Strategic use of amino acid N-substituents to limit α-carbon- centered radical formation and consequent loss of stereochemical integrity

Article abstract of DOI:10.1016/S0957-4166(03)00542-1

Ab initio calculations have been used to investigate the effect of N-substituents on the stability of α-carbon-centered amino acid radicals. Optimized structures of glycine derivatives and related substituted methanes, and the corresponding radicals, were determined with B3-LYP/6-31G(d). Single-point RMP2/6-31G(d) calculations on these structures were then used to obtain radical stabilization energies, which were compared with the relative rates of formation of the same or closely similar radicals in reactions with N-bromosuccinimde. These studies show that N-acylation and sulfonation decrease both the stability and the ease of formation of the α-carbon-centered radicals. Greater effects are seen with fluoroacyl, fluoroalkylsulfonyl and imido groups. The extent of the effect of the imido and fluoroalkylsulfonyl groups is such that N-phthaloyl- and trifluoromethanesulfonyl-protected amino acids react by hydrogen-atom abstraction from the side chain, thereby avoiding reaction at the chiral α-center and preserving its stereochemical integrity. The origins of these substituent effects are examined.

Full text of DOI:10.1016/S0957-4166(03)00542-1

TETRAHEDRON:
ASYMMETRY
Pergamon
Tetrahedron: Asymmetry 14 (2003) 2919–2926
Strategic use of amino acid N-substituents to limit
a-carbon-centered radical formation and consequent loss of
stereochemical integrity
Anna K. Croft,† Christopher J. Easton,* Katherine Kociuba and Leo Radom‡
Research School of Chemistry, Australian National University, Canberra, ACT 0200, Australia
Received 13 May 2003; accepted 13 June 2003
Abstract—Ab initio calculations have been used to investigate the effect of N-substituents on the stability of a-carbon-centered
amino acid radicals. Optimized structures of glycine derivatives and related substituted methanes, and the corresponding radicals,
were determined with B3-LYP/6-31G(d). Single-point RMP2/6-31G(d) calculations on these structures were then used to obtain
radical stabilization energies, which were compared with the relative rates of formation of the same or closely similar radicals in
reactions with N-bromosuccinimde. These studies show that N-acylation and sulfonation decrease both the stability and the ease
of formation of the a-carbon-centered radicals. Greater effects are seen with fluoroacyl, fluoroalkylsulfonyl and imido groups. The
extent of the effect of the imido and fluoroalkylsulfonyl groups is such that N-phthaloyl- and trifluoromethanesulfonyl-protected
amino acids react by hydrogen-atom abstraction from the side chain, thereby avoiding reaction at the chiral a-center and
preserving its stereochemical integrity. The origins of these substituent effects are examined.
© 2003 Elsevier Ltd. All rights reserved.
1. Introduction
the corresponding bromides 1b and 2b, without loss of
stereochemistry. While it is clear that the phthaloyl
substituent deactivates the a-position of amino acids
towards hydrogen abstraction, the effect has not been
quantified and its origins have not been examined. We
therefore set out to investigate this phthaloyl group
effect using a combination of theoretical and experi-
mental methods. During the course of this work, we
identified the trifluoromethanesulfonyl (triflyl) group as
another N-substituent that significantly deactivates the
a-position of amino acids towards radical formation,
and we also evaluate this finding.
The proteinogenic a-amino acids and their derivatives
constitute a readily available pool of optically active
starting materials for asymmetric synthesis.1,2 They are
particularly suitable for the preparation of a wide range
of the non-proteinogenic amino acids that occur natu-
rally, and have been shown to be active as enzyme
inhibitors and other moderators of physiological activ-
ity. One way to elaborate a-amino acids and their
derivatives is through the use of free radical chem-
istry,3–5 but the principal limitation of this approach is
the tendency for hydrogen-atom-transfer reactions of
these compounds to afford a-carbon-centered radicals
as a result of their stability. Generally this leads to the
destruction of the stereochemical integrity of the sub-
strate. To overcome this, we have exploited the N-
phthaloyl protecting group for the side-chain
functionalisation of amino acids.6–11 We have shown,
for example, that reaction of the derivatives of valine 1a
and leucine 2a with N-bromosuccinimide (NBS) gives
2. Results
* Corresponding author. Tel.: +61-2-6125-8201; fax: +61-2-6125-8114;
e-mail: easton@rsc.anu.edu.au
2.1. Calculation of structures and radical stabilization
energies
† Present address: Department of Chemistry, University of Wales,
Bangor, Gwynedd LL57 2UW, UK.
‡ Present address: School of Chemistry, University of Sydney, Syd-
ney, NSW 2006, Australia.
The optimized structures of the glycine derivatives and
0957-4166/$ - see front matter © 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/S0957-4166(03)00542-1

2920
A. K. Croft et al. / Tetrahedron: Asymmetry 14 (2003) 2919–2926
Table 1. Radical stabilization energies [0 K, RMP2/6-
31G(d), kJ mol−1] of substituted glycyl radicals and
related substituted methyl radicals
the related substituted methanes 3a–g, 5a–g, 9a,b, 11a,b
and 13, and the corresponding radicals 4a–g, 6a–g,
10a,b, 12a,b and 14, were determined at the B3-LYP/6-
31G(d) level. They are displayed in Figure 1 in the cases
of the radicals 10a,b and 12a,b. Radical stabilization
energies (RSEs) were calculated as the energy changes
in the isodesmic reactions shown in Eq. (1):
’
’
R +CH4
“ RH+CH3
RSE
’
4a (NH2CH2 )
4b (HCONHCH2 )
3a
3b
3c
3d
3e
3f
3g
9a
9b
13
5a
5b
5c
5d
5e
5f
44.7a
36.9
37.4a
33.7
28.7
28.0
25.4
28.5
28.8
21.6a
94.0
79.6
82.2a
72.8
72.5
74.0
66.0
33.2
46.0
’
’
’
R +CH4“RH+CH3
(1)
’
’
4c (CH3CONHCH2 )
4d (CF3CONHCH2 )
4e (HSO2NHCH2 )
4f (CH3SO2NHCH2 )
4g (CF3SO2NHCH2 )
10a (MaleimidoCH2 )
10b (MaleimidoCH CH3)
14 (CH2 CO2CH3)
6a (NH2CH CO2CH3)
6b (HCONHCH CO2CH3)
6c (CH3CONHCH CO2CH3)
6d (CF3CONHCH CO2CH3)
6e (HSO2NHCH CO2CH3)
6f (CH3SO2NHCH CO2CH3)
6g (CF3SO2NHCH CO2CH3)
12a (MaleimidoCH CO2CH3)
The RSEs correspond to the differences between the
’
CꢀH bond dissociation energies of methane and
’
’
RH,12,13 and reflect the stability of R compared with
’
’
’
CH3 (relative to the corresponding closed-shell
’
molecules). Our reported RSEs were determined
through single-point RMP2/6-31G(d) calculations on
the B3-LYP/6-31G(d) optimized structures, and corre-
spond to values at 0 K. Temperature corrections were
not applied since they should show substantial cancella-
tion for such reactions.14 Details of the theoretical
’
’
’
’
’
’
’
’
’
5g
11a
11b
’
12b (MaleimidoC (CH3)CO2CH3)
a These data are consistent with literature values.12,14,15
results (presented as Gaussian archive files) are avail-
able as Supplementary Information from the authors.
The RSEs for the radicals 4a–g, 6a–g, 10a,b, 12a,b and
14 are presented in Table 1.
The glycyl radical 6a was studied as a simple example
of an a-carbon-centered amino acid radical. The radi-
cals 6b–f were included for comparison in order to
investigate the effects of acylation and sulfonation of
the amino group. The maleimides 12a,b were examined
as simple models for N-phthaloylamino acid radicals,
with and without a side chain. The related methyl-sub-
stituted radicals 4a–g and 10a,b are analogous to the
glycine derivatives 6a–g and 12a,b, except that they lack
the methoxycarbonyl group. They were examined to
determine the effect of the methoxycarbonyl substituent
in combination with amino and substituted amino
groups. The acetate radical 14 was also studied, to
Figure 1. Optimized structures [B3-LYP/6-31G(d)] of a) the
maleimidomethyl radical 10a, b) the N-maleylglycyl radical
12a, c) the maleimidoethyl radical 10b, and d) the N-maleyl-
alanyl radical 12b.

A. K. Croft et al. / Tetrahedron: Asymmetry 14 (2003) 2919–2926
2921
examine the effect of the methoxycarbonyl substituent
alone.
brominating reagent15 and the results are shown in
Table 2.
2.2. Measurement of reaction rates
The reactions of the glycine derivatives 5c, 5d, 5g, 5h,
5i, 5j and 15 with NBS in refluxing carbon tetrachloride
under nitrogen were also studied. The amides 5c and
5d, and the sulfonamide 5g, were examined in order to
make a direct comparison between the rates of reaction
to give the radicals 6c, 6d and 6g and the RSEs of these
radicals. It was not practical to study the glycine ester
5a since the free amino group is incompatible with
these reaction conditions. The phthalimide 15 was
included to compare its rate of reaction with the RSE
of the analogous maleimide 12a. The benzamides 5h, 5i
and 5j were also investigated, as a series where the
effect of the fluorines on the ease of formation of the
radicals 6h, 6i and 6j could be probed. The amides 5c,
5d, 5h, 5i and 5j gave the corresponding bromides 7c,
7d, 7h, 7i and 7j, presumably via the radicals 6c, 6d, 6h,
6i and 6j, respectively. The bromides 7c, 7d, 7h, 7i and
7j were unstable so they were characterized as the
corresponding a-methoxyglycine derivatives 8c, 8d, 8h,
8i and 8j.16 Reaction of the sulfonamide 5g gave the
succinimidylglycine 18. Presumably the bromide 7g is
formed via the radical 6g, but loses HBr more rapidly
than the corresponding bromoglycines 7c, 7d, 7h, 7i and
7j, due to the greater acidity of the sulfonamide hydro-
gen. Succinimide then adds to the product sul-
fonylimine to give 18.16 Under the reaction conditions
used, the phthaloylglycine derivative 15 was inert; none
of the bromide 17a or the methoxyglycine derivative
17b was detected, indicating that the radical 16 does not
form. The relative rates of reaction of the glycine
derivatives 5c, 5d, 5g, 5h, 5i, 5j and 15 to give the
radicals 6c, 6d, 6g, 6h, 6i, 6j and 16 were determined in
competitive experiments using limited quantities of the
3. Discussion
The RSEs calculated in the present work for the radi-
cals 4a, 4c, 6c and 14, are consistent with literature
data.12,14,15 They show that the amino substituent in the
radical 4a has a stabilizing effect of 44.7 kJ mol−1.
N-Acylation and the consequent decrease in the p-elec-
tron-donating ability of the amido substituent reduces
the RSE by 7–8 kJ mol−1, to 36.9 and 37.4 kJ mol−1 for
the formamidomethyl and acetamidomethyl radicals 4b
and 4c, respectively. The even greater decrease in the
electron-donating ability of an imido substituent fur-
ther reduces the RSE by about 8–9 kJ mol−1, to 28.5 kJ
mol−1 for the maleimide 10a. For similar reasons, N-
sulfonation reduces radical stability more effectively
than acylation, by about 8–10 kJ mol−1, as illustrated
by a comparison of the RSEs of the amides 4b and 4c
and the sulfonamides 4e and 4f. The correlation
between the electron-donating ability of amino, amido
and sulfonamido groups and radical stability is also
apparent from the effect of the halogens on the RSEs of
the radicals 4b–g. The inductively electron-withdrawing
fluorines of the trifluoroacetamide 4d reduce its RSE by
3–4 kJ mol−1, to 33.7 kJ mol−1, relative to the corre-
sponding values of 37.4 and 36.9 kJ mol−1 for the
acetamide 4c and the formamide 4b. Similarly the RSE
of the triflamide 4g is 2–4 kJ mol−1 lower than those of
the corresponding non-fluorinated analogues 4e and 4f.
The methoxycarbonyl group has a stabilizing effect in
the radical 14 of 21.6 kJ mol−1, but its effect on the
stability of the glycyl radical 6a is much greater. With
an RSE of 94.0 kJ mol−1, the amino and methoxycar-
bonyl groups in this species provide a synergistic or
captodative stabilization15,17–19 that is 27.7 kJ mol−1
more than the sum of their individual contributions to
the RSEs of the radicals 4a (44.7 kJ mol−1) and 14 (21.6
kJ mol−1). The effects of the amino, amido and sulfon-
amido groups on the stability of the glycyl radicals 6b–g
and 12a follow a similar trend to their effects on the
corresponding methyl radicals 4b–g and 10a, but the
magnitude is larger. Whereas acylation and sulfonation
reduces the stability of the aminomethyl radical 4a by
7–8 and 16–17 kJ mol−1, respectively, the effects on the
stability of the glycyl radical 6a, as represented by the
RSEs of the radicals 6b, 6c, 6e and 6f, are 12–15 and
20–22 kJ mol−1. As discussed above, the fluorines of the
substituted aminomethyl radicals 4d and 4g reduce the
RSEs by 2–4 kJ mol−1, but comparison of the RSEs of
the glycyl radicals 6b–g shows that the effect of the
halogens is much larger in these systems, being 7–10 kJ
mol−1 for the amides 6b–d and 6–8 kJ mol−1 for the
sulfonamides 6e–f. In all of the radicals 6a–g, the
p-electron-withdrawing methoxycarbonyl group mag-
Table 2. Relative rates of reaction of amino acid deriva-
tives with N-bromosuccinimide
RH+NBS
“
RBr+Succinimide
krel
5c (CH3CONHCH2CO2CH3)
5d (CF3CONHCH2CO2CH3)
5g (CF3SO2NHCH2CO2CH3)
5h (PhCONHCH2CO2CH3)
5i (p-FC6H4CONHCH2CO2CH3)
5j (C6F5CONHCH2CO2CH3)
15 (PhthalimidoCH2CO2CH3)
7c
7d
7g
7h
7i
1.2
0.05
<0.005
1.0*
0.86
0.25
7j
17a
<0.0005
* Assigned as unity.

2922
A. K. Croft et al. / Tetrahedron: Asymmetry 14 (2003) 2919–2926
nifies the extent to which the p-electron-donating ability
of the amino, amido and sulfonamido substituents is
reflected in the RSEs.
to 1.5, while the relative rates of formation of the
N-benzoyl-, p-fluorobenzoyl- and pentafluorobenzoyl-
glycyl radicals 5h–j decrease from 1.0 to 0.86 to 0.25.
The maleyl substituent decreases the RSE of the radical
10a by 16.2 kJ mol−1 relative to that of the aminomethyl
radical 4a. Again the methoxycarbonyl group increases
the extent of this effect, to 60.8 kJ mol−1, as shown by
the difference between the RSEs of the radicals 6a and
12a, of 94.0 and 33.2 kJ mol−1, respectively. However, in
this case the cause of the increase is quite different. The
radicals 4a–g, 6a–g and 10a are each able to adopt a
minimum energy conformation in which there is good
overlap of the p-orbitals of the substituent or sub-
stituents with the semi-occupied p-orbital at the radical
center. This is illustrated in Figure 1a for the imido-
methyl radical 10a which is planar with C2v symmetry.
By contrast, the minimum energy conformation of the
maleylglycyl radical 12a is non-planar (Figure 1b) as a
result of steric repulsion between the imide and ester
oxygens. This limits the extent of orbital overlap that
stabilizes the radical 12a and decreases the RSE much
more than might otherwise have been expected.
Of the systems examined in this study, apart from
imidation, triflation appears to be the most effective way
to reduce the RSEs and ease of formation of amino acid
a-carbon-centered radicals. Analogously to the behavior
seen with the corresponding phthalimides, with the
N-triflyl-substituted dipeptide 19a and the derivatives of
valine 20a and phenylalanine 21a, triflation was suffi-
cient to prevent reaction with NBS adjacent to the
nitrogen. Reaction instead occurred to give the glycyl-
bromoglycine derivative 19b (which was characterized as
the methoxide 19c), the b-bromovaline derivative 20b,
and a 1:1 mixture of diastereomers of the b-bro-
mophenylalanine derivative 21b. With sulfuryl chloride,
the valine derivative 20a gave a mixture of the b-
chlorovaline derivative 20d and the corresponding
diastereomeric g-chlorovaline derivatives, in the ratio
1.3:1:1. Reduction of the bromide 20b with tributyltin
hydride afforded 20a with the same optical rotation as
that of the sample used to prepare the bromide 20b,
establishing that the side-chain halogenation occurs
without loss of stereochemical integrity.
The extra methyl group of the radical 10b does not
significantly increase its RSE relative to that of the
imidomethyl radical 10a. Any stabilization provided by
the methyl group is offset by the strain associated with
interactions between the methyl group and imide oxy-
gens. This strain is evident from the distortion of the
bond angles about the imide nitrogen, as shown in
Figure 1c, and the C(O)ꢀNꢀCaꢀCb torsion angle of
6.64°. The a-methyl group of the alanyl radical 12b
increases its RSE by 12.8 kJ mol−1 relative to that of the
glycyl radical 12a, as is typical of the effect of an alkyl
substituent. The minimum energy conformation of the
alanyl radical 12b (Figure 1d) is distorted from planarity,
like that of the glycyl radical 12a, but there is no
indication that the side chain of 12b increases the extent
of this distortion.
4. Concluding remarks
There is a strong correlation between the RSEs of the
radicals 6c, 6d, 6g and 12a (Table 1), and the relative
rates of reaction of the glycine derivatives 5c, 5d, 5g and
15, to give the radicals 6c, 6d and 6g and the phthaloyl
analogue 16 of the maleyl radical 12a (Table 2). There
is also a direct correlation with the pKas of the acids
corresponding to the acyl and sulfonyl groups of 5c, 5d
and 5g. These pKas decrease principally as a result of
increased stabilization of the conjugate bases due to
greater electron delocalization. This parallels the extent
of delocalization of the nitrogen electrons by the acyl
and sulfonyl groups, which is inversely proportional to
the electron-donating ability of the nitrogen to the
radical center. Thus, the pKas of acetic acid, trifluoro-
acetic acid and triflic acid decrease in order from 4.7 to
0.3 to −5.5, while the RSEs and relative rates of forma-
tion of the N-acetyl-, trifluoroacetyl- and triflyl-glycyl
radicals 5c, 5d and 5g decrease from 82.2 to 72.8 to 66.0
kJ mol−1, and from 1.2 to 0.05 to <0.005, respectively.
A similar correlation is apparent for the radicals 5h–j,
where the pKas of benzoic acid, p-fluorobenzoic acid and
pentafluorobenzoic acid decrease in order from 4.2 to 3.8
These theoretical and experimental results suggest that
N-phthaloyl- and triflyl-amino acid derivatives are suffi-
ciently protected against hydrogen-atom abstraction
from the a-position, for their stereochemical integrity to
be preserved at that site while free radical reactions take
place elsewhere on these molecules. Factors that affect
the relative RSEs of the amino-, triflamido- and imido-
substituted radicals 4a, 4g, 10a, 6a, 6g and 12a, are likely
to contribute to these effects. Specifically, the lower
p-electron-donating ability of either an imide or a trifl-
amide, as well as the inability of imido and carboxy
substituents to assume a coplanar arrangement about a
carbon radical center, as required for optimal orbital
overlap, leads to reduced radical stability.
5. Theoretical and experimental procedures
5.1. Computational details
Standard ab initio molecular orbital theory and density

A. K. Croft et al. / Tetrahedron: Asymmetry 14 (2003) 2919–2926
2923
functional theory calculations were performed with
Gaussian 9420 and MOLPRO 96.21 For the larger sys-
tems, preliminary conformational work was carried out
at AM1 using the Spartan and MacSpartan Plus pro-
grams, in order to select the most appropriate confor-
mations. The calculated results correspond to isolated
molecules in the gas phase.
nitrogen, whilst being irradiated with a sunlamp. The
mixture was then cooled to room temperature and
filtered, and the filtrate was concentrated under reduced
pressure. This afforded the a-bromoglycine derivatives
7c, 7d, 7h, 7i and 7j, from 5c, 5d, 5h, 5i and 5j,
respectively. These bromides were unstable so they were
characterized as the corresponding methoxides 8c, 8d,
8h, 8i and 8j, through addition of methanol to the
corresponding crude reaction products.15 Reaction of
the sulfonamide 5g gave the succinimide 18, which was
isolated through chromatography. The phthalimide 15
was recovered unchanged under these reaction
conditions.
Optimized structures and harmonic vibrational frequen-
cies were determined at the B3-LYP/6-31G(d) level.
The structures were confirmed to be true local minima
by means of the vibrational frequency calculations that
showed all the frequencies to be real. Zero-point vibra-
tional energies were obtained by scaling the B3-LYP/6-
0.9806.22
Single-point
5.2.1.1. N-Acetyl-a-bromoglycine methyl ester 7c. Pale
brown oil, lH 2.10 (3H, s), 3.87 (3H, s), 6.48 (1H, d,
J=10.0) and 7.10 (1H, br d, J=10.0).
31G(d)
values
by
RMP2/6-31G(d) calculations on the B3-LYP/6-31G(d)
structures [RMP2/6-31G(d)//B3-LYP/6-31G(d)] were
used to obtain improved relative energies. This
approach has been found in previous work to produce
radical stabilization energies (RSEs) of useful
reliability.13
5.2.1.2. N-Acetyl-a-methoxyglycine methyl ester 8c.
(74% from 5c) Colorless needles from ethyl acetate/hex-
anes, mp 44–45°C (lit.27 44–46°C); lH 2.10 (3H, s), 3.46
(3H, s), 3.82 (3H, s), 5.53 (1H, d, J=9.0) and 6.80 (1H,
br d, J=9.0).
5.2. General
5.2.1.3. N-Trifluoroacetyl-a-bromoglycine methyl ester
7d. Pale brown oil, lH 3.94 (3H, s), 6.36 (1H, d,
J=10.0) and 7.10 (1H, br d, J=10.0).
Melting points were determined on a Kofler hot-stage
melting-point apparatus under a Reichert microscope
and are uncorrected. Microanalyses were performed by
the Research School of Chemistry Microanalytical Ser-
vice at the Australian National University, on a Carlo-
Erba 1106 autoanalyser. Nuclear Magnetic Resonance
(NMR) spectra were recorded in CDCl3 on either a
Varian Gemini 300 or a Varian VXR 500S spectrome-
ter. Electron impact (EI) mass spectra were obtained
using either a VG Autospec double-focusing trisector
mass spectrometer operating at 70 eV or a Vacuum
Generators ZAB2-SEQ mass spectrometer. Electro-
spray (ES) mass spectra were recorded on a VG Quatro
2 triple-quadrupole mass spectrometer. Optical rota-
5.2.1.4. N-Trifluoroacetyl-a-methoxyglycine methyl
ester 8d. (80% from 5d) Colorless solid from ethyl
acetate/hexanes, mp 96–99°C (Found: C, 33.58; H,
3.88; N, 6.63. C6H8F3NO4 requires C, 33.48; H, 3.75;
N, 6.51%); lH 3.53 (3H, s), 3.87 (3H, s), 5.53 (1H, d,
J=9.0) and 7.40 (1H, br d, J=9.0).
5.2.1.5. N-Benzoyl-a-bromoglycine methyl ester 7h.
Pale brown oil, lH 3.91 (3H, s), 6.68 (1H, d, J=10.0),
7.15 (1H, br d, J=10.0), 7.42–7.60 (3H, m) and 7.81–
7.85 (2H, m).
tions were measured using
polarimeter.
a Perkin Elmer 241
5.2.1.6. N-Benzoyl-a-methoxyglycine methyl ester 8h.
(85% from 5h) Colorless solid from ethyl acetate/hex-
anes, mp 72–73°C (lit.28 86–87°C) (Found: C, 59.23; H,
5.98; N, 6.51. C11H13NO4 requires C, 59.19; H, 5.87; N,
6.27%); lH 3.56 (3H, s), 3.87 (3H, s), 5.78 (1H, d,
J=8.5), 7.12 (1H, br d, J=8.5), 7.44–7.59 (3H, m) and
Preparative chromatography was performed using
either dry flash column chromatography23 or radial
chromatography on a Harrison research model 7924T
chromatotron, with Merck Kieselgel 60 PF254 contain-
ing gypsum, eluting with a gradient of ethyl acetate/
hexanes. An Osram Ultra-Vitalux (240 V, 300 W, E27)
sunlamp was used as the light source, to initiate radical
reactions, at a distance of between 5 and 10 cm from
the reaction vessel. Organic extracts were dried using
anhydrous magnesium sulfate.
+
’
7.83–7.86 (2H, m); m/z (EI) 223 (M , 0.6%), 164 (45),
105 (100) and 77 (40).
5.2.1.7. N-(4-Fluorobenzoyl)-a-bromoglycine methyl
ester 7i. Pale brown oil, lH 3.93 (3H, s), 6.66 (1H, d,
J=10.0), 6.95 (1H, br d, J=10.0), 7.14 (2H, d, J=9.0)
and 7.19 (2H, d, J=9.0).
Glycine, glycylglycine, (S)-valine and phenylalanine
were purchased from Aldrich Chemical Co. The glycine
derivatives 5c, 5d, 5g, 5h, 5i, 5j and 15 were prepared
from glycine using standard methods and had physical
and spectroscopic data consistent with those reported
previously.7,24–26
5.2.1.8. N-(4-Fluorobenzoyl)-a-methoxyglycine methyl
ester 8i. (87% from 5i) Colorless solid from ethyl ace-
tate/hexanes, mp 98–100°C (Found: C, 54.85; H, 5.09;
N, 6.01. C11H12FNO4 requires C, 54.75; H, 5.02; N,
5.81%); lH 3.54 (3H, s), 3.86 (3H, s), 5.76 (1H, d,
J=9.0), 6.85 (1H, br d, J=9.0) and 7.10–7.20 (4H, m).
5.2.1. Treatment of the glycine derivatives 5c, 5d, 5g, 5h,
5i, 5j and 15 with NBS. A mixture of the glycine
derivative (0.42 mmol) and NBS (80 mg, 0.45 mmol) in
CCl4 (5 ml) was heated at reflux for 0.5–3 h, under
5.2.1.9.
N-(Pentafluorobenzoyl)-a-bromoglycine
methyl ester 7j. Pale brown oil, lH 3.82 (3H, s), 6.57
(1H, d, J=5.0) and 6.73 (1H, br d, J=5.0).

2924
A. K. Croft et al. / Tetrahedron: Asymmetry 14 (2003) 2919–2926
5.2.1.10.
N-(Pentafluorobenzoyl)-a-methoxyglycine
5.2.4. (S)-N-Trifluoromethanesulfonylvaline methyl ester
20a. (S)-Valine was treated with methanol which had
been pretreated with thionyl chloride, and the crude
product ester was treated with trifluoromethanesulfonic
anhydride, as described above for the synthesis of the
glycylglycine derivative 19a, to give the title compound
20a as colorless crystals (54%) mp 43–45°C (lit.29 44–
46°C); [h]2D5=−39.4 (c 0.50, MeOH); lH 0.92 (3H, d,
J=7.0), 1.04 (3H, d, J=7.0), 2.21 (1H, m), 3.81 (3H, s),
4.07 (1H, dd, J=4.5, 9.5) and 5.48 (1H, br d, J=9.5);
lC 17.5, 19.2, 32.0, 53.5, 62.9, 120.0 (q, J=320) and
171.7.
methyl ester 8j. (77% from 5j) Colorless solid from ethyl
acetate/hexanes, mp 104–105°C (Found: C, 42.28; H,
2.63; N, 4.57. C11H8F5NO4 requires C, 42.16; H, 2.57;
N, 4.47%); lH 3.55 (3H, s), 3.86 (3H, s), 5.71 (1H, d,
J=9.0) and 7.40 (1H, br d, J=9.0).
5.2.1.11. N-Trifluoromethanesulfonyl-a-(N-succinimy-
dyl)glycine methyl ester 18. (98%) (Found m/z (EI)
+
’
259.0000. C6H6F3N2O4S (M −CO2Me) requires m/z
258.9999); lH 2.83 (4H, s), 3.85 (3H, s), 6.08 (1H, d,
J=9.0) and 7.75 (1H, br d, J=9.0); lC 28.0, 54.5, 58.3,
+
’
119.8 (q, J=319), 164.8 and 175.1; m/z (EI) 259 (M −
5.2.5. N-Trifluoromethanesulfonylphenylalanine methyl
ester 21a. Phenylalanine (0.98 g, 5.9 mmol) was treated
with methanol which had been pretreated with thionyl
chloride, and the crude product ester was treated with
trifluoromethanesulfonic anhydride, as described above
for the synthesis of the glycylglycine derivative 19a, to
give the title compound 21a as a colourless powder (350
mg, 19%), mp 45–46°C from hexane (Found: C, 42.73;
H, 3.68; N, 4.45. C11H12F3NO4S requires C, 42.45; H,
3.89; N, 4.50%.); lH 3.12 (1H, dd, J=6.0, 13.5), 3.18
(1H, dd, J=6.0, 13.5), 3.77 (3H, s), 4.50 (1H, t, J=6.0);
5.77 (1H, br), 7.12–7.15 (2H, m) and 7.29–7.34 (3H, m);
lC 40.0, 53.5, 58.3, 115.6 (q, J=320), 128.3, 129.4,
CO2Me, 100%), 221 (12), 162 (12) and 69 (24); m/z (ES)
(+ve) 319 (M+H+); m/z (ES) (−ve) 317 (M−H+).
5.2.2. Treatment of mixtures of the amino acid deriva-
tives 5c, 5d, 5g, 5h, 5i, 5j and 15 with NBS. Equimolar
mixtures of 5c and 5d, 5c and 5h, 5c and 5i, 5c and 5j,
5d and 5j, 5d and 5g, and 5g and 15, were treated with
NBS (1 mol equiv.) in CCl4, in the presence of N-tert-
butylbenzamide (0.1 mol equiv.) as an internal stan-
dard, under the conditions used for the reactions of 5c,
1
5d, 5g, 5h, 5i and 5j described above. H NMR spectro-
scopic analysis of the product mixtures was used to
determine the relative rates of reaction, as reported
previously,15 and the results are shown in Table 2.
Within the limits of detection, with 5d and 5g, the
trifluoroacetamide 5d was quantitatively converted to
the bromide 7d, while the sulfonamide 5g was inert.
With 5g and 15, the sulfonamide 5g was quantitatively
converted to the succinimide 18, while the phthalimide
15 was inert. On this basis the relative rates of reaction
of 5d, 5g and 15 were determined to be 1:<0.1:<0.01.
+
’
129.9, 134.4 and 170.9; m/z (EI) 311 (M , 6%), 252
(61), 182 (19), 163 (38), 162 (100), 131 (41), 119 (61),
118 (73), 103 (22), 92 (52), 91 (55), 77 (15), 69 (56) and
65 (68).
5.2.6.
N-Trifluoromethanesulfonylglycyl-a-methoxy-
glycine methyl ester 19c. A mixture of N-trifluoro-
methanesulfonylglycylglycine methyl ester 19a (50 mg,
0.18 mmol) and NBS (35 mg, 0.20 mmol) in CCl4 and
CH2Cl2 (1:3, v/v, 5 ml) was heated at reflux for 4 h
under nitrogen in a quartz tube whilst being irradiated
with a sunlamp. The mixture was then allowed to cool
to room temperature, before being filtered, to afford a
solution of the bromide 19b; lH 3.89 (3H, s), 4.09 (2H,
d, J=5.5), 6.42 (1H, d, J=10.0), 6.75 (1H, br t, J=5.5)
and 7.79 (1H, d, J=10.0). The yield of the bromide 19b
was determined to be 92% through the addition of
N-tert-butylbenzamide as an external standard.
5.2.3. N-Trifluoromethanesulfonylglycylglycine methyl
ester 19a. Glycylglycine (0.80 g, 6.1 mmol) was added
to methanol (20 ml) which had been pretreated with
thionyl chloride (0.85 g, 7.1 mmol) and the mixture was
let stand overnight, then it was concentrated under
reduced pressure, to yield a white powder. The powder
was dissolved in methanol and the solution was concen-
trated under reduced pressure. The residual solid was
suspended in dichloromethane (20 ml) and triethyl-
amine (1.26 g, 12.5 mmol) was added with stirring. The
suspension was cooled to −78°C and trifluoromethane-
sulfonic anhydride (1.67 g, 5.9 mmol) in
dichloromethane (10 ml) was added dropwise, then the
mixture was allowed to warm to room temperature and
let stand overnight. The resultant solution was washed
with dilute hydrochloric acid (2×10 ml) and saturated
aqueous sodium chloride (2×10 ml), dried over magne-
sium sulfate and concentrated under reduced pressure,
to give the title compound 19a as a pale yellow powder
(0.29 g, 18%) mp 121°C (Found: C, 25.96; H, 3.22; N,
10.08. C6H9F3N2O5S requires C, 25.90; H, 3.26; N,
10.07%); lH 3.80 (3H, s), 4.02 (2H, d, J=5.0), 4.11 (2H,
d, J=5.0), 6.17 (1H, br t, J=5.0) and 6.32 (1H, br t,
J=5.0); lC 41.0, 45.7, 52.1, 119.6 (q, J=320), 167.8 and
170.0; m/z (EI) 279 (M+H+, 11%), 219 (100), 209 (29),
191 (11), 162 (60), 133 (17), 116 (72) and 88 (99).
Methanol (1 ml) was added to the solution of the
bromide 19b and the mixture was stirred for 1 h at
room temperature before it was concentrated under
reduced pressure. Chromatography of the residue
afforded the title compound 19c as colorless needles (20
mg, 36%), mp 94–95°C from ethyl acetate/hexanes
+
’
(Found: m/z (EI) 277.0108. C6H8F3N2O5S (M −OMe)
requires m/z 277.0106); lH 3.49 (3H, s), 3.85 (3H, s),
4.05 (2H, d, J=5.0), 5.54 (1H, d, J=9.0), 6.04 (1H, br
t, J=5.0) and 6.82 (1H, br d, J=9.0); lC 46.2, 53.9,
57.8, 120.0 (q, J=320), 167.5 and 168.5; m/z (EI) 293
+
’
(M −Me, 12%), 279 (23), 278 (50), 277 (64), 250 (11),
249 (83), 162 (22), 104 (14) and 103 (54); m/z (ES) (+ve)
309 (M+H+). The regioselectivity of methoxylation was
determined by mass spectrometry (Found: m/z (EI)
+
’
103.0396. C4H7O3 (CH(OMe)CO2Me ) requires m/z
103.0395).

A. K. Croft et al. / Tetrahedron: Asymmetry 14 (2003) 2919–2926
2925
5.2.7.
(S)-N-Trifluoromethanesulfonyl-3-bromovaline
under reduced pressure. Chromatography of the residue
afforded the valine derivative 20a as colorless crystals
(17 mg, 44%), mp 42–44°C, [h]2D5=−40.1 (c 0.09,
MeOH). These data are consistent with those obtained
above, using a sample prepared directly from (S)-
valine.
methyl ester 20b. A mixture of the valine derivative 20a
(0.20 g, 0.76 mmol) and NBS (135 mg, 0.76 mmol) in
carbon tetrachloride (5 ml) was heated at reflux for 5 h
whilst being irradiated with a sunlamp. The mixture
was then allowed to cool to room temperature, before it
was filtered. The filtrate was concentrated under
reduced pressure and the residue was subjected to chro-
matography, to give the title compound 20b as colorless
crystals (180 mg, 69%), mp 47.5–48°C from hexanes
(Found: C, 24.62; H, 3.08; N, 4.20. C7H11BrF3NO4S
requires C, 24.57; H, 3.24; N, 4.09%.); [h]2D5=−65.6 (c
0.75, MeOH); lH 1.91 (3H, s), 1.93 (3H, s), 3.87 (3H, s),
4.05 (1H, d, J=10.0) and 5.94 (1H, d, J=10.0); lC
32.3, 32.9, 53.7, 62.7, 66.7, 119.9 (q, J=320) and 168.5;
5.2.10. N-Trifluoromethanesulfonyl-3-bromophenylala-
nine methyl ester 21b. To a solution of N-trifluoro-
methanesulfonylphenylalanine methyl ester 21a (100
mg, 0.32 mmol) in carbon tetrachloride (10 ml) was
added NBS (64 mg, 0.35 mmol). The mixture was
heated at reflux for 6 h under nitrogen whilst being
irradiated with a sunlamp, then it was allowed to cool
before it was filtered. The filtrate was concentrated
under reduced pressure and the residue was chro-
matographed to give a 1:1 mixture of diastereomers of
the title compound 21b as a colorless solid (85 mg,
68%), mp 113–126°C (Found: C, 33.98; H, 2.60; N,
3.38. C11H11BrF3NO4S requires C, 33.86; H, 2.84; N,
3.59%.); lH 3.79 (3H, s), 3.86 (3H, s), 4.51 (1H, dd,
J=3.5, 10.0), 4.71 (1H, dd, J=5.5, 10.0), 5.31 (1H, d,
J=5.5), 5.52 (1H, d, J=3.5), 5.83 (1H, br d, J=10.0),
5.97 (1H, br d, J=10.0) and 7.37–7.47 (10H, m); lC
51.7, 54.0, 54.4, 63.7, 64.2, 119.7 (q, J=320), 119.8 (q,
m/z (EI) 342 (M , Br, 7%), 340 (M , 79Br, 7), 284
(49), 282 (62), 262 (37), 221 (67), 220 (33), 202 (100),
149 (23), 123 (37), 121 (40), 88 (52) and 69 (40).
+
81
+
’
’
5.2.8. Treatment of (S)-N-trifluoromethanesulfonylvaline
methyl ester 20a with sulfuryl chloride. To a solution of
(S)-N-trifluoromethanesulfonylvaline methyl ester 20a
(100 mg, 0.38 mmol) in carbon tetrachloride (5 ml) was
added sulfuryl chloride (51 mg, 0.38 mmol) and benzoyl
peroxide (ca. 5 mg). The mixture was heated at reflux
for 5 h under nitrogen whilst being irradiated with a
sunlamp, then it was cooled and concentrated under
reduced pressure. The residue was chromatographed to
give unreacted starting material 20a (24 mg, 24%),
(S)-N-trifluoromethanesulfonyl-3-chlorovaline methyl
ester 20d (31 mg, 27%) as colorless crystals, mp 64–
65°C (Found: C, 28.50; H, 3.83; N, 4.89.
C7H11ClF3NO4S requires C, 28.24; H, 3.72; N, 4.71%.);
[h]2D5=−27.8 (c 0.24, MeOH); lH 1.73 (3H, s), 1.74 (3H,
s), 3.85 (3H, s), 4.16 (1H, br) and 5.95 (1H, br); lC 30.7,
31.3, 53.7, 66.1, 68.6, 120.0 (q J=320) and 168.4; m/z
(EI) 240 (11%), 238 (49), 202 (100), 149 (47), 79 (52)
and 77 (89); and a 1:1 mixture of diastereomers of
(2S,3S)- and (2S,3R)-N-trifluoromethanesulfonyl-4-
chlorovaline methyl ester (46 mg, 41%) as a colorless
solid, mp 52–71°C (Found: C, 28.20; H, 3.74; N, 4.56.
C7H11ClF3NO4S requires C, 28.24; H, 3.72; N, 4.71%);
m/z (EI) 300 (M+H+, 0.3%), 298 (M+H+, 0.8), 274 (3),
272 (5), 240 (49), 238 (100), 220 (30), 202 (49) and 162
(51); one diastereomer had lH 1.12 (3H, d, J=7.0), 2.49
(1H, m), 3.55 (2H, m), 3.84 (3H, s), 4.33 (1H, d, J=4.5)
and 5.82 (1H, br); lC 14.6, 39.3, 45.7, 53.3, 59.3, 119.3
(q, J=320) and 170.1; the other diastereomer had lH
1.00 (3H, d, J=7.0), 2.46 (1H, m), 3.46 (1H, dd, J=6.0,
11.5), 3.56 (1H, dd, J=8.0, 11.5), 3.85 (3H, s), 4.52
(1H, br d, J=4.5) and 5.60 (1H, br); lC 12.6, 39.4, 45.6,
J=320), 128.8, 128.9, 129.4, 129.6, 130.1, 130.2, 135.7,
+
’
136.3, 168.6 and 168.8; m/z (EI) 391 (M , 2%), 389
+
’
(M , 2), 332 (4), 330 (4), 310 (10), 251 (10), 250 (13),
242 (8), 240 (8), 192 (6), 171 (97), 169 (100), 118 (62),
117 (42), 105 (36), 91 (71) and 77 (22). Repeated
chromatography afforded a single diastereomer (23 mg,
18%) mp 123–123.5°C; lH 3.86 (3H, s), 4.51 (1H, dd,
J=3.5, 10.0), 5.52 (1H, d, J=3.5), 5.83 (1H, br d,
J=10.0) and 7.37–7.47 (5H, m).
References
1. Williams, R. M. Synthesis of Optically Active h-Amino
Acids; Baldwin, J. E.; Magnus, P. D., Eds.; Organic
Chemistry Series, Vol. 7; Pergamon Press: Oxford, 1989.
2. Seebach, D.; Beck, A. K.; Studer, A. In Modern Synthetic
Methods; Ernst, B.; Leumann, C., Eds.; VCH: Weinheim,
1995; Vol. 7, pp. 1–178.
3. Easton, C. J. Chem. Rev. 1997, 97, 53.
4. Easton, C. J. In Radicals in Organic Synthesis; Renaud,
P.; Sibi, M. P., Eds.; Wiley-VCH: Weinheim, 2001; Vol.
2, pp. 505–522.
5. Wessig, P. In Radicals in Organic Synthesis; Renaud, P.;
Sibi, M. P., Eds.; Wiley-VCH: Weinheim, 2001; Vol. 2,
pp. 523–537.
1
53.5, 58.2, 119.5 (q, J=320) and 170.5. A H NMR
6. Easton, C. J.; Tan, E. W.; Hay, M. P. J. Chem. Soc.
spectrum of the crude product mixture showed that the
b-chlorovaline derivative and the diasteromers of the
corresponding g-chlorovaline derivative were present in
the ratio 1.3:1:1.
Chem. Commun. 1989, 385.
7. Easton, C. J.; Hutton, C. A.; Rositano, G.; Tan, E. W. J.
Org. Chem. 1991, 56, 5614.
8. Easton, C. J.; Hutton, C. A.; Roselt, P. D.; Tiekink, E.
R. T. Tetrahedron 1994, 50, 7327.
5.2.9. Treatment of the bromovaline derivative 20b with
tributyltin hydride. A mixture of the bromide 20b (50
mg, 0.15 mmol) and tributyltin hydride (0.1 ml, 3.7
mmol) in benzene (10 ml) was heated at reflux under
nitrogen for 6 h, then it was cooled and concentrated
9. Easton, C. J.; Hutton, C. A.; Merrett, M. C.; Tiekink, E.
R. T. Tetrahedron 1996, 52, 7025.
10. Easton, C. J.; Hutton, C. A. Synlett 1998, 457.
11. Easton, C. J.; Fryer, N. L.; Ivory, A. J.; Tiekink, E. R. T.
J. Chem. Soc., Perkin Trans. 1 1998, 3725.

2926
A. K. Croft et al. / Tetrahedron: Asymmetry 14 (2003) 2919–2926
12. Henry, D. J.; Parkinson, C. J.; Mayer, P. M.; Radom, L.
J. Phys. Chem. A 2001, 105, 6750.
13. Parkinson, C. J.; Mayer, P. M.; Radom, L. J. Chem.
Soc., Perkin Trans. 2 1999, 2305.
14. Armstrong, D. A.; Yu, D.; Rauk, A. Can. J. Chem. 1996,
74, 1192.
15. Croft, A. K.; Easton, C. J.; Radom, L. J. Am. Chem. Soc.
2003, 125, 4119.
16. Easton, C. J.; Scharfbillig, I. M.; Tan, E. W. Tetrahedron
94, Revision E.2; Gaussian, Inc.: Pittsburgh PA, 1995.
21. MOLPRO 96 is a package of ab initio programs written
by H.-J. Werner and P. J. Knowles, with contributions
from J. Almlo¨f, R. D. Amos, A. Berning, M. J. O.
Deegan, F. Eckert, S. T. Elbert, C. Hampel, R. Lindh W.
Meyer, A. Nicklass, K. Peterson, R. Pitzer, A. J. Stone,
P. R. Taylor, M. E. Mura, P. Pulay, M. Schuetz, H. Stoll,
T. Thorsteinsson, and D. L. Cooper, © University of
Birmingham: Birmingham, UK, 1996.
22. Scott, A. P.; Radom, L. J. Phys. Chem. 1996, 100, 16502.
23. Still, W. C.; Kahn, M.; Mitra, A. J. Org. Chem. 1978, 43,
2923.
Lett. 1988, 29, 1565.
17. Schulze, R.; Beckhaus, H.-D.; Ru¨chardt, C. Chem. Ber.
1993, 126, 1031.
24. Huang, H. T.; Niemann, C. J. Am. Chem. Soc. 1952, 74,
4634.
18. Welle, F. M.; Beckhaus, H.-D.; Ru¨chardt, C. J. Org.
Chem. 1997, 62, 552.
25. Applewhite, T. H.; Martin, R. B.; Niemann, C. J. Am.
19. Yu, D.; Rauk, A.; Armstrong, D. A. J. Am. Chem. Soc.
1995, 117, 1789.
Chem. Soc. 1958, 80, 1457.
26. Burgess, V. A.; Easton, C. J.; Hay, M. P. J. Am. Chem.
Soc. 1989, 111, 1047.
27. Legall, P.; Sawhney, K. N.; Conley, J. D.; Kohn, H. Int.
20. Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Gill, P. M.
W.; Johnson, B. G.; Robb, M. A.; Cheeseman, J. R.;
Keith, T.; Petersson, G. A.; Montgomery, J. A.;
Raghavachari, K.; Al-Laham, M. A.; Zakrzewski, V. G.;
Ortiz, J. V.; Foresman, J. B.; Cioslowski, J.; Stefanov, B.
B.; Nanayakkara, A.; Challacombe, M.; Peng, C. Y.;
Ayala, P. Y.; Chen, W.; Wong, M. W.; Andres, J. L.;
Replogle, E. S.; Gomperts, R.; Martin, R. L.; Fox, D. J.;
Binkley, J. S.; Defrees, D. J.; Baker, J.; Stewart, J. P.;
Head-Gordon, M.; Gonzalez, C.; Pople, J. A. Gaussian
J. Pept. Protein Res. 1988, 32, 279.
28. Chemiakine, M. M.; Tchaman, E. S.; Denisova, L. I.;
Ravdel, G. A.; Rodionow, W. J. Bull. Soc. Chim. Fr.
1959, 530.
29. (a) Sartor, D.; Saffrich, J.; Helmchen, G. Synlett 1990,
197; (b) Helmchen, G. University of Heidelberg, personal
communication.