Aromaticity in azlactone anions and its significance for the Erlenmeyer synthesis

Article abstract of DOI:10.1016/j.tetlet.2006.06.006

Azlactone anions-the key intermediates in the classical Erlenmeyer synthesis of amino acids-apparently possess aromatic stabilization, as indicated by the relative rate of base catalyzed deuterium exchange in the following analogs: 1-methyl-2-phenyl-5(4H)

Full text of DOI:10.1016/j.tetlet.2006.06.006

Tetrahedron Letters 47 (2006) 5763–5766
Aromaticity in azlactone anions and its significance
for the Erlenmeyer synthesis
Sosale Chandrasekhar* and Phaneendrasai Karri
Department of Organic Chemistry, Indian Institute of Science, Bangalore 560 012, India
Received 7 April 2006; revised 22 May 2006; accepted 1 June 2006
Available online 21 June 2006
Abstract—Azlactone anions—the key intermediates in the classical Erlenmeyer synthesis of amino acids—apparently possess
aromatic stabilization, as indicated by the relative rate of base catalyzed deuterium exchange in the following analogs: 1-methyl-
2-phenyl-5(4H)-imidazolone > 2-phenyl-5(4H)-oxazolone (azlactone) > 3,3-dimethyl-2-phenyl-4(3H)-pyrrolone. This is paralleled
by the relative rate of condensation of these compounds with hexadeuteroacetone. Reported pKa data also suggest that the azlactone
products of the Erlenmeyer synthesis are analogs of the fulvenes.
Ó 2006 Elsevier Ltd. All rights reserved.
The classical Erlenmeyer azlactone synthesis (1893) is an
important method for preparing aromatic a-amino
acids,1 and azlactones continue to find diverse applica-
tions in synthesis as revealed by several recent
reports.2–4 These include asymmetric syntheses of
various natural and unnatural amino acids2,3 and pep-
tide synthesis.4 The general revival of azlactone chemis-
try is exemplified by a report on the asymmetric
alkoxyallylation of azlactones with a chiral Pd(II)
catalyst.5
of the resulting anion I to an aromatic aldehyde; the
initially formed arylidene derivative 2 is subsequently
transformed into the amino acid 3 (Scheme 1). It is inter-
esting to note that deprotonation of 1 is effected by the
rather weakly basic acetate ion (pKa ꢀ 5):6 it is possible
that the acidity of 1 is derived from the aromatic stabil-
ization of its anion I, which possesses six electrons in
cyclic conjugation (with the lone pair of the ring oxygen
atom included).7 We have addressed this possibility as
follows.
The key steps of the Erlenmeyer synthesis are the base-
We have studied the relative rate of base-cata-
catalyzed deprotonation of azlactone 1, and the addition
lyzed deprotonation in azlactone 1 and its analogs
OH
O-
O
O
O
H
H
Ar
Ar
iii
i
ii
H
H
O
O
O
O
N
N
N
N
Ph
1
Ph
Ph
Ph
2
I
iv, v
(i) AcO-; (ii) ArCHO/H+; (iii) (-H2O);
(iv) OH-; (v) reduction
-
Ar
CO2
+
3
NH3
Scheme 1. The key steps in the Erlenmeyer azlactone synthesis of amino acids.
Keywords: Amino acid; Aromaticity; Azlactone; Deuterium exchange; Erlenmeyer synthesis; Fulvene; Hexadeuteroacetone.
*
Corresponding author. Tel.: +91 80 2293 2689/2402; fax: +91 80 23600 529; e-mail: sosale@orgchem.iisc.ernet.in
0040-4039/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2006.06.006

5764
S. Chandrasekhar, P. Karri / Tetrahedron Letters 47 (2006) 5763–5766
O-
O
H
H
(i)
H
Z
Z
1,4,5/I-III/6-8
1, I, 6
Z
N
N
O
Ph
I, II, III
Ph
1, 4, 5
4, II, 7
NMe
C(Me)2
(ii)
5, III, 8
(iii)
(1, 4, 5)-d1
(ii)
O
D
(i) Pyridine/CDCl3/25 oC/(-C5H5NH+);
(ii) D2O; (iii) (CD3)2C=O
CD3
N
O
D
D3C
Z
N
Z
Ph
Ph
6-8
(1, 4, 5)-d2
Scheme 2. Reactions of azlactone 1 and its analogs 4 and 5: (i) and (ii) base catalyzed H-D exchange to form (1,4,5)-d2, via the corresponding
monodeutero derivatives (1,4,5)-d1; (iii) condensation with acetone-d6 to form the isopropylidenes 6–8.
imidazolone 4 and pyrrolidone 5 (Scheme 2), in two
processes, using 1H NMR spectroscopy: deuterium
exchange and condensation with hexadeuteroacetone.
As both these processes would be mediated by the an-
ions of 1, 4 and 5, their relative rates would reflect the
relative stabilities of these anions. (This follows from
Hammond’s postulate, as discussed under ‘Mechanistic
aspects’ below). Compounds 1, 4 and 5 are expected to
exist in the ‘oxo’ forms shown, rather than as the enol
tautomers;8 compounds 19 and 410 were prepared as
previously reported and pyrrolidone 5 via an intramo-
lecular aza-Wittig reaction (Scheme 3).10–12
were recorded at five time periods as indicated. The
results obtained are shown in Table 1 (first row).
These results clearly show that the rate of the base-cat-
alyzed deuterium exchange is in the order imidazolone
4 > azlactone 1 > pyrrolidone 5. (Thus, 90 min after
the start of the reaction, 4 had all but disappeared and
1 had nearly half disappeared; at 24 h both 1 and 4
had totally disappeared, whereas 5 had barely reacted.)
This reactivity order is explicable on the basis of the sta-
bilization of the derived anions I and II (Scheme 2) by
aromaticity, which is absent in III. The higher reactivity
of 4 versus 1 would arise from the greater aromaticity of
the imidazole nucleus (cf. II) relative to the oxazole (cf.
I).13,14 Note also that the least deshielded a proton ex-
changes the fastest, ruling out a simple inductive effect.
Deuterium exchange. Generally, the rate of deuterium
exchange would follow (inversely) the order of pKas:
ketone < ester (lactone) < amide (lactam).6 On this ba-
sis, the rate of deuterium exchange would be pyrrolidone
5 > azlactone 1 > imidazolone 4; however, the order
would be 4 > 1 > 5 were aromatic stabilization to be
included (vide infra). The deuterium exchange in 1, 4
Condensation with acetone-d6. The above trend is paral-
leled by the relative rate of condensation of 1, 4 and 5
with acetone-d6 (hexadeuteroacetone: ‘HDA’), to form
the corresponding isopropylidene derivatives 6–8
(Scheme 2). Although far slower than deuterium ex-
change, this reaction occurred without added base upon
addition of HDA to a solution of 1, 4 and 5 in CDCl3,
the reaction being followed by 1H NMR analysis at
400 MHz.
1
and 5 was studied by H NMR spectroscopy in CDCl3,
by monitoring the disappearance of the resonance of
the proton a to the carbonyl group upon the addition
of D2O and a catalytic amount of pyridine. These
resonances occurred at d 4.42 (1), 4.36 (5) and 4.29
(4), as sharp singlets that gradually broadened and
diminished as the deuterium exchange progressed. The
three analogs could thus be studied together as
the chemical shifts of the protons undergoing exchange
were well separated (by > d 0.06). The NMR spectra
This reaction is analogous to the penultimate step in the
Erlenmeyer synthesis (cf. Scheme 1), in which the azlac-
tone condenses with an aldehyde. In the absence of base,
O
H
H
O
O
O
O
O
O
i
ii
iii
Br
N3
Ph
Ph
Ph
N
Ph
79%
5 (80%)
(i) Br2/AcOH; (ii) NaN3/DMSO; (iii) Ph3P/PhH/-(N2 + Ph3PO)
Scheme 3. Synthesis of pyrrolidone 5 via an aza-Wittig reaction.

S. Chandrasekhar, P. Karri / Tetrahedron Letters 47 (2006) 5763–5766
5765
Table 1. Relative reactivity of azlactone 1, imidazolone 4 and pyrrolidone 5 with deuterium oxide (D2O) and with hexadeuteroacetone (HDA)a,b
Reactants
Ratio 1:4:5 at time
T3
1.00:1.00:1.00 (0 min) 0.93:0.72:1.00 (5 min) 0.74:0.35:1.00 (60 min) 0.54:0.10:1.00 (90 min) 0.00:0.00:0.95 (24 h)
1.07:0.77:1.00 (18 h) 0.98:0.60:1.00 (25 h) 0.96:0.45:1.00 (36 h) 0.73:0.16:1.00 (48 h)
T1
T2
T4
T5
1 + 4 + 5 + D2O
1 + 4 + 5 + HDA 1.20:1.05:1.00 (0 h)
a As measured by the changes in the reactant ratios by 1H NMR spectroscopy [300 MHz (D2O reaction)/400 MHz (HDA reaction)] at 25 °C: the
intensities of the resonances of the protons a to the carbonyl group were recorded at the indicated time intervals, and are shown above relative to 5.
The exchange reaction was performed in 0.1 M CDCl3 solution (0.5 ml) in an NMR tube, with pyridine (2 ll) and D2O (5 ll). The condensation
with HDA was similarly performed in 0.12 M CDCl3 solution (0.4 ml) with added HDA (0.1 ml).
The initial rates approximation17 can be employed to arrive at a rough measure of the relative reactivity: in the D2O reaction the ratio of the percent
reaction at 5 min (T2) is 1:4 = 7:28 (relative rate = 1:4); in the HDA reaction at 18 h the percent ratio is 1:4 = 13:28, which upon correction for the
ratio of the starting concentrations (1.20:1.05) becomes 10.8:26.7 (relative rate = 2.47).
b Preparative methods. 2-Phenyl-5(4H)-oxazolone (azlactone 19) and 1-methyl-2-phenyl-5(4H)-imidazolone (imidazolone 410) were prepared as
previously reported, and characterized from melting point and spectral data (IR, 300 MHz 1H NMR). 3,3-Dimethyl-2-phenyl-4(3H)-pyrrolone
(pyrrolidone 5) was prepared by the intramolecular aza-Wittig reaction employed by Japanese workers for the synthesis of 4, but starting from 1,1-
dimethylbenzoylacetone (Scheme 3):10–12 this was first converted, with Br2/AcOH, to the 3-bromo derivative (not purified), which upon treatment
with NaN3/DMSO at 25 °C furnished the corresponding azide in 79% yield; this was treated with Ph3P/benzene at 25 °C to effect the aza-Wittig
reaction to form 5 in 80% yield (Scheme 3). Pyrrolidone 5 was thoroughly characterized by IR, 1H and 13C NMR, LRMS and HRMS.
The condensation products 6 and 7 were isolated via evaporation of the solvent from the reaction mixture, and characterized by IR, 1H and 13C
NMR, LRMS and HRMS (8 was formed in minuscule amounts and was not isolated); 6 and 7 were identified by their IR spectra, which showed
carbonyl bands at 1781 cmÀ1 and 1699 cmÀ1, respectively, being much lower than the carbonyl bands in the corresponding parent compounds 1
(1814 cmÀ1) and 4 (1741 cmÀ1).
it possibly occurs via the corresponding enols (not
shown, but cf. their conjugate bases I–III, Scheme 2),
the relative stabilities of these also being governed by
the above aromaticity criteria. The relative reactivity
was again imidazolone 4 > azlactone 1 > pyrrolidone 5,
as measured by the rate of disappearance of the reso-
nances of the methylene protons a to the carbonyl group
(Table 1, second row).
sponding relative reactivity in the HDA reaction is
4:1 = 2.5:1.)
The above results also explain the easy racemization of
4-substituted azlactones (which are unwanted by-prod-
ucts and intermediates in certain peptide syntheses).18
Interestingly also, the initial Erlenmeyer products 2
may well be stabilized by conjugation that enhances
the aromaticity of the oxazole moiety (cf. Scheme 4), a
charge transfer process rather reminiscent of fulvenes.19
Intriguingly, the pKa of ꢀ12 reported20 for compounds 2
indicates that they are more basic than trialkylamines
(pKa ꢀ 11),6 and suggests that canonical form 2b
(Scheme 4) is an important contributor. Therefore,
although it is generally believed that oxazoles are only
marginally aromatic,13,14 the present results suggest that
The isopropylidene derivatives 6–8 were clearly distin-
guished from the corresponding starting materials 1, 4
and 5 by having a lower wavenumber (by ꢀ35 cmÀ1
)
carbonyl stretch. The N-Me signal in the imidazolone
derivative 7 showed a discernible downfield shift of
0.1 ppm relative to the parent 4. This deshielding seems
to indicate the presence of a ring current in 7, most likely
a consequence of its aromaticity (vide infra).7,15
`
this may have to be reconsidered—at least vis-a-vis the
azlactone anions.
Mechanistic aspects. Both the exchange and condensa-
tion involve a sequence of two steps—deprotonation-
deuteration (exchange) and deprotonation-addition
(condensation). By the Hammond postulate,16 the tran-
sition states for all these processes would resemble the
reactive intermediate in each case, that is, I–III, and
the relative stabilities of these would determine the rela-
tive reactivity. These arguments would apply equally to
both the above mentioned steps, regardless of which is
rate determining. (We note that in all the cases, the con-
centrations of the deprotonating base and the electro-
philic acceptor remain the same for the three
competing substrates: the relative reactivity in each case
would then depend solely on the relative free energy of
activation for the overall process.)
Such is the ease of the Erlenmeyer synthesis that the
azlactones are generated, deprotonated and reacted
in situ, typically in a mixture of N-benzoylglycine, acetic
anhydride, sodium acetate and aromatic aldehyde.1
Also, as azlactones are apparently isoelectronic analogs
of acetic anhydride, the Erlenmeyer synthesis has often
been considered to be an extension of the Perkin con-
densation.1,21 However, this ignores the critical impor-
tance of the cyclic conjugation that is present in the
azlactone anion I, as demonstrated in the present study.
O-
O
Ar
Ar
O
O
In fact, the data in Table 1 lead to a rough measure of
the relative reactivities based on the initial rates approx-
imation,17 particularly in the case of the deuterium
exchange: the ratio at 5 min (T2) indicates that imidazo-
lone 4 is around four times as reactive as azlactone 1, as
explained briefly in the experimental note. (The corre-
N
N
Ph
2b
Ph
2a
Scheme 4. Possibility of fulvenoid aromaticity in Erlenmeyer products
2.

5766
S. Chandrasekhar, P. Karri / Tetrahedron Letters 47 (2006) 5763–5766
8. Katritzky, A. R. Handbook of Heterocyclic Chemistry;
Pergamon Press: Oxford, 1985; p 123.
9. VandenBerg, G. E.; Harrison, J. B.; Carter, H. E.;
Magerlein, B. J. Org. Synth. 1973, Coll. Vol. V, 946–948.
10. Takeuchi, H.; Hagiwara, S.; Eguchi, S. Tetrahedron 1989,
45, 6375–6386.
11. Hanson, R. N. J. Heterocycl. Chem. 1984, 21, 57–59.
12. Takasaki, M. Jpn. Kokai Tokkyo Koho JP 2003040865
(13 February 2003; CAN 138: 137165).
Thus, this study fills a puzzling lacuna in the mechanistic
understanding of this fascinating reaction, which contin-
ues to be an important method for the synthesis of ami-
no acids. Apart from the renewal of azlactone chemistry
mentioned above,2–5 the theoretical extension of the aro-
maticity concept7 to heterocyclic and other systems is of
much contemporary interest.8,14,22
13. Ref. 8, pp 118–119.
References and notes
14. Boyd, G. V. In Comprehensive Heterocyclic Chemistry;
Katritzky, A. R., Rees, C. W., Potts, K. T., Eds.;
Pergamon: Oxford, 1984; Vol. 6, pp 184–185.
15. Kemp, W. Organic Spectroscopy, 3rd ed.; Macmillan:
London, 1991; p 125.
1. Rosen, T. In Comprehensive Organic Synthesis; Trost, B.
M., Fleming, I., Heathcock, C. H., Eds.; Pergamon:
Oxford, 1991; Vol. 2, pp 402–407, and references cited
therein.
16. Ref. 6, p 215.
17. Hammes, G. G. Principles of Chemical Kinetics; Academic
Press: New York, 1978; p 7.
18. Goodman, M.; Levine, L. J. Am. Chem. Soc. 1964, 86,
2918–2922.
19. Neuenschwander, M. In Chemistry of Double-Bonded
Functional Groups; Patai, S., Ed.; John Wiley: Chichester,
1989; Chapter 16, pp 1131–1268.
2. Seebach, D.; Jaeschke, G.; Gottwald, K.; Matsuda, K.;
Formisano, R.; Chaplin, D. A. Tetrahedron 1997, 53,
7539–7556.
3. Boaz, N. W.; Debenham, S. D.; Large, S. E.; Moore, M.
K. Tetrahedron: Asymmetry 2003, 14, 3575–3580.
4. Yamada, K.; Oku, H.; Shinoda, S.-S.; Katakai, R. J. Pept.
Res. 2005, 65, 167–174.
20. Ertekin, K.; Alp, S.; Yalcin, I. Dyes Pigments 2005, 65,
33–38.
5. Trost, B. M.; Ja¨kel, C.; Plietker, B. J. Am. Chem. Soc.
2003, 125, 4438–4439.
21. Chandrasekhar, S.; Karri, P. Tetrahedron Lett. 2006, 47,
2249–2251.
6. March, J. Advanced Organic Chemistry, 4th ed.; John
Wiley: New York, 1992; p 251.
22. Schleyer, P. v. R. Chem. Rev. 2001, 101, 1115–1118.
7. Ref. 6, pp 40–67, and references cited therein.