes of Health (National Institute of
General Medical Sciences, GM068443).
Supporting Information Available: Experimental pro-
cedures and analytical data for all new compounds. This
material is available free of charge via the Internet at
http://pubs.acs.org.
(11) Mascarenhas, C. M.; Miller, S. P.; White, P. S.; Morken, J. P.
Angew. Chem., Int. Ed. 2001, 40, 601–603.
OL802828D
830
Org. Lett., Vol. 11, No. 4, 2009