Stereospecificity in the silicon tethered α-(methyl)allylation of aldehydes

Article abstract of DOI:10.1039/b306920f

Heating E- and Z-crotyl(diphenyl)silyloxy aldehydes, in the absence of an added catalyst, results in sterospecific intramolecular allyl transfer with moderate to high stereoselectivity. The successful preparation of suitable substrates was described to test the given proposal. This paper reports their behavior under both Lewis acidic and thermal conditions.

Full text of DOI:10.1039/b306920f

View Article Online / Journal Homepage / Table of Contents for this issue
Stereospecificity in the silicon tethered ꢀ-(methyl)allylation of
aldehydes†
Jeremy Robertson,*a Michael J. Halla and Stuart P. Greenb
a Dyson Perrins Laboratory, South Parks Road, Oxford, UK OX1 3QY
b Chemical Research & Development Department, Pfizer Global Research and Development,
Sandwich, Kent, UK CT13 9NJ
Received 17th June 2003, Accepted 15th August 2003
First published as an Advance Article on the web 3rd September 2003
Heating E- and Z-crotyl(diphenyl)silyloxy aldehydes, in
the absence of an added catalyst, results in stereospecific
intramolecular allyl transfer with moderate to high stereo-
selectivity.
sought to prepare the E- and Z-crotyl analogues in the belief
that these would cyclise with improved ratios of diastereomeric
products (Scheme 2, X = H).4 In this communication we
describe the successful preparation of suitable substrates in
order to test this proposal, and report their behaviour under
both Lewis acidic and thermal conditions.
Introduction
We have described Lewis acid-mediated silicon tethered
ene cyclisations of β-prenylsilyl aldehydes that proceed with
moderate stereoselectivity in favour of a trans-disposition of
the newly-formed hydroxyl and isopropenyl groups.1 More
recently, we have observed that α-prenylsilyloxy aldehyde pre-
cursors show broadly similar behavior and benefit from an
easy oxidative cleavage pathway to generate stereodefined triols
(Scheme 1).2
Stereoselective synthesis of E- and Z-crotyl precursors
Substrates 4 (Scheme 3) and 10 (see Scheme 5) were identified as
the initial targets since our earlier work had shown that phenyl
substituents on silicon offered the optimum balance of stability
and eventual ease of oxidative cleavage of the C–Si bond.5 In
addition, we had found that Piers’ method6 for hydroxyl silyl-
ation could be relied upon even with allylic silanes, therefore
access to the Z-crotyl derivative 4 required the preparation of
Z-crotyldiphenylsilane. The requisite silane (2) was prepared
by reduction7 of 2-butynyldiphenylsilane (1)8 and, after purifi-
cation by chromatography, was found to be free of the E-isomer
as judged by examination of the 600 MHz 1H NMR spectrum.
( )-Ethyl mandelate was successfully silylated with this
reagent and the ester (3) reduced to provide the desired Z-crotyl
precursor (4) in acceptable yield (Scheme 3).
Scheme 1
Diastereomeric ene cyclisation products could arise by pro-
ton abstraction from either of the diastereotopic methyl groups
on the prenyl substituent in either a trans- or cis-decalin-like
transition structure (Scheme 2, X = Me).3 On this basis we
Scheme
3
Reagents: (i) DIBAL; H3Oϩ; (ii) PhCH(OH)CO2Et,
B(C6F5)3 cat.; (iii) DIBAL.
Efficient access to the E-analogue proved to be more
problematic since existing methods9 for E-crotylsilane synthesis
were either inapplicable to our specific requirements or gave
poor results in our hands. However, inspired by Hodgson’s
report10 that the double bond in 1,1-bis(trialkylsilyl)alkenes
could be migrated into the allylic position we proposed that it
might be possible to isomerise the double bond in readily-
prepared 3-butenylsilanes (i.e. homoallylic silanes) into the
allylic position. Indeed, there was good precedent to this
concept in the work of Matsuda in which a range of Ir() and
Rh() complexes had been evaluated for their potential in the
isomerisation of several simple tetraalkylsilanes.11 A short
model study established that commercially-available [(COD)Ir-
(PPh2Me)2]ϩPF6Ϫ was effective in achieving complete isomeris-
ation of 3-butenylsilane 5 (Scheme 4) to give E-crotylsilane 6,
even though this was not identified as the optimum pre-catalyst
Scheme 2 Prediction of improved stereoselectivity in the ene cyclis-
ation of crotylsilyl precursors (X = H) compared with that of prenylsilyl
precursors (X = Me).
† This is one of a number of contributions from the current members
of the Dyson Perrins Laboratory to mark the end of almost 90 years of
organic chemistry research in that building, as all its current academic
staff move across South Parks Road to a new purpose-built laboratory.
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 3
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 3 6 3 5 – 3 6 3 8
3635

View Article Online
Table 1
Enhancement/%
Irradiate
H1
H2
CH2CH᎐
Me3C
᎐
Scheme
4
Reagents: (i) t-BuLi; Ph2MeSiCl; (ii) [(COD)Ir(PPh2-
Me)2]ϩPF6Ϫ cat.
H1
H2
—
2.0
8.3
19.7
1.5
—
6.9
21.7
4.8
4.9
—
6.9
8.6
2.3
—
CH2CH᎐
in Matsuda’s study. Key to the success of this reaction was
careful control of activation of the pre-catalyst with hydrogen,
removal of excess hydrogen, and identification of the correct
temperature (0 ЊC) and time (ca. 30–45 min) for the isomeris-
ation; in this way, the product was obtained free of vinylsilyl
contaminants and accompanied by only ca. 2–3% of the Z-
crotylsilyl isomer.
᎐
t-Bu
4.6
Enhancement/%
Subsequently we were delighted to find that the product (8)
of silylation of ethyl mandelate with 3-butenyldiphenylsilane
(7)12 could be isomerised under these optimised conditions to
provide E-crotyl substrate 9 in high yield. DIBAL reduction of
the ester to give aldehyde 10 proceeded well to complete a
practical and efficient synthesis of the E-crotyl precursor
(Scheme 5).
Irradiate
H1
H2
CH2CH᎐
Me2CH
᎐
H1
H2
—
2.3
—
5.0
3.4
5.5
5.1
—
6.3
4.7
0.7
—
2.2
5.5
4.0
CH2CH᎐
᎐
CHMe2
—
α-silyloxyaldehyde to a pair of 2-sila-1,3-dioxolanes, which
were tentatively assigned, on the basis of subsequent results, as
being epimeric at the benzylic position, the stereochemistry at
the allylic position being determined by that of the crotyl unit
in the starting material (Scheme 6). Desilylation of these inter-
mediates gave a high yield of the diols 12 and 13 (from 4) and 15
and 16 (from 10) that could be readily separated by chromato-
graphy.
Confirmation of the stereochemistry in these compounds
was sought by correlation with simpler analogues and, initially,
unsubstituted allyl substrates were examined so that the pre-
dominant formation of siladioxolane precursors of syn-diols
could be established. Preparation of test substrates was most
easily achieved by silylation of cyanohydrins 17–2013 (Scheme 7)
with allyldiphenylchlorosilane (prepared in situ from dichloro-
diphenylsilane and allylmagnesium bromide) followed by
DIBAL reduction.14 This sequence worked well for cyano-
hydrins 17 and 18, derived respectively from pivalaldehyde and
isobutyraldehyde, but the DIBAL reductions were not product-
ive in generating the more labile aldehydes derived ultimately
from propionaldehyde and benzaldehyde. Nevertheless, the two
substrates 25 and 26 proved sufficient for the purposes of this
study, behaving as expected when heated to 80 ЊC and, after 36
h at this temperature, removal of the solvent gave a single com-
pound in each case (27 and 28 respectively), the stereochemistry
being confirmed by NOE experiments, as indicated (Table 1).
Scheme 5 Reagents: (i) PhCH(OH)CO2Et, B(C6F5)3 cat.; (ii) [(COD)-
Ir(PPh2Me)2]ϩPF6Ϫ cat.; (iii) DIBAL.
Cyclisations
Having expended considerable effort in developing short,
efficient routes to both E- and Z-crotyl precursors we were dis-
appointed to find that they were totally unsuited to ene cyclis-
ation both under the “standard” conditions (Me2AlCl, CH2Cl2,
RT) and with a variety of alternative Lewis acids at various
temperatures; in all cases the precursors either decomposed
completely or gave rise to complex product mixtures that were
of no synthetic value. To try to effect ene cyclisation under
conditions in which the substrates would not be prone to
decomposition, aldehydes 4 and 10 were heated at 80 ЊC in
CDCl3 and the reactions monitored by 1H NMR. Interestingly,
in each case, there was a smooth progression from the
Scheme 6 Reagents: (i) CHCl3, 80 ЊC, 36 h; (ii) KF, KHCO3, aq. MeOH–Et2O.
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 3 6 3 5 – 3 6 3 8
3636

View Article Online
Fig. 1
Scheme
7
Reagents: (i) Ph2(allyl)SiCl, imidazole; (ii) DIBAL;
(iii) C6H6, 80 ЊC, 36 h.
Because the tert-butyl-substituted compounds were stable,
easy to handle, gave essentially perfect stereocontrol during the
rearrangement step and simple NMR spectra with the minimal
amount of coupling of the ring protons, tert-butyl variants (31,
Scheme 8 and 34, Scheme 9) of the Z- and E-crotyl precursors
were prepared. Once again, cyclisation proceeded extremely
cleanly, this time to give single diastereoisomers of the sila-
dioxolane intermediates (not shown) which were desilylated in
the presence of hydrogen peroxide to facilitate separation of the
diols from phenylsilyl residues.
Scheme 10 Reagents: (i) NaH, allyl bromide; (ii) 36 cat., 20 ЊC, 4 h.
these unoptimised RCM reactions result from competing
cyclisation to give seven membered ring products (not shown).15
The coupling constant between the allylic and ring-oxygen
methine protons was large (J 9.4 Hz) in the case of 37 and
small (J 2.7 Hz) in the case of 38 which supported anti- and
syn-assignments respectively of the precursor diols 32 and 35.
Mechanism and stereochemistry
In view of the observed stereochemistry of the siladioxolanes,
the process seems most likely to be intramolecular and to be
sure that intermolecular processes could not compete a mixture
of tert-butyl/E-crotyl substrate 35 and isopropyl/allyl substrate
26 was heated for 36 h at 80 ЊC. Analysis of the reaction mixture
1
both by high field H NMR spectroscopy and high resolution
GC mass spectrometry revealed that only two siladioxolanes
(39 and 28) had been produced in support of a purely intra-
molecular reaction (Scheme 11).
Scheme 8 Reagents: (i) silane 2, B(C6F5)3 cat.; (ii) DIBAL; Swern
oxidation; (iii) C6H6, 80 ЊC, 36 h then KF, H2O2, aq. MeOH–Et2O.
Scheme 11 Crossover experiment.
Our working model for these reactions involves cooperative16
pre-activation of both the aldehyde and silicon atom by their
mutual coordination and subsequent allylic transfer through a
chair-like assembly from the face opposite the alkyl substituent
(Fig. 2). This arrangement results in anti-Felkin–Anh selectivity
and, because the group bound to the acyl carbon is forced into
a pseudoaxial position, the usual sense of stereoselectivity
between the allylic and newly-formed hydroxylated positions
Scheme 9 Reagents: (i) silane 7, B(C6F5)3 cat.; (ii) [(COD)Ir(PPh2-
Ϫ
Me)2]ϩPF6 cat.; (iii) DIBAL; Swern oxidation; (iv) C6H6, 80 ЊC, 36 h
then KF, H2O2, aq. MeOH–Et2O.
Whilst the relative stereochemistry at the two hydroxylated
positions was clear by NMR analysis of the intermediate
siladioxolanes, to secure the stereochemistry at the allylic
position diols 32 and 35 were bis-allylated and the allyl ethers
subjected to ring closing metathesis (RCM) with the second
generation Grubbs’ catalyst 36 (Fig. 1) to give samples of the
dihydropyrans 37 and 38 (Scheme 10). The modest yields in
(E
anti, Z
syn) is reversed when compared with nominally
intermolecular allylations that proceed via coordinated nucleo-
philic delivery within a six-membered ring transition state.17
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 3 6 3 5 – 3 6 3 8
3637

View Article Online
3 For recent pertinent examples of carbonyl ene cyclisations contain-
ing discussion of stereochemistry and mechanism under a variety of
conditions see: (a) J. T. Williams, P. S. Bahia and J. S. Snaith,
Org. Lett., 2002, 4, 3727–3730; (b) P. Kocovsky, G. Ahmed,
J. Srogl, A. V. Malkov and J. Steele, J. Org. Chem., 1999, 64, 2765–
2775.
4 B. B. Snider, M. Karras, R. T. Price and D. J. Rodini, J. Org. Chem.,
1982, 47, 4538–4545.
5 Review: G. R. Jones and Y. Landais, Tetrahedron, 1996, 52, 7599–
7662.
6 J. M. Blackwell, K. L. Foster, V. H. Beck and W. E. Piers, J. Org.
Chem., 1999, 64, 4887–4892 . See also V. Gevorgyan, M. Rubin,
S. Benson, J.-X. Liu and Y. Yamamoto, J. Org. Chem., 2000, 65,
6179–6186.
7 S. Rajagopalan and G. Zweifel, Synthesis, 1984, 113–115.
8 Prepared by the general procedure described in S. Rajagopalan and
G. Zweifel, Synthesis, 1984, 111–112.
9 E.g. by the deprotonation and silylation of E-but-2-ene: H.-J.
Knölker, E. Baum and O. Schmitt, Tetrahedron Lett., 1998, 39,
7705–7708 . We are very grateful to Professor Knölker for providing
full experimental details and useful advice concerning the deproton-
ation chemistry. Whilst trialkylchlorosilanes are trapped efficiently
under these conditions, trapping with dialkylchlorosilanes led to
significant double addition.
Fig. 2
This model predicts that the level of facial selectivity should be
determined by the effective size of the alkyl group and, indeed,
the t-Bu and i-Pr cases proceeded with excellent stereo-
selectivity in contrast to the Ph cases.
10 D. M. Hodgson, S. F. Barker, L. H. Mace and J. R. Moran, Chem.
Commun., 2001, 153–154.
11 I. Matsuda, T. Kato, S. Sato and Y. Izumi, Tetrahedron Lett., 1986,
27, 5747–5750.
12 (a) K. Kahle, P. J. Murphy, J. Scott and R. Tamagni, J. Chem. Soc.,
Perkin Trans. 1, 1997, 997–999; (b) C. Kim, S. K. Choi, E. Park and
I. Jung, J. Korean Chem. Soc., 1997, 41, 88–97.
13 L. Meerpoel and G. Hoornaert, Synthesis, 1990, 905–908 and
references cited therein.
14 M. Hayashi, T. Yoshiga, K. Nakatani, K. Ono and N. Oguni,
Tetrahedron, 1994, 50, 2821–2830.
Summary
This process is closely related to an increasing number of allyl-
ations that proceed in the absence of added catalysts. For
example, if the silicon atom is rendered sufficiently Lewis
acidic, either by virtue of attached electronegative substi-
tuents18 or by being constrained within a four membered
ring,16,19 then complexation to an aldehyde precedes allyl deliv-
ery and an ordered intramolecular reaction follows with con-
sequently high stereocontrol. More recently, Leighton has been
very active in this general field and has reported, inter alia, a
variety of elegant systems for intramolecular carbonyl allyl-
ation from silicon intermediates.20 Nevertheless, these examples
appear to be the first in which a simple diarylsilyloxy substi-
tuent activates the proximal carbonyl group in this way.
A full description of these results is currently in preparation
that will include β-silyloxyaldehyde analogues, substrates bear-
ing functionalised side-chains, and examples of α-(trialkylsilyl)-
allyl transfer.
15 Cf. L. Jafarpour, M.-P. Heck, C. Baylon, H. M. Lee,
C. Mioskowski and S. P. Nolan, Organometallics, 2002, 21, 671–
679.
16 K. Matsumoto, K. Oshima and K. Utimoto, J. Org. Chem., 1994, 59,
7152–7155.
17 For a summary of aspects of stereocontrol involving reactions
between aldehydes and allylsilane derivatives see pp. 2112–2118 in
I. Fleming, A. Barbero and D. Walter, Chem. Rev., 1997, 97, 2063–
2192.
18 (a) S. Kobayashi and K. Nishio, J. Org. Chem., 1994, 59, 6620–6628;
(b) A. Hosomi, S. Kohra, K. Ogata, T. Yanagi and Y. Tominaga,
J. Org. Chem., 1990, 55, 2415–2420; (c) M. Kira, T. Hino and
H. Sakurai, Tetrahedron Lett., 1989, 30, 1099–1102; (d ) M. Kira,
K. Sato and H. Sakurai, J. Am. Chem. Soc., 1988, 110, 4599–4602;
(e) M. Kira, M. Kobayashi and H. Sakurai, Tetrahedron Lett., 1987,
28, 4081–4084.
References
1 J. Robertson, G. O’Connor, C. L. Ringrose and D. S. Middleton,
Tetrahedron, 2000, 56, 8321–8333.
2 J. Robertson, M. J. Hall, P. M. Stafford and S. P. Green, Org. Biomol.
Chem., 2003, 1, DOI: 10.1039/b306922m.
19 For a theoretical study see K. Omoto, Y. Sawada and H. Fujimoto,
J. Am. Chem. Soc., 1996, 118, 1750–1755.
20 See, for example M. J. Zacuto, S. J. O’Malley and J. L. Leighton,
J. Am. Chem. Soc., 2002, 124, 7890–7891.
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 3 6 3 5 – 3 6 3 8
3638