α-Allylation of α-amino acids via 1,5-hydrogen atom transfer

Article abstract of DOI:10.1016/j.tetlet.2009.02.110

A straightforward method for the radical-based α-allylation of proteinogenic α-amino acids is described in which the key step involves 1,5-hydrogen atom transfer from the C-4 position of an oxazolidin-5-one.

Full text of DOI:10.1016/j.tetlet.2009.02.110

Tetrahedron Letters 50 (2009) 3400–3403
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
a-Allylation of
a-amino acids via 1,5-hydrogen atom transfer
Muhammad I. Chowdhry a, Peter N. Horton b, Michael B. Hursthouse b, Mark E. Wood a,
*
a School of Biosciences, University of Exeter, Geoffrey Pope Building, Stocker Road, Exeter, EX4 4QD, UK
b EPSRC X-Ray Crystallography Service, Department of Chemistry, University of Southampton, Highfield, Southampton, SO17 1BJ, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
A straightforward method for the radical-based a-allylation of proteinogenic a-amino acids is described
in which the key step involves 1,5-hydrogen atom transfer from the C-4 position of an oxazolidin-5-one.
Received 13 January 2009
Revised 11 February 2009
Accepted 18 February 2009
Available online 21 February 2009
Ó 2009 Elsevier Ltd. All rights reserved.
The development of general synthetic methodologies for the
preparation of highly substituted, in particular, -disubstituted
-amino acids still represents a formidable synthetic challenge.
give the required quaternary centre in product 5 with stereoselec-
tivity being controlled by the C-2 stereocentre, following Seebach’s
principle of ‘self-regeneration of stereocentres’5 (Scheme 1). From
previous studies,2 it was anticipated that there would not be a
problem with regioselectivity in the hydrogen atom transfer step.
The required oxazolidin-5-ones 1 were prepared by treatment
of the sodium salts of the pivaldehyde-derived imines of the appro-
priate amino acids 2 with 2-iodobenzoyl chloride (Scheme 2), a
modification of an existing literature procedure for the preparation
of such compounds.6
a,a
a
The majority of approaches reported to date involve the alkylation,
often asymmetric, of enolate derivatives of existing proteinogenic
amino acids.1 The inherent limitations of using such methodology
for the generation of quaternary centres in systems containing
reactive functional groups, however, often result in alanine being
the most complex amino acid that can be further functionalised.
We have recently shown that the use of a radical-based methodol-
ogy facilitates the stereocontrolled generation of quaternary cen-
Table 1 summarises the results obtained using glycine,
L-ala-
tres
a
- to nitrogen in b-amino alcohols, leading to a short
nine, -valine and -phenylalanine (entries 1 to 4, respectively). In
L
L
synthetic route to protected
The need to oxidise the functionalised b-amino alcohols to amino
acids, however, still limits the applicability of this approach and
hence, related methods for the direct
a,a-disubstituted,
a
-amino acids.2
all cases where appropriate, it was assumed that the stereochem-
istry of the original amino acid (C-4 in the products 1) remained in-
tact and the major diastereoisomer of the product 1 was the one in
which the C-2 and C-4 substituents were in a cis-relationship, in
line with previous reports for similar systems.5 The degree of ste-
reocontrol for the heterocycle synthesis followed, as expected, the
steric bulk of the amino acid 2 side-chain, with only a single diaste-
a-functionalisation of pro-
teinogenic amino acids were investigated. Radical 1,5-hydrogen
atom transfer represents a powerful method for carbon-centred
radical generation at remote sites and prior to our studies de-
scribed below, Giraud and Renaud3 showed that pendant glycine
and alanine derivatives, built into oxazolidin-4-ones containing a
C-2 2-bromobenzyl protecting-radical-translocating (PRT) group,4
could be stereoselectively
no acid functionality into this system via an
ever, still restricted the scope of this process and so our attention
turned to the direct allylation of a variety of -amino acids incor-
reoisomer being produced with L-valine (entry 3). The diastereo-
isomeric products (entries 2 and 4) proved to be inseparable
using conventional silica gel chromatography and they were,
therefore, used as a mixture in subsequent experiments.7 (Note:
No further experiments were attempted with the L-phenylala-
nine-derived oxazolidin-5-one 1, R = Bn.)
As for our previously reported radical experiments using 1,3-
oxazolidines,8,9 the efficiency of 1,5-hydrogen atom transfer from
C-4 was examined by reduction of the glycine-derived oxazoli-
a-allylated. The introduction of the ami-
a-bromoester, how-
a
porated into the ring of oxazolidin-5-ones 1 bearing an N-(2-iodo-
benzoyl) PRT group (Fig. 1).
In forming these substrates 1, it was expected that the a-stereo-
centre in the original amino acid 2 would exert some control over
the C-2 stereochemistry. Aryl radical 3 generation from the iodide
would then be followed by hydrogen atom transfer from the C-4
position of the oxazolidin-5-one to generate the key a-aminoalkyl
radical intermediate 4 which would subsequently be trapped to
tBu
O
2
N
R
O
4
5
I
O
1
* Corresponding author. Tel.: +44 1392 263450; fax: +44 1392 263434.
E-mail address: m.e.wood@exeter.ac.uk (M.E. Wood).
Figure 1. General structure of N-(2-iodobenzoyl)oxazolidin-5-ones.
0040-4039/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2009.02.110

M. I. Chowdhry et al. / Tetrahedron Letters 50 (2009) 3400–3403
tBu tBu
3401
tBu
tBu
O
I
O
O
O
H2N
CO2H
N
O
N
O
Ph
N
O
Ph
N
O
R1
H
R1
O
O
R1
O
O
R1
R1
R2
2
1
3
4
5
Scheme 1. General scheme for radical-based
a-functionalisation of
a-amino acids via 1,5-hydrogen atom transfer.
tBu
tBu
any of the desired C-4 alkylated oxazolidin-5-one product with
the reduced starting material being the major product obtained
in each case. Given the more electrophilic nature of amino acid
O
I
O
I
R
i, ii,
iii
+
N
R
O
N
R
O
-radicals compared with those obtained from amino alcohols10
H2N
CO2H
a
and the electrophilic nature of these radicalphiles, these results
were not surprising and subsequent experiments focussed on the
use of electron-rich allyltin reagents as nucleophilic radicalphiles.
Photochemical radical-generating conditions at room tempera-
ture (using a standard medium pressure mercury vapour lamp)
were found to give the best results for the C-4 allylation of the gly-
cine-derived oxazolidin-5-one 1 (R = H) with an equimolar quan-
tity of allyltributyltin, at a relatively high substrate concentration
of 130 mM in benzene (Scheme 4). It was found to be necessary
to use an equimolar quantity of AIBN, although slow addition of
allyltributyltin proved to be unnecessary, the reaction being com-
plete in 8 h.11 The desired, racemic allylated product 9 was ob-
tained in a single disatereoisomeric form in 42% yield, the trans-
relative stereochemistry between the C-2 tert-butyl and C-4 allyl
substituents being established by X-ray crystallography (Fig. 2).12
This stereochemical outcome is consistent with those obtained in
enolate alkylations using related oxazolidin-5-one substrates5,6
and it suggests that the bulky C-2 tert-butyl substituent confers fa-
2
O
O
1a
1b
Scheme 2. Reagents and conditions: (i) NaOH, H2O, EtOH; (ii) tBuCHO, pentane,
reflux; (iii) 2-iodobenzoyl chloride, CHCl3, reflux.
Table 1
Results of the preparation of N-(2-iodobenzoyl)oxazolidin-5-ones
Entry
R
Ratio 1a:1b
Yield (%)
1
2
3
4
H
—
48
50
53
34
Me
iPr
Bn
1.2:1
1:0
6:1
t Bu
O
Ph
N
O
cial selectivity on trapping of the intermediate a-aminoalkyl radi-
cal 4 (R1 = H).
D
O
Identical radical allylations were carried out with the
L-alanine
6
1 (R = CH3) and
L
-valine 1 (R = iPr) derived oxazolidin-5-ones. The
+
expected products 10a/b and 11a were obtained in 64% and 5%
yields, respectively, a 4:1 ratio of inseparable diastereoisomers
(presumed by analogy to 9 to be trans:cis with respect to the C-4
allyl and C-2 tert-butyl substituents, 10a:10b) being obtained in
the former case and only a single diastereoisomer (presumed to
be trans, 11a) in the latter (Scheme 5—only the major diastereoiso-
mer of 1, R = CH3 shown, Table 2). Clearly, the yield of product 11a
was disappointingly low but this result does demonstrate the fact
that remarkably hindered quaternary centres can be constructed
using this radical-based methodology.
t Bu
t Bu
O
I
O
N
O
N
O
D
O
O
7
1 (R=H)
+
t Bu D
O
Ph
N
O
The introduction of more highly functionalised allylic substitu-
ents was also investigated using (2-methylallyl)tributyltin (Scheme
6—only the major diastereoisomer of 1 shown where appropriate),
which is known to show high reactivity towards electrophilic radi-
cals.13 The results, summarised in Table 3, show that comparable
yields and diastereoisomer ratios were obtained in comparison
with the previous results found using allyltributyltin. Again, the
O
8
Ratio 6:7:8 = 7.2 : 2 : 0.8
Scheme 3. Reagents and conditions: Bu3SnD (1.5 equiv), AIBN, C6H6, reflux; overall
yield 94%.
L
-valine-derived oxazolidin-5-one 1 (R = iPr) gave a disappointingly
din-5-one (1, R = H) with tributyltin deuteride (Scheme 3). The ex-
pected mixture of deuterated reduction products was obtained in
an overall yield of 94%, the predominant product being the re-
quired C-4 deuterated material 6. Direct reduction accounted for
only 20% of the product mixture (compound 7) and only a very
small level of hydrogen atom transfer from C-2 (compound 8)
t Bu
t Bu
O
O
( )
N
O
Ph
N
O
I
was observed, suggesting that generation of the required C-4
a-
O
O
aminoalkyl radical intermediate 4 (R1 = H) was the major reaction
1 (R=H)
pathway.
42%
9
Attempts to trap the same glycine-derived C-4
a-aminoalkyl
radical intermediate 4 (R1 = H) using methyl acrylate and acryloni-
Scheme 4. Reagents and conditions: allyltributyltin (1 equiv), AIBN (1 equiv), C6H6,
, rt, 8 h.
trile in the presence of tributyltin hydride and AIBN failed to give
hm

3402
M. I. Chowdhry et al. / Tetrahedron Letters 50 (2009) 3400–3403
tBu
O
O
O
i, ii
+
Ph
N
O
Ph
NH
Ph
NH
CO2Me
CO2Me
R1
R2
9 (R1=H, R2=H)
12a/b (R1=H, R2=Me)
13a/b (R1=Me, R2=Me)
R1
R2
R1
R2
O
15a (R1=H, R2=H)
16a (R1=H, R2=Me)
15b (R1=H, R2=H)
16b (R1=H, R2=Me)
17a (R1=Me, R2=Me) 17b (R1=Me, R2=Me)
Scheme 7. Reagents and conditions: (i) LiOH (3 equiv), H2O, MeOH, 40 °C, 20 h; (ii)
Me3SiCHN2, hexane:C6H6:MeOH (1:3:1 v/v), rt.
Table 4
Results of the oxazolidin-5-one hydrolysis/amino acid esterification
Entry
R1
R2
Product ratio
ee/%
Yield (%)
1
2
3
H
H
Me
H
Me
Me
15a:15b 1:1
16a:16b 1:1
17a:17b 1.6:1
0
0
23
98
98
92
Figure 2. ORTEP representation of 9; ellipsoids drawn at 30% probability level.
t Bu
t Bu
t Bu
O
I
O
O
+
N
R
O
Ph
N
O
Ph
N
O
products has the potential for conversion into a range of biologi-
cally important heterocycles,14 expanding the scope of highly
substituted amino acids that can be accessed using this
methodology.
O
O
O
R
R
1
10a (R=Me)
11a (R=iPr)
10b (R=Me)
11b (R=iPr)
In contrast to the harsh acidic conditions required to hydrolyse
the N-benzoyl-1,3-oxazolidines prepared in previous studies,2 the
C-4 allylated oxazolidin-5-ones were readily cleaved by alkaline
hydrolysis with lithium hydroxide. In order to facilitate the isola-
tion of the N-benzoylated amino acid products, the free acids thus
obtained were converted into their methyl esters using (trimethyl-
silyl)diazomethane and excellent yields of the desired, fully pro-
tected amino acids were obtained after the two reaction steps.
Scheme 7 and Table 415 illustrate and summarise the results.
In summary, we have shown through these preliminary studies
Scheme 5. Reagents and conditions: allyltributyltin (1 equiv), AIBN (1 equiv), C6H6,
, rt, 8 h.
h
m
Table 2
Results of the radical allylation of C-4 alkyl N-(2-iodobenzoyl)oxazolidin-5-ones
Entry
R
Product ratio
Yield (%)
1
2
Me
iPr
10a:10b 4:1
11a:11b 1:0
64
5
that it is possible to
rectly via 1,5-hydrogen atom transfer using allyltributylstannanes.
This leads to the construction of highly substituted -disubsti-
a-allylate proteinogenic a-amino acids di-
a,a
tBu
tBu
tBu
tuted (quaternary) amino acids in a minimal number of steps. Fur-
ther work is required in order to improve and optimise the
allylation step; however, the methodology should be applicable
to a wide range of amino acid substrates.
O
I
O
O
+
N
R
O
O
Ph
N
O
Ph
N
O
O
O
R
R
Acknowledgement
1
12a (R=H)
12b (R=H)
We thank the EPSRC National Mass Spectrometry Service Centre
(Swansea, UK) for mass spectra, essential to these studies.
13a (R=Me)
13b (R=Me)
14a (R=iPr)
14b (R=iPr)
Scheme 6. Reagents and conditions: (2-methylallyl)tributyltin (1 equiv), AIBN
(1 equiv), C6H6, hm, rt, 8 h.
References and notes
1. Cativiela, C.; Díaz-de-Villegas, M. D. Tetrahedron: Asymmetry 1998, 9, 3517–
3599; Vogt, H.; Bräse, S. Org. Biomol. Chem. 2007, 5, 406–430.
2. Wood, M. E.; Penny, M. J.; Steere, J. S.; Horton, P. N.; Light, M. E.; Hursthouse, M.
B. Chem. Commun. 2006, 2983–2985.
Table 3
Results of the radical 2-methylallylation of N-(2-iodobenzoyl)oxazolidin-5-ones
3. Giraud, L.; Renaud, P. J. Org. Chem. 1998, 63, 9162–9163.
4. Snieckus, V.; Cuevas, J.-C.; Sloan, C. P.; Liu, H.; Curran, D. P. J. Am. Chem. Soc.
1990, 112, 896–898.
5. Seebach, D.; Sting, A. R.; Hoffmann, M. Angew. Chem., Int. Ed. 1996, 35, 2708–
2748.
Entry
R
Product ratio
Yield (%)
1
2
3
H
12a:12b 1:0 ( )
13a:13b 1.6:1
14a:14b 1.6:1
43
39
3
Me
iPr
6. O’Donnell, M. J.; Fang, Z.; Ma, X.; Huffman, J. C. Heterocycles 1997, 46, 617–630.
7. Diastereoisomer ratios were determined by 1H NMR.
8. Gosain, R.; Norrish, A. M.; Wood, M. E. Tetrahedron Lett. 1999, 40, 6673–6676.
9. Gosain, R.; Norrish, A. M.; Wood, M. E. Tetrahedron 2001, 57, 1399–1410.
10. Easton, C. J. Chem. Rev. 1997, 97, 53–82.
low yield of products 14a and 14b, and as before, the relative ste-
reochemistry between the C-2 tert-butyl and C-4 2-methylallyl
substituents was presumed to be trans in the major products. Sig-
nificantly, the 2-methylallyl functionality in the radical allylation
11. Typical procedure: Allyltributyltin (440 mg, 1.33 mmol) and 2,20-
azobisisobutyronitrile (220 mg, 1.34 mmol) were added to
solution of N-(2-iodobenzoyl)-2-tert-butyl-1,3,-oxazolidine
a
degassed
(500 mg,
1.34 mmol) in benzene (10 cm3) in a quartz tube. The reaction mixture was

M. I. Chowdhry et al. / Tetrahedron Letters 50 (2009) 3400–3403
3403
t
stirred at room temperature and irradiated with a medium pressure mercury
vapour lamp for 8 h. The solvent was removed in vacuo and the residue was
dissolved in diethyl ether and stirred vigorously with a 10% w/v aqueous
solution of potassium fluoride for 1 h. The separated organic phase was
evaporated in vacuo and the residue obtained was purified by column
chromatography on silica gel containing 10% w/w potassium fluoride,16
eluting with 75% petroleum ether (bp 40–60 °C): 25% ethyl acetate. This gave
9 (160 mg, 42%) as a white, crystalline solid (found MH+ (ES+) 288.1596,
(oxazolidin-5-one C@O); m/z (EI) 230 ((Mꢀ Bu)+, 8%), 105 (100), 77 (54), 57
(48), 51 (13) and 41 (49); m/z (CI) 305 (M+NH4+, 13%), 289 (19), 288 (M+H+,
100), 105 (15), 96 (8), 86 (10) and 79 (4).
12. Crystal data for 9: C17H21NO3, M = 287.35, monoclinic, space group P21,
a = 5.9385(8) Å, b = 13.0338(19) Å, c = 10.4518 (16) Å, b = 104.583(6)°,
U = 782.9(2) Å3, Dc = 1.219 Mg mꢀ3, Z = 2, Mo–K
a
radiation (k = 0.71073 Å),
l
= 0.083 mm-1 T = 120(2) K, 1880 observed reflections (Rint = 0.1014)
R1 = 0.0579 [I > 2r(I)], wR2 = 0.1537 (all data). Crystallographic data
C17H22NO3 requires 288.1594);
m
max(thin film)/cmꢀ1 3040–2765 (m) (C–H),
(excluding structure factors) have been deposited with the Cambridge
Crystallographic Data Centre as supplementary publication number no. CCDC
716085. Copies of the data can be obtained free of charge on application to
CCDC, 12 Union Road, Cambridge, CB2 1EZ, UK (fax: +44 1223 336033 or e-
mail: deposit@ccdc.cam.ac.uk).
1788 (s) (oxazolidin-5-one C@O), 1634 (s) (amide C@O), 1445 (m), 1338 (s),
1240 (m), 1189 (m), 1148 (m), 1045 (m), 1015 (m), 928 (m), 871 (w), 794 (w),
722 (m) and 651 (m); dH (400 MHz; CDCl3) 1.03 (9H, s, (CH3)3C), 1.99 (1H, br m,
CHaHbCH@CH2), 2.47 (1H, br m, CHaHbCH@CH2), 4.50 (1H, dd, J = 1.2 and
5.1 Hz, CNHC@O), 4.96 (1H, d, J = 17.8 Hz, CH@CHaHb trans), 5.15 (1H, d,
J = 10.2 Hz, CH@CHaHb cis), 5.46 (1H, br m, CH@CH2), 6.16 (1H, br s, NCHO),
7.50 (2H, ca q, J = 7.5 Hz, phenylCH), 7.57 (1H, ca t, J = 7.5 Hz, phenylCH) and
7.64 (2H, d, J = 7.0 Hz, phenylCH); dC (100 MHz; CDCl3; some resonances not
resolved through line broadening) 24.7 ((CH3)3C)), 39.8 ((CH3)3C), 95.0 (NCHO),
121.7 (CH@CH2), 127.7, 129.0, 132.1 (phenylCH and CH@CH2) and 172.6
13. Renaud, P.; Gerster, M.; Ribezzo, M. Chimia 1994, 48, 366–369.
14. Baldwin, J. E.; Fryer, A. M.; Pritchard, G. J. J. Org. Chem. 2001, 66, 2588–2596.
15. Enantiomer ratios were determined by HPLC using the chiral support
ChiralcelÒ ODTM with elution by appropriate hexane/propan-2-ol mixtures
and detection at 254 nm.
16. Harrowven, D. C.; Guy, I. L. Chem. Commun. 2004, 1968–1969.