Si-based benzylic 1,4-rearrangement/cyclization reaction

Article abstract of DOI:10.1021/ol802289f

(Chemical Equation Presented) The trans-selective hydrosilylation of ynones (1) yields ss-silylated enones (2) that undergo a benzylic 1,4-rearrangement/cyclization reaction in the presence of base, yielding 2,5-dihydro-1,2-oxasiloles (3).

Full text of DOI:10.1021/ol802289f

ORGANIC
LETTERS
2009
Vol. 11, No. 3
511-513
Si-Based Benzylic 1,4-Rearrangement/
Cyclization Reaction
Barry M. Trost* and Andreas Bertogg
Department of Chemistry, Stanford UniVersity, Stanford, California 94305-5080
bmtrost@stanford.edu
Received October 2, 2008
ABSTRACT
The trans-selective hydrosilylation of ynones (1) yields ꢀ-silylated enones (2) that undergo a benzylic 1,4-rearrangement/cyclization reaction
in the presence of base, yielding 2,5-dihydro-1,2-oxasiloles (3).
Use of silicon compounds for the formation of C-C bonds
has largely revolved around electrophilic substitutions in-
Scheme 1. Si-Based Benzylic 1,2-Shift
volving allyl, vinyl, or aryl systems.1 More recently, their
use as nucleophilic partners in cross-coupling reactions has
shown great promise.2 In conjunction with our studies
directed at a trans hydrosilylation,3 we investigated a geminal
alkylation-hydroxylation of alkynyl carbonyl compounds as
shown in Scheme 1.3b
In conjunction with our studies of exploiting this process
in complex natural product chemistry, we discovered a totally
different and, to the best of our knowledge, unprecedented
1,4-shift in such substrates (2f3) in the presence of
nonfluoride base to give 2,5-dihydro-1,2-oxasiloles (see
abstract).4
The initial observation evolved from our efforts to examine
a Peterson-style olefination as shown in Figure 1, path A.
(1) (a) Brook, M. A. Silicon in Organic, Organometallic, and Polymer
Chemistry; John Wiley & Sons: Toronto, 2000; pp 569-595. Strained
silacycles have also been used to transfer allyl and crotyl fragments with a
high degree of enantioselectivity. (b) Burns, N. Z.; Hackman, B. M.; Ng,
P. Y.; Powelson, I. A.; Leighton, J. L. Angew. Chem., Int. Ed. 2006, 45,
3811–3813. (c) Park, P. K.; O‘Malley, S. J.; Schmidt, D. R.; Leighton,
J. L. J. Am. Chem. Soc. 2006, 128, 2796–2797. (d) Wang, X.; Meng, Q.;
Perl, N. R.; Xu, Y.; Leighton, J. L. J. Am. Chem. Soc. 2005, 127, 12806–
12807.
(2) (a) Hiyama, T.; Shirakawa, E. In Cross-Coupling Reactions: A
Practical Guide; Miyaura, N., Ed.; Topics in Current Chemistry; Springer-
Verlag: Berlin, Heidelberg, 2002, Vol. 219. (b) Hiyama, T. In Metal-
Catalyzed Cross-Coupling Reactions; Diederich, F., Stang, P., Eds.; Wiley-
VCH: Weinheim, 1998. (c) Hatanaka, Y.; Fukushima, S.; Hiyama, T.
Heterocycles 1990, 30, 303–306. (d) Denmark, S.; Sweis, R. F. Acc. Chem.
Res. 2002, 35, 835. (e) Denmark, S. E.; Tymonko, S. A. J. Am. Chem. Soc.
2005, 127, 8004–8005.
(3) (a) Trost, B. M.; Ball, Z. T. J. Am. Chem. Soc. 2001, 123, 12726–
12727. (b) Trost, B. M.; Ball, Z. T. J. Am. Chem. Soc. 2004, 126, 13942–
13944. (c) Trost, B. M.; Ball, Z. T. J. Am. Chem. Soc. 2005, 127, 17644–
17655.
Figure 1. Discovery of the benzylic 1,4-shift.
10.1021/ol802289f CCC: $40.75
Published on Web 01/06/2009
2009 American Chemical Society

Surprisingly, none of the expected enoate was observed. The
mass spectrum and elemental analysis indicated a molecular
formula of C17H20O2Si. The infrared spectrum showed no
carbonyl peaks. Combined with the 1H and 13C NMR spectra,
the only structure that can be assigned is depicted as 3a (see
Table 1) which was further confirmed by subsequent
transformations (vide infra).5 Since the olefinating agent was
clearly not involved, the reaction was repeated just with LDA
and an even more hindered amide base, lithium t-butyl-
tritylamide (LiTTA). Adding the base at -78° and allowing
the reaction to warm to 0° before quenching gave the same
product 3a, which in the case of LiTTA as base was isolated
in 86% yield (98% based on recovered starting material). A
nonamide base, trityllithium, also promoted the reaction. In
this case, keeping the reaction at -78° and quenching at
that temperature gave a 66% conversion after 5.5 h.
Table 1. Novel 1,4-Rearrangement/Cyclization Reactiona b
,
To explore the generality of the process, a series of ynones
were hydrosilylated using the Ru-catalyzed trans hydrosi-
lylation protocol3b using the appropriately functionalized
benzyldimethylsilane as summarized in Table 1. It is
interesting to note that in the 1,4-rearrangement/cyclization
reaction electron-deficient benzyl silanes work as well as
benzylsilane itself. While LDA or trityllithium was generally
used, LiTTA gave the cleanest reactions with the furyl
substrates 2a-2d. Given the basic nature of the reaction
conditions, the effect of the presence of enolizable hydrogens
in both R and R′ were explored. Interestingly, aryl-aryl,
aryl-alkyl, alkyl-aryl, and alkyl-alkyl (for R and R′,
respectively) all work with the alkyl-alkyl enones being the
slowest. Regarding the migrating benzyl group, the more
electron-deficient ones migrate the fastest.
To provide further proof of the structures of the 2,5-
dihydro-1,2-oxasiloles 3, we performed degradation reactions
on several silacycles. Applying protodesilylation conditions
to the silyl ethers provided trans-allylic alcohols (Table 2),
Table 2. Trans-Allylic Alcohols: Proof of Structure
entry substituent R substituent R′ aryl substituent Ar yield
a A solution of the enone was treated with 1.00-1.25 equiv of the base
indicated at -78 °C. The solution was either stirred and quenched at -78 °C
(entries 3, 4, 7, 8, 9, 10, 11) or quenched after warming to 0 °C (entry 1), to
4 °C (entry 6) or rt (entries 2, 5). b Protocol of a typical rearrangement/
cyclization reaction (2h f 3h): A solution of (Z)-3-(dimethyl(4-(trifluorom-
ethyl)benzyl)silyl)-1,3-diphenylprop-2-en-1-one (2h; 160 mg, 377 µmol, 1.00
equiv) in 0.8 mL of THF was treated with Li-diisopropylamide (798 mM
solution in THF/hexanes; 590 µL, 471 µmol, 1.25 equiv) at -78 °C. After
stirring for 1 h 40 min, the reaction was quenched by addition of phosphate
buffer (pH ) 7), extracted with diethyl ether, dried over Na2SO4, filtered, and
concentrated in vacuo. Purification by column chromatography yielded 2,2-
dimethyl-3,5-diphenyl-5-(4-(trifluoromethyl)benzyl)-2,5-dihydro-1,2-oxas-
ilole (3h; 144 mg, 338 µmol, 90%) as a colorless solid. 1H NMR (CDCl3, 400
MHz): δ ) 0.00 (s, 3 H, SiCH3), 0.40 (s, 3 H, SiCH3), 3.29 (d, J ) 13.2 Hz,
1 H, CH2), 3.38 (d, J ) 13.2 Hz, 1 H, CH2), 7.23-7.36 (m, 9 H, 9 CH),
7.38-7.42 (m, 2 H, 2 CH), 7.49 (d, J ) 7.6 Hz, 2 H, 2 CH), 7.53-7.56 (m,
2 H, 2 CH). 13C{1H,19F} NMR (CDCl3, 100 MHz): δ ) 0.65 (SiCH3), 1.48
(SiCH3), 49.7 (CH2), 89.1 (Cquat), 124.6, 124.7, 125.1, 127.0, 127.1, 127.7, 128.7,
128.9, 129.0, 131.7, 137.3, 141.0, 141.4, 146.0, 147.1. 19F{1H} NMR (CDCl3,
376.3 MHz): δ ) -62.6. IR (neat): ν (cm-1) 3058, 3026, 2954, 1618, 1491,
1446, 1418, 1326, 1254, 1164, 1122, 1067, 1020, 955, 922, 881, 849, 824,
789, 756, 697, 635, 609. HRMS (EI): (MNa)+ calcd for C25H23OF3SiNa
447.1368, found 447.1360. C25H23F3OSi (424.53): calcd C 70.73, H 5.46; meas.
C 70.52, H 5.62. mp ) 90.0 °C. c LiTTA is lithium tert-butyl(trityl)amide.
d This reaction was performed on a 1.52 g (4.92 mmol) scale. e 2.00 equiv
of base was employed.
1
2
3
4
C6H5
CH2CH3
CH3
CH2CH3
CH2CH3
C6H5
C6H5
C6H5
4-(F3C)C6H4
90%
81%
87%
88%
3-furyl
3-furyl
3-furyl
thus proving both the correctness of our structural assignment
and the unusual nature of the benzylic 1,4-shift. The
transformation 3 f 4 represents a method for accessing
stereodefined allylic alcohols from ynones that is comple-
(4) A 1,4-shift of allyl or crotyl substituents from Si to carbonyl C under
thermal conditions was reported recently. See: Bashiardes, G.; Chaussebourg,
V.; Laverdan, G.; Pornet, J. Chem. Commun. 2004, 122–123. An interesting
allylic 1,4-rearrangement as part of a two-step process including a Rh-
catalyzed silylformylation (notably, in this transformation no silacycle is
isolated) was described. See: O’Malley, S. J.; Leighton, J. L. Angew. Chem.,
Int. Ed. 2001, 40, 2915–1917. Applying thermal conditions (such as
described by Bashiardes et al.) to the γ-silylated enones used in this study
did not result in a rearrangement.
(5) The structural assignment for the 2,5-dihydro-1,2-oxasilole products
by NMR experiments was verified based on two-dimensional NMR
measurements for 3b.
512
Org. Lett., Vol. 11, No. 3, 2009

mentary to known protocols with respect to reaction condi-
tions and reagents.
Furthermore, we subjected silacycle 3f to Tamao-Fleming
oxidation conditions (Scheme 2), leading to the isolation of
ment (1,2- vs 1,4-shift) is determined by the reaction
conditions, not by the structure of the substrate, and thus
proved that the two processes under comparison are comple-
mentary. Mechanistically, initiation of the transformation by
deprotonation appears unlikely given that nonenolizable
substrates react the fastest. The lack of deuterium incorpora-
tion in the benzylic position upon quenching the reaction
with D2O at partial conversion also rules out any deproto-
nation there.
Scheme 2. Further Proof for Structure: Use of the
Rearrangement As an Alternative Access to the Aldol Structural
Motif
The attractive interaction between carbonyl functions and
silyl substituents in respective cis-olefins6 increases the
electrophilicity of the Si nucleus. The coordination of a
relatively small fluoride anion to Si3b presumably puts the
benzyl substituent in a pseudoaxial position thus facilitating
a benzylic 1,2-shift. Since the use of sterically demanding
bases as reagents excludes a coordination to Si, a single-
electron transfer mechanism seems most likely, leading to a
benzylic 1,4-shift. Following the reduction of the substrate
to a radical anion, the benzyl substituent migrates as an anion.
This view is supported by the fact that electron-withdrawing
substituents at the benzyl substituents increase the rate of
the reaction.
the aldol 5. The independent synthesis of this aldol structure
by the addition of a benzyl Grignard reagent to 1,3-diphenyl-
1,3-propanedione provided another proof for the structure
of the silacycles 3.
Earlier, we reported3b a benzylic 1,2-rearrangement which
had been applied to substrates that are similar to the enones
2 used for the benzylic 1,4-rearrangement/cyclization reaction
presented here. To verify that the two rearrangement reac-
tions are independent processes, we subjected a ꢀ-silylated
enone used herein to the reported reaction conditions (Figure
2). The experiment showed that the nature of the rearrange-
In summary, we have discovered a benzylic 1,4-rearrange-
ment/cyclization reaction. Our current efforts focus on
deepening our mechanistic understanding of this transforma-
tion.
Acknowledgment. We thank the NSF and the General
Medical Sciences Institute of NIH (GM13598) for their
support of our programs. We also thank the Swiss National
Science Foundation and the Roche Research Foundation for
fellowship support. Johnson-Matthey is greatly acknowledged
for their generous gift of Ru salts.
Supporting Information Available: Experimental pro-
cedures and full spectroscopic data for all new compounds.
This material is available free of charge via the Internet at
http://pubs.acs.org.
OL802289F
Figure 2. Benzylic 1,4-shift vs 1,2-shift: a mechanistic proposal.
(6) Shindo, M.; Matsumoto, K.; Shishido, K. Angew. Chem., Int. Ed.
2004, 104–106.
Org. Lett., Vol. 11, No. 3, 2009
513