Tris(trimethylsilyl)silane: An unprecedented enhancement in the diastereoselectivity of radical cyclisations to give 2,4-disubstituted piperidines

Article abstract of DOI:10.1039/b409714a

Cyclisation of bromides 4a-f mediated by tributyltin hydride affords predominantly the trans piperidines 5a-f with modest diastereomeric ratios, while cyclisation with tris(trimethylsilyl)silane affords the same products with diastereomeric ratios of up t

Full text of DOI:10.1039/b409714a

View Article Online / Journal Homepage / Table of Contents for this issue
C O M M U N I C A T I O N
Tris(trimethylsilyl)silane: an unprecedented enhancement in the
diastereoselectivity of radical cyclisations to give 2,4-disubstituted
piperidines
Lucile A. Gandon, Alexander G. Russell and John S. Snaith*
School of Chemistry, The University of Birmingham, Edgbaston, Birmingham, UK B15 2TT.
E-mail: j.s.snaith@bham.ac.uk; Fax: +44 121 414 4403; Tel: +44 121 414 4363
Received 28th June 2004, Accepted 2nd July 2004
First published as an Advance Article on the web 19th July 2004
Cyclisation of bromides 4a–f mediated by tributyltin hydride
affords predominantly the trans piperidines 5a–f with modest
diastereomeric ratios, while cyclisation with tris(trimethylsilyl)-
silane affords the same products with diastereomeric ratios of
up to 99:1.
Tris(trimethylsilyl)silane (TTMSS) has emerged as an important
hydrogen atom source in radical chemistry.1 Advantages over the
more commonly used tributyltin hydride (TBTH) include lower
toxicity and a reduced rate constant for hydrogen atom donation.
The latter can be beneficial in radical cyclisation reactions, giv-
ing improved selectivity for products of cyclisation rather than
direct reduction. We were keen to capitalise on the advantages of
TTMSS-mediated radical cyclisations in the synthesis of nitrogen
heterocycles, specifically 2,4-disubstituted piperidines, highlighted
as an important scaffold for drug discovery,2 and forming the core
of several pharmaceuticals.3
Free radical cyclisations are an attractive method for heterocyclic
Scheme 1 Synthesis of cyclisation precursors. (a) I(CH2)2CH(OCH3)2,
synthesis,4 and they have formed the basis of a number of piperidine
syntheses,5 although the use of TTMSS has received scant atten-
tion.6 It was envisaged that cyclisation of a primary alkyl radical
onto an activated double bond would form the C3–C4 bond of
the piperidine, with the substituent adjacent to nitrogen exerting
stereocontrol over the forming stereogenic centre at C4. The C2
substituent of the piperidine can be derived from the wide variety of
commercially available natural and unnatural amino acids.
NaH, DMF, 70 °C, 59–70%; (b) 2 M HCl, THF, H2O, 25 °C, 67–80%;
(c) Ph3PCHCO2R′, C6H5CO2H, C6H5CH3, 90 °C, 78–90%; (d) Ph3P, CBr4,
CH2Cl2, 25 °C, 76–99%; (e) MsCl, Et3N, CH2Cl2, 0 °C, then LiBr, THF,
reflux, 75–95%.
Table 1 Cyclisations of 4a–f with TBTHa
The cyclisation precursors were synthesised in 4 steps, Scheme 1.
N-Alkylation of the N-tosyl amino alcohols7 1a–c with the dimethyl
acetal of 3-iodopropanal,8 followed by acetal hydrolysis resulted
in cyclisation to an epimeric mixture of hemiacetals 2a–c, which
were directly subjected to the Wittig reaction to afford the ,-
unsaturated methyl and tert-butyl esters 3a–f in excellent yield. The
E:Z selectivity in the Wittig reactions performed under standard
conditions was surprisingly poor (typically 3:1 E:Z), but could be
significantly improved by the introduction of benzoic acid under
the modified conditions reported by Martin;9 it was convenient to
carry the E:Z mixture forward and separate at the bromide stage.
The methyl esters 3a–c were converted into the corresponding
bromides 4a–c by treatment with carbon tetrabromide and tri-
phenylphosphine. This procedure proved to be unreliable and low
yielding for the tert-butyl esters, which were instead converted into
the bromides 4d–f by a two-step procedure involving mesylation of
the alcohol followed by displacement with lithium bromide.
We first explored the cyclisation of the bromides using TBTH.
Syringe pump addition of TBTH and AIBN to a solution of the
substrate in refluxing benzene gave excellent yields of the dia-
stereomeric piperidines 5a–f and 6a–f, Table 1. Virtually identical
results were obtained on substituting toluene (at 90 °C) for the more
toxic benzene. The major diastereomer 5 was identified as the trans
product from a combination of 1H–1H coupling constants and NOE
measurements. This product arises from the preference of the 2-
substituent to adopt an axial disposition in the chair-like transition
state, thus avoiding the pseudo A1,3 strain with the sulfonamide;10
the radical cyclises onto an equatorial olefin, Fig. 1. The minor
isomer was the cis piperidine 6, resulting from a transition state in
Entry
Bromidea
R
R′
5:6b
Yield (%)c
1
2
3
4
5
6
4a
4b
4c
4d
4e
4f
Me
Bn
iPr
Me
Bn
iPr
Me
Me
Me
tBu
tBu
tBu
70:30
80:20
86:14
75:25
83:17
86:14
98
89
95
99
85
84
a Reactions were performed by syringe pump (10 h) addition of benzene
solutions of Bu3SnH and AIBN to a solution of the bromide in refluxing
benzene (final concentration 0.015 M). b Ratio determined by 1H NMR (4a,
4b, 4d) or HPLC after chromatography to remove tin residues. c Isolated
yields following chromatography.
which the 2-substituent adopts an equatorial disposition. Increasing
the bulk of the 2-substituent increases the pseudo A1,3 strain and
favours the trans product (entries 1–3), while the bulk of the ester
does not appear to have a significant effect on the stereoselectivity
(entries 1–3 versus entries 4–6).
Moving on to TTMSS, we cyclised the esters 4a–f in toluene at
90 °C, with addition of TTMSS by syringe pump, Table 2.† Cyclisa-
tion yields were a little lower than those obtained with TBTH, but it
2 2 7 0
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2 , 2 2 7 0 – 2 2 7 1
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 4

View Article Online
Acknowledgements
We thank the Engineering and Physical Sciences Research Council
for financial support (grant GR/M70964/01 and studentship to
A.G.R.).
Fig. 1 Proposed pathway to major isomer.
Notes and references
Table 2 Cyclisations with TTMSSa
† TTMSS cyclisation procedure. Preparation of (2R,4R)-4-(tert-butoxy-
carbonylmethyl)-2-isopropyl-1-(p-toluenesulfonyl)-piperidine (5f). To
a stirred solution of the bromide 4f (0.222 g, 0.468 mmol) in degassed
toluene (10 mL) at 90 °C under Ar were added solutions of AIBN (0.008 g,
0.049 mmol) in toluene (10 mL) and TTMSS (0.26 mL, 0.842 mmol) in
toluene (10 mL) by syringe pump at a rate of 0.5 mL h−1. The solution was
stirred at 90 °C for a further 12 h. The solvent was removed in vacuo and
the residue purified by column chromatography (petrol:ethyl acetate, 8:1,
Rf = 0.25) to afford the title compound as a colourless oil. (0.13 g, 75%)
[]D18 +34.9 (c 0.64 in CHCl3); (Found: C, 63.8; H, 8.5; N, 3.5. C21H33NO4S
requires C, 63.8; H, 8.4; N, 3.5%); max(film)/cm−1 2970 (C–H), 2872
(C–H), 1726 (CO), 1598 (CC aromatic), 1494 (CC aromatic), 1454,
1391, 1367, 1337, 1259, 1204, 1152, 1093, 1054; H(300 MHz, CDCl3)
0.71–0.96 (8H, envelope), 1.33–1.48 (10H, envelope), 1.73 (1H, d, J 11.4),
1.77–2.13 (4H, envelope), 2.40 (3H, s), 2.91–3.05 (1H, m), 3.53–3.62 (1H,
m), 3.74–3.86 (1H, m), 7.25 (2H, d, J 8.3), 7.69 (2H, d, J 8.3); C(125 MHz,
CDCl3) 19.9, 20.2, 21.5, 26.6, 27.4, 28.0, 30.3, 31.5, 40.7, 42.7, 59.5,
80.4, 126.9, 129.5, 139.0, 142.7, 171.4; m/z (ES+) 418 (34%, [M + Na]+),
362 (100, [M − CH2C(CH3)2]+) [HRMS Found: (M + Na)+ 418.2029.
C21H33NNaO4S requires M, 418.2028].
Entry
Bromidea
R
R′
5:6b
Yield (%)c
1
2
3
4
5
6
4a
4d
4b
4e
4c
4f
Me
Me
Bn
Bn
iPr
Me
tBu
Me
tBu
Me
tBu
72:28
77:23
92:8
96:4
97:3
99:1
97
86
76
63
73
75
iPr
a Reactions were performed by syringe pump (10 h) addition of toluene
solutions of AIBN and TTMSS to a solution of the bromide in toluene at
90 °C (final concentration 0.015 M). b Ratio determined by H NMR (4a,
1
4b, 4d) or HPLC both before and after chromatography. c Isolated yields
following chromatography.
was in the diastereoselectivities that the difference between the two
reagents was most noticeable.
In the examples with the methyl 2-substituent (entries 1 & 2), the
diastereoselectivities with TTMSS were similar to those obtained
with TBTH, but in all other cases there was a marked improvement
on switching from TBTH to TTMSS. Thus benzyl derivative 4b
favoured the trans product 5b with a dr of 92:8 (entry 3), and this
increased to 97:3 for the isopropyl derivative 4c (entry 5). Interest-
ingly, the tert-butyl esters of these two derivatives exhibited higher
stereoselectivities. For the benzyl compound 4e, the dr was 96:4 in
favour of trans piperidine 5e, whilst for the isopropyl derivative 4f,
the dr rose to 99:1 (entries 4 & 6).
Such an enhancement in the stereoselectivity of the cyclisation
process on switching from TBTH to TTMSS is remarkable
and apparently without precedent. Others have noted increased
diastereoselectivity in the reduction of halides on switching from
TBTH to TTMSS,11 an observation attributed to the increased
preference for the bulkier TTMSS to approach the intermediate
radical from the less hindered face. This explanation cannot
account for the enhanced diastereoselectivities observed during the
cyclisation of 4a–f.
It is very unlikely that the increased diastereoselectivities are a
result of the cyclisation occurring under thermodynamic control.
Reversibility in radical cyclisations is very unusual, and most
examples are associated with the cyclisation of highly stabilised
radicals in conjunction with slow trapping by poor atom donors.12
Reversibility in this case would be energetically very unfavour-
able as the radical produced by the cyclisation is much more stable
(due to delocalisation with the ester) than the initial primary alkyl
radical.
1 C. Chatgilialoglu, D. Griller and M. Lesage, J. Org. Chem., 1988,
53, 3641; C. Chatgilialoglu, C. Ferreri and T. Gimisis, The Chemistry
of Organic Silicon Compounds, volume 2, Ed. Z. Rappoport and
Y. Apeloig, Wiley, Chichester, UK, 1998.
2 P. S. Watson, B. Jiang and B. Scott, Org. Lett., 2000, 2, 3679.
3 Palinavir: P. C. Anderson, F. Soucy, C. Yoakim, P. Lavallee and
P. L. Beaulieu, European Patent EP0560268, 1993; Selfotel:
A. J. Hutchison, M. Williams, C. Angst, R. deJesus, L. Blanchard,
R. H. Jackson, E. J. Wilusz, D. E. Murphy, P. S. Bernard, J. Schneider,
T. Campbell, W. Guida and M. A. Sills, J. Med. Chem., 1989, 32,
2171; Argatroban: S. Okamoto, A. Hijikata, R. Kikumoto, Y. Tamao,
K. Ohkubo, T. Tezuka and S. Tonamura, US Patent 4101653, 1978.
4 For a review of radical cyclisations applied to the synthesis of hetero-
cycles see: W. R. Bowman, A. J. Fletcher and G. B. S. Potts, J. Chem.
Soc., Perkin Trans. 1, 2002, 2747.
5 A. R. Katritzky, Z. Luo, Y. Fang, D. Feng and I. Ghiviriga, J. Chem.
Soc., Perkin Trans. 2, 2000, 1375; M. Koreeda, Y. Wang and L. Zhang,
Org. Lett., 2002, 4, 3329; S. E. Yoo, K. Y. Yi, S. H. Lee and N. Jeong,
Synlett, 1990, 575; A. F. Parsons and R. M. Pettifer, J. Chem. Soc.,
Perkin Trans., 1, 1998, 651; M. Ihara, F. Setsu, M. Shoda, N. Taniguchi,
Y. Tokunaga and K. Fukumoto, J. Org. Chem., 1994, 59, 5317; E. Lee,
T. S. Kang, B. J. Joo, J. S. Tae, K. S. Li and C. K. Chung, Tetrahedron
Lett., 1995, 36, 417; E. Lee, E. J. Jeong, S. J. Min, S. Hong, J. Lim,
S. K. Kim, B. G. Choi and K. C. Koo, Org. Lett., 2000, 2, 2169;
N. Prevost and M. Shipman, Org. Lett., 2001, 3, 2383.
6 For a recent example of the use of TTMSS in the synthesis of oxygen
heterocycles see: P. A. Evans and T. Manangan, J. Org. Chem., 2000, 65,
4523.
7 M. B. Berry and D. Craig, Synlett, 1992, 41; . The amino alcohols were
either obtained commercially or prepared by reduction of the correspond-
ing amino acid by the method of Meyers: M. J. McKennon,A. I. Meyers,
K. Drauz and M. Schwarm, J. Org. Chem., 1993, 58, 3568.
We believe that the reduced rate constant for hydrogen atom
donation by TTMSS compared with TBTH could allow the product
radicals to undergo side reactions before they are trapped; it is
reasonable to assume that these could be different, or could occur at
different rates, for the two diastereomers. The slightly lower yields
obtained in the TTMSS cyclisations, coupled with the improved
diastereoselectivities, suggest that the minor diastereomer may be
decomposing by some, as yet unidentified, pathway.
In summary, we have discovered a highly diastereoselective
synthesis of trans 2,4-disubstituted piperidines from simple acyclic
precursors, which should have application to the synthesis of
more complex molecules. Experimental and theoretical studies
are currently underway in an effort to elucidate the origins of the
difference in stereoselectivity between TTMSS and TBTH.
8 For a related example see: J. A. Gomez, M. A. Trujillo, J. Campos,
M. A. Gallo andA. Espinosa, Tetrahedron, 1998, 54, 13295; . For prepa-
ration of the iodoacetal see: D. L. J. Clive, C. Chua Paul and Z. Wang,
J. Org. Chem., 1997, 62, 7028.
9 C. Harcken and S. F. Martin, Org. Lett., 2001, 3, 3591.
10 F. Johnson, Chem. Rev., 1968, 68, 375.
11 Y. Apeloig and M. Nakash, J. Am. Chem. Soc., 1994, 116, 10781;
E. Kawashima, S. Uchida, M. Miyahara and Y. Ishido, Tetrahedron
Lett., 1997, 38, 7369; K. Mima, T. Ishihara, S. Kuwahata, T. Konno and
H. Yamanaka, Tetrahedron, 2002, 58, 2369.
12 M. Julia, Acc. Chem. Res., 1971, 4, 386; D. P. Curran, T. M. Morgan,
C. E. Schwartz, B. B. Snider and M. A. Dombroski, J. Am. Chem.
Soc., 1991, 113, 6607; K. Sung and Y. Y. Wang, J. Org. Chem., 2003,
68, 2771; . The 5-exo cyclisation of highly stabilised allylic O-stannyl
ketyls with TBTH is also believed to be reversible: E. J. Enholm and
K. S. Kinter, J. Org. Chem., 1995, 60, 4850.
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2 , 2 2 7 0 – 2 2 7 1
2 2 7 1