Duality of mechanism in the tetramethylfluoroformamidinium hexafiuorophosphate-mediated synthesis of N-benzyloxycarbonylamino acid fluorides

Full text of DOI:10.1021/jo0009393

J . Org. Chem. 2001, 66, 5905-5910
5905
Sch em e 1
Du a lity of Mech a n ism in th e
Tetr a m eth ylflu or ofor m a m id in iu m
Hexa flu or op h osp h a te-Med ia ted Syn th esis
of N-Ben zyloxyca r bon yla m in o Acid
F lu or id es
Roberto Fiammengo, Giulia Licini,*, Alessia Nicotra,
Giorgio Modena, Lucia Pasquato, Paolo Scrimin,*,
Quirinus B. Broxterman,# and Bernard Kaptein#
acyl fluorides must be used (e.g., histidine or arginine
derivatives). Furthermore, the in situ formed fluorides
have been successfully employed for the coupling of some
CR-tetrasubstituted amino acids, like aminoisobutyric
acid, although this coupling reaction appears to be
sensitive to steric hindrance.7
Dipartimento di Chimica Organica, Universita` di Padova,
and Centro Meccanismi Reazioni Organiche del CNR,
via Marzolo 1, I-35131 Padova, Italy, and Organic
Chemistry & Biotechnology Section, DSM Research,
P.O. Box 18, 6160 MD Geleen, The Netherlands
Our interest in the use of short peptide frameworks
for the construction of catalysts or molecular devices8 has
prompted us to investigate the scope of the fluorination
reaction with TFFH. Particularly important was to
establish the reactivity of CR-tetrasubstituted amino
acids. Because of their ability to induce helical conforma-
tion even in short sequences9 (seven to eight amino acids,
i.e., two helical turns), they are particularly appealing
for the realization of simple and nevertheless highly
organized structures. The easy synthetic access to such
unnatural peptides is thus becoming a hot and, so far,
still unsolved problem. In fact, despite considerable
progress, both the synthesis of CR-tetrasubstituted amino
acid fluorides and their coupling reactions are very slow,
particularly for the more hindered compounds.10
Preliminary work on key substrates suggested the
occurrence of at least two concurrent pathways to the acyl
fluoride (2): (i) the direct conversion of the activated
intermediate into the product and (ii) the intramolecular
cyclization to the corresponding oxazolone (3) followed
by fluoride nucleophilic ring opening. The contribution
of the second process appeared to depend strongly on the
nature of the amino acid, the process being absent with
the natural amino acids while becoming more and more
relevant with the R-alkyl-R-methyl derivatives with the
increase of the bulkiness of the substitutents at the
R-carbon.
giulia.licini@unipd.it
Received J une 21, 2000
(Revised Manuscript Received J uly 2, 2001)
The use of protected amino acid fluorides as activated
forms of the N-protected R-amino acids both for solution-
phase1 and for solid-phase synthesis2 of peptides has
recently acquired a remarkable expansion.3 The main
reasons for their success may be ascribed to their unique
properties, namely, (i) their inherent stability after
purification even in the presence of tert-butyl-based
protecting groups, (ii) the scarce or even absent oxazolone
formation in the presence of tertiary bases, (iii) their
ability to afford efficient coupling even in the absence of
base, and (iv) the ability to react in high yield and
relatively short reaction time, also with sterically hin-
dered amino acids.1-3 Common procedures for the syn-
thesis of the fluorides call for the use of cyanuryl fluoride
in the presence of pyridine4 or (diethylamino)sulfur
trifluoride5 in the absence of base. More recently, Car-
pino’s group has introduced tetramethylfluoroformami-
dinium hexafluorophosphate (TFFH) as a fluoride source
in the presence of stoichiometric amounts of a tertiary
amine for the in situ formation of N-protected amino acid
fluorides from the corresponding carboxylic acids (Scheme
1).6 The procedure has the advantage of allowing the use
of acyl fluorides directly in the coupling reaction without
prior isolation. The reaction conditions are mild, and
therefore, such a process has been considered a benign
alternative to fluoride synthesis when non-shelf-stable
Resu lts a n d Discu ssion
We have examined two classes of benzyloxycarbonyl-
protected (Z) derivatives: those deriving from proteino-
genic R-amino acids (Z-glycine, 1a , and Z-valine, 1b) and
those deriving from unnatural R,R-disubstituted ones of
increasing bulkiness of the CR-substituents (Z-aminoisobu-
tyric acid (1c), Z-R-methyl-valine (1d ), and Z-R-methyl-
tert-leucine (1e)).
* To whom correspondence should be addressed. Fax (Italy): +39
049 8275239. E-mail for G.L.: giulia.licini@unipd.it. E-mail for P.S.:
paolo.scrimin@unipd.it.
University of Padova.
#
DSM Research.
All amino acids were available with the exception of
R-methyl-tert-leucine, 6, which was synthesized starting
(1) Carpino, L. A.; Sadat-Aalaee, D.; Guang Chao, H.; DeSelms. R.
H. J . Am. Chem. Soc. 1990, 112, 9651. Wenschuh, H.; Beyermann,
M.; El-Faham, A.; Ghassemi, S.; Carpino, L. A.; Bienert, M. J . Chem.
Soc., Chem. Commun. 1995, 669.
(2) Wenschuh, H.; Beyermann, M.; Krause, E.; Brudel, M.; Winter,
R.; Schu¨mann, M.; Carpino, L. A.; Bienert, M. J . Org. Chem. 1994,
59, 3275. Wenschuh, H.; Beyerman, M.; Haber, H.; Seydel, J . K.;
Krause, E.; Bienert, M.; Carpino, L. A.; El-Faham, A.; Albericio, F. J .
Org. Chem. 1995, 60, 405. Wenschuh, H.; Beyermann, M.; Rothemund,
S.; Carpino, L. A.; Bienert, M. Tetrahedron Lett. 1995, 36, 1247.
(3) Carpino, L. A.; Beyermann, M.; Wenschuh, H.; Bienert, M. Acc.
Chem. Res. 1996, 29, 268.
(7) Wenschuh, H.; Beyermann, M.; Winter, R.; Bienert, M.; Ionescu,
D.; Carpino, L. A. Tetrahedron Lett. 1996, 37, 5483.
(8) Scrimin, P.; Veronese, A.; Tecilla, P.; Tonellato, U.; Monaco, V.;
Formaggio, F.; Crisma, M.; Toniolo, C. J . Am. Chem. Soc. 1996, 118,
2505. Rossi, P.; Felluga, F.; Tecilla, P.; Formaggio, F.; Crisma, M.;
Toniolo, C.; Scrimin, P. J . Am. Chem. Soc. 1999, 121, 6948.
(9) Toniolo, C.; Crisma, M.; Formaggio, F.; Valle, G.; Cavicchioni,
G.; Pre´cigoux, G.; Aurby, A.; Kamphuis, J . Biopolymers 1993, 33, 1061.
Karle, I.; Balaram, P. Biochemistry 1990, 29, 6747.
(10) Polese, A.; Formaggio, F.; Crisma, M.; Valle, G.; Toniolo, C.;
Bonora, G. M.; Broxterman, Q. B.; Kamphuis, J . Chem.sEur. J . 1996,
2, 1104.
(4) Olah, G. A.; Nojima, M.; Kerekes, I. Synthesis 1973, 487.
(5) Kaduk, C.; Wenschuh, H.; Beyermann, M.; Forner, K.; Carpino,
L. A.; Bienert, M. Lett. Pept. Sci. 1996, 2, 385.
(6) Carpino, L. A.; El-Faham, A. J . Am. Chem. Soc. 1995, 117, 5401.
10.1021/jo0009393 CCC: $20.00 © 2001 American Chemical Society
Published on Web 08/17/2001

5906 J . Org. Chem., Vol. 66, No. 17, 2001
Notes
F igu r e 1. Time course of the TFFH-mediated fluorination of
1b (Z-ValF) in the presence of pyridine in CD2Cl2. Conditions
are as follows: [1b]0 ) 0.10 M; [pyridine]0 ) [TFFH]0 ) 0.15
M; T ) 25 °C.
F igu r e 2. Time course of the TFFH-mediated fluorination of
1d (Z-(RMe)Val) in the presence of pyridine in CD2Cl2. Condi-
tions are as follows: [1d ]0 ) 0.10 M; [pyridine]0 ) 0.13 M;
[TFFH]0 ) 0.17 M; T ) 25 °C.
Sch em e 2
Sch em e 3
from tert-butyl methyl ketone (Scheme 2).11 Thus, reac-
tion with potassium cyanide/ammonium carbonate af-
forded (()-5-tert-butyl-5-methyl-hydantoin, 5 (52% yield),
which was then hydrolyzed to the amino acid (()-6. The
hydantoin hydrolysis occurred only partially under basic
conditions (Ba(OH)2 in a 1:1 dioxane/water mixture for
15 days). Complete hydrolysis was achieved by further
treatment under acidic conditions (6 N HCl in a sealed
tube at 110 °C). Racemic R-methyl-tert-leucine hydro-
choride, (()-6, was obtained in 67% yield. The compound
(()-1e was synthesized via in situ silylation followed by
reaction with benzyl chloroformate (Scheme 2).12 Despite
the hindrance of the CR-substituents, carbamate (()-1e
was obtained in satisfactory yield (54%).
similar for both substrates indicating that there is little
influence of the bulkiness of the substituents at CR on
the rate and that no intermediates accumulate during
the reaction.
B. Cr-Tetr a su bstitu ted Am in o Acid s. The picture
of the fluorinations for the R-methyl-R-alkyl-substituted
amino acids is more complicated. As it may be appreci-
ated from the inspection of Figure 2, which shows the
behavior of 1d , the reaction becomes slower and, besides
the acyl fluoride 2d , two new R-methyl-valine derivatives
are formed. The first one, intermediate 3d , identified as
the 2-benzyloxy-4-isopropyl-4-methyl-5(4H)-oxazolone,
builds up at the early stages of the reaction, while the
second one, corresponding to the 4-isopropyl-4-methyl-
oxazolidine-2,5-dione 4d ([R-methyl-valine N-carboxy-
anhydride (NCA)], is detected at the last stages.
F lu or in a tion Rea ction s. The study of the fluorina-
tion reaction was carried out in CD2Cl2 at 25 °C in the
presence of a slight excess of pyridine as a base and 1,2-
dichoroethane as an internal standard (Scheme 1) and
1
monitored via H NMR. The choice of pyridine was made
on the basis of its absence of signals in the 1H NMR
spectrum in the region where selected signals of the
reagents and products appear.
A. P r otein ogen ic Am in o Acid s. The reaction of
TFFH with 1a (see Supporting Information) and 1b (the
time course of the latter is shown in Figure 1) proceeds
smoothly toward completion with formation of the acyl
fluorides (2a ,b) as the only detectable products. The rates
of product formation and reagent disappearance are very
The kinetic profile provides insights into the course of
the fluorination. It indicates that the activated interme-
diate formed upon reaction of the amino acid with TFFH
may either be directly converted into the acyl fluoride or
give an intramolecular cyclization to the oxazolone 3d
(see Scheme 3). Slow reaction of 3d with F- eventually
leads to the formation of the acyl fluoride. A concurrent,
acid-catalyzed, even slower reaction leads to the forma-
tion of the corresponding N-carboxyanhydride 4d .13,14
(11) Obrecht, D.; Spiegler, C.; Scho¨nholzer, P.; Mu¨ller, K.; Heim-
gartner, H.; Stierli, F. Helv. Chim. Acta 1992, 75, 1666.
(12) Bolin, D. R.; Sytwu, I.; Humiec, F.; Meienhofer, J . Int. J . Pept.
Protein Res. 1989, 33, 353.

Notes
J . Org. Chem., Vol. 66, No. 17, 2001 5907
Ta ble 1. Ra te Con sta n tsa Ca lcu la ted for th e
TF F H-Med ia ted F lu or in a tion of Z-Am in o Acid s 1a -e in
th e P r esen ce of P yr id in e in CH2Cl2 a t 25 °C
Sch em e 4
substrate k1 (M-1 s-1
)
k2 (s-1
)
k3 (s-1
)
k2/k3 k4 (M-1 s-1
)
1a
1b
1c
1d
1e
0.021
0.016
0.019
0.015
0.012
0.238
0.080
0.175
0.0057
b
b
0.050
0.016
3.50
0.35
0.0074
0.0077
0.0072
0.000 31 0.0061 0.05
a
Rate constant values have been calculated by least-squares
b
fitting ((20%) (ref 18). No formation of oxazolone.
The capability of oxazolones to react with nucleophiles
is well documented,15,16 and they have also been used as
activated intermediates in coupling reactions.16 The time
course of the reaction of 1d is qualitatively similar to
those observed for the fluorination of 1c and 1e (see
Supporting Information). There are, however, important
differences as to the time scale of the overall process and
the relative amount of oxazolone and N-carboxyanhy-
dride17 observed. The half-times for the formation of the
acylfluorides increase with the bulkiness of the substit-
uents at CR [1e (2900 s) > 1d (1500 s) > 1c (210 s)] as do
the amounts of oxazolone 3 and N-carboxyanhydride 4
formed. For instance, the maximum amount of oxazolone
3 formed from 1c, 1d , and 1e is, respectively, 20% (300
s), 52% (600 s), and 54% (1500 s).
Mech a n ism of th e Rea ction . The above experimen-
tal observations point to the mechanism outlined in
Scheme 3, which shows the dual pathway toward the acyl
fluoride: direct (a) and indirect (b, c). The indirect
pathway is only operative with the CR-tetrasubstituted
derivatives. As already mentioned, the fluorination of the
most hindered derivatives 1d and 1e yields also the
corresponding N-carboxyanhydride at the last stages of
the reaction. The latter may arise from acid-catalyzed
debenzylation of 3d ,e with formation of benzyl fluoride
which is eventually converted to the pyridinium salt
(Scheme 4).
Such a reaction pathway could be independently
demonstrated by following via 1H NMR the reaction of
3d ,e in the presence of 1 equiv of pyridine hydrochloride
in CD2Cl2.
Careful analysis of the concentration vs time profiles
for the different amino acids allowed us to determine
kinetic rate constants for the different steps of the
fluorination process (see Scheme 3).18 These values are
reported in Table 1.
A closer look at this table shows that while the
disappearance of the amino acid (k1) is little influenced
by the substituents at CR, all other reactions are strongly
dependent on the bulkiness of the groups present on that
carbon. These, in fact, involve the attack at the CdO
carbon with formation of a tetrahedral intermediate
which is very sensitive to steric hindrance in the R-posi-
tion as is well documented in the literature.19 Thus, k2
becomes more than 700 times slower for 1e compared
with 1a . The intramolecular cyclization to the oxazolone
3 (k3) is influenced not only by steric hindrance at CR but
also by the gem-dialkyl effect.20 These two effects operate
in opposite directions. The geminal alkyl substituents at
the R-carbon favor the conformation that leads to ring
closure. On the other hand, they make the intramolecular
attack to the carbonyl slower because of steric hindrance.
As a consequence, k3 is unmeasurably slow with the
derivatives of proteinogenic amino acids, becomes rela-
tively fast with 1c, and then decreases again with the
more hindered CR-tetrasubstituted derivatives. The k2/
k3 ratio reflects the relative amount of oxazolone vs
fluoride formation, and it clearly indicates that fluoride
formation for the three CR-tetrasubstituted amino acids
is only favored in the case of 1c, while for the other two
amino acids the ring closure reaction is largely the most
favored process.
Ma xim iza tion of Acyl F lu or id e P r od u ction . From
the mechanistic analysis we have described in the previ-
ous section, it is immediately clear that with CR-tetra-
substituted amino acids the formation of the oxazolones
is a problem that has to be dealt with. If one wants to
improve the production of the acyl fluoride, because the
oxazolone formation is an intrinsic property of the
particular substrate, he can only operate on step c.
We thus examined more closely the oxazolone ring
opening reaction. Oxazolone 3d was prepared by treat-
ment of 1d with 1-(3-dimethylaminopropyl)-3-ethylcar-
bodiimide hydrochloride (68% yield).14 Subsequently, we
followed via 1H NMR the reaction of 3d with TFFH in
the presence of 1 equiv of pyridine. The kinetic picture
is reported in Figure 3. The kinetic behavior shows the
slow formation of acyl fluoride 2d in low yield (16% after
42 000 s) together with a much larger amount of N-
carboxyanhydride 4d (39%). After approximately 12 000
s, no more acyl fluoride 2d is produced and no more
TFFH is consumed. However, upon addition of water (0.5
equiv at 60 000 s), the reaction starts again. Therefore,
very likely the acyl fluoride 2d formed initially derives
from the reaction of the nucleophilic fluoride ion that is
produced in small amounts at the beginning of the
reaction as a consequence of TFFH hydrolysis by small
quantities of water present in the reaction mixture.
Further addition of water generates new fluoride, restor-
ing the fluorination process. When the reaction is carried
out starting directly from the N-protected amino acid, the
production of the nucleophilic fluoride is the consequence
of the attack of the carboxylate on TFFH with formation
(13) 2-tert-Butoxy-5(4H)-oxazolones can be converted into the cor-
responding N-carboxyanhydride on standing in tert-butyl alcohol,
anhydrous acid, or in vacuo (see ref 14).
(14) Benoiton, L. N.; Chen, F. M. C. Can. J . Chem. 1981, 59, 384.
(15) For example, upon reaction with hydrogen chloride, 2-tert-
butoxy-5(4H)-oxazolones can afford the corresponding acyl chlorides
(see: Chen, F. M. F.; Lee, Y. C.; Benoiton, L. N. Int. J . Pept. Protein
Res. 1991, 38, 97 and ref 16).
(16) Benoiton, L. N. Biopolymers 1996, 40, 245 and references
therein.
(17) N-carboxyanhydrides 4d and 4e were detected only in the
fluorination of 1d and 1e (15% and 60% yields, respectively).
(18) Kinetic rate constants have been calculated by least-squares
fittings. (ScientistTM; MicroMath Scientific Software: Salt Lake City,
UT, 1995.)
(19) March, J . Advanced Organic Chemistry, Reactions, Mechanisms
and Structure, 4th ed.; J ohn Wiley & Sons: New York, 1992.
(20) Bruice, T. C.; Lightstone, F. C. Acc. Chem. Res. 1999, 32, 127.
Sammes, P.; Weller, D. J . Synthesis 1995, 1205 and references therein.

5908 J . Org. Chem., Vol. 66, No. 17, 2001
Notes
F igu r e 3. Time course of the TFFH-mediated fluorination of
3d (Z-(RMe)ValOx) in the presence of pyridine in CD2Cl2.
F igu r e 4. Time course of the TFFH-mediated fluorination of
1d (Z-(RMe)Val) in the presence of pyridine in CD2Cl2. After
21 min, HF/Py (1.5 equiv) was added. Conditions are as
follows: [1d ]0 ) [pyridine]0 ) 0.10 M; [TFFH]0 ) 0.15 M; T )
25 °C.
Conditions are as follows: [3d ]0 ) 0.10 M; [pyridine]0
[TFFH]0 ) 0.15 M; T ) 25 °C.
)
of the active intermediate. Thus, an extra source of
nucleophilic fluoride is required in order to accelerate the
fluorination process. For this purpose, the use of “naked”
fluoride nucleophiles such as tetrabutylammonium fluo-
ride (TBAF)21 or potassium fluoride22 was explored.
Reaction with different TBAF sources (TBAF‚3H2O and
TBAF/SiO2) afforded an extended decomposition of oxa-
zolone 3d without any detectable acyl fluoride formation.
On the other hand, reaction with dry KF in the presence
of 18-crown-6 in benzene did not result in any formation
of products, and the starting oxazolone 3d was recovered
unchanged. However, addition of a slight excess of a HF/
pyridine 1:1 complex resulted in the fast conversion of
the oxazolone 3d into the acyl fluoride 2d (40% after
16 000 s)23 together with the formation of only 4% of
N-carboxyanhydride 4d .
process being absent in the natural amino acid deriva-
tives while in the R-alkyl-R-methyl ones it parallels the
steric bulkiness of the alkyl substituent (Me < i-Pr <
t-Bu). The formation of the oxazolone with the subse-
quent slow ring opening seems to be one of the main
causes of the decreased reactivity of the R,R-disubstituted
amino acids. Furthermore, because of the longer reaction
time, the formation of the corresponding N-carboxy-
anhydrides may become a significant side reaction.
Addition of “HF” species to the reaction mixture ap-
preciably accelerates the ring opening of the oxazolone
and therefore the R,R-disubstituted acyl fluoride forma-
tion. Studies concerning the optimization of the reaction
conditions for the synthesis of highly hindered R,R-
disubstituted amino acid fluorides are currently being
pursued in our laboratories.
To determine whether the formation of the acyl fluoride
could be maximized also starting from the amino acid
and TFFH, we ran the reaction directly on 1d . Figure 4
1
shows the time course of the process monitored by H
Exp er im en ta l Section
NMR and the effect of the addition of HF/pyridine after
ca. 1200 s. It is clearly apparent that HF/pyridine
converts the formed oxazolone into the acyl fluoride at a
much higher rate in this case, too. An analogous behavior
was observed with 1e (see Supporting Information). The
failure of Bu4N+F- or KF/18-crown-6 to afford the fluoride
2d indicates that, to occur, the F- nucleophilic attack to
the ozaxolone requires acid catalysis. It is noteworthy
that the conversion of the intermediate oxazolone to the
corresponding N-carboxyanhydride, which is also an acid-
catalyzed process, is much less accelerated and, conse-
quently, the overall amount of 4d formed is much lower.
In conclusion, our study points out that in the TFFH-
mediated Z-amino acid fluorination at least two concur-
rent pathways yielding the acyl fluoride can take place:
the direct conversion of the activated intermediate into
the product and the intramolecular cyclization to the
corresponding oxazolone followed by fluoride nucleophilic
ring opening. The contribution of the second process
strongly depends on the nature of the amino acid, the
Gen er a l Meth od s. 1H NMR spectra were recorded at 200 or
250 MHz, 13C NMR at 62.9 MHz, and 2D-correlated spectra at
400 MHz, at 25 °C using Me4Si as internal standard reference.
Cq, CH, CH2, and CH3 were determined by distortionless
enhancement by polarization transfer (DEPT) pulse sequence.
Mass spectra were obtained by EI, matrix-assisted laser desorp-
tion ionization (MALDI) (matrix, R-cyano-hydroxycinnamic acid),
or electrospray (ESI) (mobile phase, acetonitrile) ionizations.
Radial chromatography has been performed using silica gel
plates. Melting points are uncorrected.
Ch em ica ls. Enantiopure L-R-methyl valine, 1d , was prepared
by DSM Research by enzymatic resolution of (()-R-methyl valine
amide.24 TFFH has been synthesized following Carpino’s pro-
cedure.6 Z-Glycine fluoride (2a ),25 Z-L-valine fluoride (2b),25
Z-aminobutyric fluoride (2c),7 Z-L-R-methyl valine (1d ),10 Z-L-
R-methyl valine fluoride (2d ),10 and 2-benzyloxy-4,4-dimethyl-
5(4H)-oxazolone (3c)26 have been prepared following previously
reported procedures. Dichloromethane was distilled over CaH2
and stored over molecular sieves. Z-Glycine (1a ), Z-L-valine
(1b), Z-aminoisobutyric acid (1c), dry pyridine, diisopropylethyl-
(24) Kruizinga, W. Z.; Boster, J .; Kellogg, R. M.; Kamphuis, J .;
Boesten, W. H. I.; Mejier, E. M.; Schoemaker, H. E. J . Org. Chem. 1988,
53, 1826. Kaptein, B.; Boesten, W. H. J .; Broxterman, Q. B.; Peters, P.
J . H.; Schoemaker, H. E.; Kamphuis, J . Tetrahedron: Asymmetry 1993,
4, 1113.
(25) Carpino, L. A.; Mansour, E.-S. M. E.; Sadat-Aalaee, D. J . Org.
Chem. 1991, 56, 2611.
(21) Clark, J . H.; Smith, D. K. Tetrahedron Lett. 1985, 26, 2233.
(22) Liotta, C. L.; Harris, H. P. J . Am. Chem. Soc. 1974, 96, 2250.
Smyth, T. P.; Carey, A.; Hodnett, B. K. Tetrahedron 1995, 51, 6363.
(23) The reaction of 2-(9-fluorenylmethoxy)-4-isopropyl-5(4H)-ox-
azolone with HCl has been reported to afford quantitatively the
corresponding N-9-fluorenylmethoxycarbonyl valine chloride (ref 15).
(26) Formaggio, F.; Pantano, M.; Crisma, M.; Bonora, G. M.; Toniolo,
C.; Kamphuis, J . J . Chem. Soc., Perkin Trans. 2 1995, 1097.

Notes
J . Org. Chem., Vol. 66, No. 17, 2001 5909
amine (DIEA), trimethylsilyl chloride (TMSCl), 1-(3-dimethyl-
aminopropyl)-3-ethylcarbodiimide hydrochloride (EDC‚HCl), and
cyanuryl fluoride are commercial products and have been used
without further purification. Pyridine/hydrogen fluoride 1:1
complex has been obtained by addition of 1.1 mL of dry pyridine
to 0.4 mL of pyridinium poly(hydrogen fluoride).
Syn th esis of (()-5-ter t-Bu tyl-5-m eth yl-h yd a n toin , 5. tert-
Butyl methyl ketone (50.0 mL, 0.4 mmol) was dissolved in 1.2
L of a 1:1 ethanol/water solution to which ammonium carbonate
(160.6 g, 1.7 mmol) and potassium cyanide (55.4 g, 0.8 mmol)
were added. The mixture was heated at 60 °C for 2.5 h until
disappearance of the starting material (TLC, CHCl3/EtOH 9:1).
The solution was concentrated under vacuum, and the precipi-
tate was filtered. Crystallization from EtOH/H2O afforded
hydantoin 5 (35.4 g, 52%) as a white solid, which readily
sublimes; mp 224-226 °C (lit.11 218-219 °C). 1H NMR (CD3-
OD, δ, ppm): 1.54 (3H, s), 1.21 (9H, s). 13C NMR (CD3OD, δ,
ppm): 180.69 (C), 159.69 (C), 69.53 (C), 37.86 (C), 25.40 (CH3),
19.62 (CH3). FT-IR (KBr, cm-1): 3220, 1732. Anal. Calcd for
was removed, and the resulting product was purified by radial
chromatography (EtOAc/petroleum ether 4:96).
2-Ben zyloxy-4-isop r op yl-4-m eth yl-5(4H)-oxa zolon e, 3d .
Oxazolone 3d was obtained in 84% yield as a colorless oil. 1H
NMR (CDCl3, δ, ppm): 7.40 (5H, m), 5.34 (2H, m), 1.98 (1H, m),
1.42 (3H, s), 0.98 (3H, d, J ) 6.7 Hz), 0.84 (3H, d, J ) 6.7 Hz).
13C NMR (CDCl3, δ, ppm,): 178.56 (C), 156.91 (C), 134.25 (C),
128.83 (CH), 128.58 (CH), 128.46 (CH), 73.35 (C), 71.44 (CH2),
35.16 (CH), 22.09 (CH3), 17.06 (CH3), 16.39 (CH3). [R]25 -30.1
D
(c 1.1, CHCl3). FT-IR (neat, cm-1): 3446, 1830, 1686. EI-MS m/z
(relative intensity): 347 (M+, 3), 205 (7), 91 (100). Anal. Calcd
for C14H17NO3: C, 68.00; H, 6.93; N, 5.66. Found: C, 68.12; H,
6.99; N, 5.61.
2-Ben zyloxy-4-ter t-bu tyl-4-m eth yl-5(4H)-oxa zolon e, 3e.
Oxazolone 3e was obtained in 98% yield as a colorless oil. 1H
NMR (CDCl3, δ, ppm): 7.39 (5H, m), 5.34 (2H, m), 1.40 (3H, s),
0.97 (9H, s). 13C NMR (CDCl3, δ, ppm): 177.97 (C), 156.60 (C),
134.39 (C), 128.83 (CH), 128.56 (CH), 75.35 (C), 71.35 (CH2),
36.75 (C), 24.62 (CH3), 19.66 (CH3). FT-IR (neat, cm-1): 3446,
1828, 1684. EI-MS m/z (relative intensity): 261 (M+, 4), 205 (9),
91 (100). Anal. Calcd for C15H19NO3: C, 68.94; H, 7.33; N, 5.36.
Found: C, 69.07; H, 7.41; N, 5.27.
C
18H14N2O2: C, 56.45; H, 8.29; N, 16.46. Found: C, 56.55; H,
8.36; N, 16.54.
Syn th esis of (()-r-Meth yl-ter t-leu cin e, 6. Hydantoin (()-5
(2.16 g, 13 mmol) was dissolved in 30 mL of water containing
Ba(OH)2‚8H2O (32.31 g, 86 mmol). The mixture was refluxed
for 15 days. The reaction was followed by TLC (CHCl3/EtOH
9:1 for reagent; t-BuOH/H2O/AcOH 3:1:1 for product). Then, the
mixture was cooled to room temperature, filtered, concentrated
under vacuum, and treated twice with solid CO2 to precipitate
barium as BaCO3. After the solvent was removed under vacuum,
the crude of the reaction, containing a mixture of reagent, ureic
acids, and hydantoin, was treated with 50 mL of HCl (6 N) in a
sealed vial at 110 °C for 22 h. The solvent was removed under
vacuum, and the solid was treated with water three times, and
the solvent was evaporated again to dryness. Washing with
EtOAc and lyophilization of the aqueous phase afforded (()-6
hydrochloride11 (1.54 g, 67%) as a white solid; mp 310-313 °C
(lit.27 309-312 °C). 1H NMR (CD3OD, δ, ppm): 1.56 (3H, s), 1.06
(9H, s). 13C NMR (CD3OD, δ, ppm): 173.40 (C), 67.30 (C), 36.70
(C), 25.99 (CH3), 18.83 (CH3). FT-IR (KBr, cm-1): 3424, 2976,
1735.
Syn th esis of (() Z-r-Meth yl-ter t-leu cin e, 1e. Hydrochlo-
ride 6 (1.41 g, 7.8 mmol) was dissolved in 50 mL of dry CH2Cl2.
DIEA (7.0 mL, 40.9 mmol) and TMSCl (3.00 mL, 23.6 mmol)
were slowly added, and the mixture was refluxed for 1 h. After
the mixture was cooled to 0 °C, benzyl chloroformate (Z-Cl, 1.10
mL, 7.7 mmol) was added. The reaction was stirred at 0 °C for
1 h and then warmed to room temperature. The course of the
reaction was monitored by TLC (CHCl3/EtOAc/AcOH 9:0.8:0.2),
and the pH was maintained in the 8-9 interval by addition of
DIEA. After 3 days, additional Z-Cl (0.8 mL, 5.6 mmol) was
added. After 8 days, the mixture was treated with 100 mL of
KHSO4 and vigorously stirred for 2 h. The organic phase was
separated, and the aqueous phase was extracted with CH2Cl2.
The aqueous phases were then acidified with 5% HCl to pH 1-2
and extracted with CH2Cl2. The organic phases were extracted
with 0.5 M NaOH, washed with brine and water, and dried over
MgSO4, and the solvent was removed under vacuum. Carbamate
1e was obtained as a white foam (1.19 g, 55%); mp 105-108 °C.
1H NMR (CDCl3, δ, ppm): 7.33 (5H, m), 5.38 (1H, br s), 5.10
(2H, s), 1.64 (3H, s), 1.05 (9H, s). 13C NMR (CD3OD, δ, ppm):
177.99 (C), 155.77 (C), 136.21 (C), 128.50 (CH), 128.16 (CH),
66.88 (CH2), 65.09 (C), 36.71 (C), 25.65 (CH3), 18.09 (CH3). FT-
IR (neat, cm-1): 3336, 1706, 1657. MALDI-TOF MS m/z: 279.9
[MH]+; calcd 280.3. Anal. Calcd for C15H21NO4: C, 64.50; H, 7.58;
N, 5.01. Found: C, 64.71; H, 7.62; N, 4.89.
Syn th esis of (()-Z-r-Meth yl-ter t-leu cin e F lu or id e, 2e.
Z-R-Methyl-tert-leucine, 1e (96 mg, 0.4 mmol), was dissolved in
dry CH2Cl2 (10 mL). After the mixture was cooled to 0 °C,
pyridine (60 µL, 0.7 mmol) and cyanuryl fluoride (80 µL, 0.9
mmol) were added. The solution was slowly warmed to room
temperature. After 3 h, crushed ice was added to the reaction
mixture along with additional CH2Cl2. The organic phase was
washed with ice-cold water and dried over MgSO4, and the
solvent was removed under vacuum. (()-Z-R-Methyl-tert-leucine
fluoride 3e (78 mg, 84%) was recovered as a colorless oil. 1H
NMR (CDCl3, δ, ppm): 7.34 (5H, m), 5.11 (2H, s), 1.50 (3H, d,
J H-F ) 1.5 Hz), 1.06 (9H, s). 13C NMR (CDCl3, δ, ppm): 163.18
(C, d, J ) 380.0 Hz), 155.56 (C), 135.80 (C), 128.75 (CH), 128.57
(CH), 67.23 (CH2), 36.85 (C, J ) 47.1 Hz), 24.93 (CH3, J ) 56.9
Hz), 18.57 (CH3, J ) 37.7 Hz). FT-IR (neat, cm-1): 3446, 1837,
1684. MALDI-TOF MS m/z: 282.0, calcd 282.1.
Syn th esis of 4-Alk yl-4-m eth yl-oxa zolid in e-2,5-d ion es 4d
a n d 4e fr om th e Cor r esp on d in g Oxa zolon es 3d a n d 3e. In
a screw-capped 5 mm NMR tube, 0.01 mmol of oxazolones 3d
or 3e was dissolved in 0.50 mL of CD2Cl2. Pyridine hydrochoride
(11.5 mg, 0.01 mmol) was added, and the reactions were
monitored via 1H NMR. The immediate disappearance of the
starting materials was observed together with the product
formation, which was complete after 24 h. The N-carboxy-
anhydrides 4d and 4e and benzyl pyridinium chloride 729 are
stable in CD2Cl2 for more than 1 month (see Supporting
Information). Filtration over silica gel of both reaction mixtures
allowed the recovery in 25% and 28% yields, respectively, of
N-carboxyanhydrides 4d and 4e for which the spectroscopic data
are consistent with those of the products obtained via the
procedure described later.29
Heteronuclear multiple-bond correlation (HMBC) and nuclear
Overhauser effect spectroscopy (NOESY) experiments were
useful in the identification of products 4d and 7 in the reaction
mixture deriving from 1d and for the assignment of their 1H
and 13C NMR signals (Figure 5, see also Supporting Informa-
tion). The spectral data reported below are those of the reaction
mixtures.
4-Isop r op yl-4-m eth yl-oxa zolid in e-2,5-d ion e, 4d . 1H NMR
(CD2Cl2, δ, ppm): 9.23 (1H, br s), 2.03 (1H, sept, J ) 6.8 Hz),
1.47 (3H, s), 1.02 (3H, d, J ) 6.8 Hz), 0.91 (3H, d, J ) 6.8 Hz).
13C NMR (CD2Cl2, δ, ppm,): 174.00 (C), 151.86 (C), 66.85 (C),
35.20 (CH), 21.67 (CH3), 16.97 (CH3), 16.31 (CH3).
Syn th esis of Oxa zolon es 3d a n d 3e (via EDC‚HCl).14 The
Z-amino acid (1d or 1e, 2.0 mmol) was dissolved in dry CH2Cl2
(20 mL). After the solution was cooled to 0 °C, EDC‚HCl (460
mg, 3.0 mmol) was added, and the reaction mixture was allowed
to slowly warm to room temperature. After the mixture was
stirred for 2-3 h, a cold 0.1 M KHSO4 solution was added. The
organic phase was separated and washed with cold water, with
cold 5% NaHCO3 aqueous solution, and with cold water a second
time. After the organic phase was dried over MgSO4, the solvent
(28) Buckley, N.; Maltby, D.; Burlingame, A. L.; Oppenheimer, N.
J . J . Org. Chem. 1996, 61, 2753.
(29) 1H NMR resonances of authentic 4d and 4e in CD2Cl2 and
CHCl3 have significant shifts with respect to the in situ formed
products obtained from oxazolones 3d or 3e in the presence of pyridine
hydrochoride. The identity of the two compounds was ascertained by
adding authentic 4d and 4e to the reaction mixtures obtained from
3d and 3e, respectively. The remarkable differences of the 1H and 13C
NMR resonances are very likely due to specific interactions with benzyl
pyridinium chloride, 7.
(27) Prochazka, M. Collect. Czech. Chem. Commun. 1977, 42, 2394.

5910 J . Org. Chem., Vol. 66, No. 17, 2001
Notes
added. The reactions were monitored via 1H NMR at 25 °C by
integration of the signals at the selected chemical shifts listed
in Table S-1 (see Supporting Information) at the reaction times
reported in Figures 1 and 2 and Figures S-1, S-2, and S-3 (see
Supporting Information).
P r oced u r e for th e Kin etic Stu d y of TF F H-Med ia ted
F lu or in a tion Rea ction of 2-Ben zyloxy-4-isop r op yl-4-m eth -
yl-5(4H)-oxa zolon e, 3d . In a screw-capped 5 mm NMR tube,
3d (13.43 mg, 0.054 mmol) was dissolved in 0.50 mL of CD2Cl2
together with 2 µL of 1,2-dichoroethane as an internal standard.
TFFH (20.45 mg, 0.077 mmol) and, after 5 min, dry pyridine (4
µL, 0.039 mmol) were added. After 64 200 s, 0.5 equiv of water
(0.5 µL) was added. The reaction was monitored via 1H NMR at
25 °C by integration of the signals at the selected chemical shifts
listed in Table S-1 at the reaction times reported in Figure 3.
P r oced u r e for th e Kin etic Stu d y of HF /P yr id in e 1:1
Com p lex-Med ia ted F lu or in a tion Rea ction of 2-Ben zyloxy-
4-isop r op yl-4-m eth yl-5(4H)-oxa zolon e, 3d . In a screw-capped
5 mm NMR tube, oxazolone 3d (13.43 mg, 0.054 mmol) was
dissolved in 0.500 mL of CD2Cl2, together with 2 µL of 1,2-
dichoroethane as an internal standard. A total of 7.3 µL (1.4
equiv) of a 1:1 HF/pyridine solution was added. The reaction
F igu r e 5. HMBC correlations (- - -) and NOESY (-) observed
for 4d and 7.
4-ter t-Bu tyl-4-m eth yl-oxa zolid in e-2,5-d ion e, 4e. 1H NMR
(CD2Cl2, δ, ppm): 9.21 (1H, br s), 1.48 (3H, s), 1.04 (9H, s). 13C
NMR (CD2Cl2, δ, ppm,): 173.23 (C), 151.90 (C), 69.05 (C), 37.25
(C), 22.58 (CH3), 18.99 (CH3).
Ben zyl P yr id in iu m Ch lor id e, 7.28 1H NMR (CD2Cl2, δ,
ppm): 9.51 (2H, d, J ) 5.5 Hz), 8.42 (1H, t, J ) 8.0 Hz), 8.05
(2H, m), 7.57-7.70 (2H, m), 7.31-7.43 (3H, m), 6.20 (2H, s). 13
C
NMR (CD2Cl2, δ, ppm,): 145.30 (CH), 145.15 (CH), 133.29 (C),
129.87 (CH), 129.52 (CH), 129.48 (CH), 128.43 (CH), 64.31 (CH2).
ESI-TOF MS m/z: 170.107 (100%), 171.1092 (12%); calcd
170.096 (100%), 171.100 (12%).
1
was monitored via H NMR at 25 °C by integration of the signals
at the selected chemical shifts listed in Table S-1 at the reaction
times reported in Figure S-4 (see Supporting Information).
P r oced u r e for th e Kin etic Stu d y of (TF F H + HF /P y)-
Med ia ted F lu or in a tion Rea ction of Z-Am in o Acid s 1d a n d
1e. In a screw-capped 5 mm NMR tube, 0.05 mmol of Z-amino
acid 1d or 1e was dissolved in 0.500 mL of CD2Cl2 together with
2 µL of 1,2-dichoroethane (DCE) as an internal standard. TFFH
(19.88 mg, 0.075 mmol) and, after 5 min, dry pyridine (5 µL,
0.049 mmol) were added. After 21 min, 0.075 mmol of HF/
pyridine 1:1 complex was added. The reactions were monitored
Syn th esis of N-Oxa zolid in e-2,5-d ion es 4d a n d 4e fr om
Z-r-Meth yl Am in o Acid s 1d a n d 1e. In a 1 mL screw-capped
vial, Z-R-methyl amino acids 1 (1.06 mml) were added to SOCl2
(0.500 mL) under magnetic stirring. The mixture was warmed
to 60 °C, and then, after 30 min, the reaction was cooled to room
temperature, and the solvent was removed under vacuum. The
products were purified by crystallization.
4-Isopr opyl-4-m eth yl-oxazolidin e-2,5-dion e, 4d. Yield 57%,
mp 54-56 °C (crystallized from ethyl acetate/pentane), [R]25
D
1
via H NMR at 25 °C by integration of the signals at the selected
-33.5 (c 1.0, chloroform). 1H NMR (CDCl3, δ, ppm): 6.60 (1H,
br s), 2.08 (1H, sept, J ) 6.5 Hz), 1.53 (3H, s), 1.02 (3H, d, J )
6.5 Hz), 0.99 (3H, d, J ) 6.5 Hz). 13C NMR (CDCl3, δ, ppm,):
170.66 (C), 151.86 (C), 66.79 (C), 34.99 (CH), 21.99 (CH3), 17.07
(CH3), 16.14 (CH3). IR (neat, cm-1): 3300, 1853, 1780. Anal.
Calcd for C7H11NO3: C, 53.50; H, 7.05; N, 8.91. Found: C, 53.62;
H, 7.11; N, 8.83.
chemical shifts listed in Table S-1 (see Supporting Information)
at the reaction times stated in Figures 4 and S-5 (see Supporting
Information).
Ack n ow led gm en t. We thank Dr. Patrizia Polverino
de Laureto for the MS MALDI-TOF spectra, Dr. Mas-
similiano Forcato and Dr. Alessandro Scarso for the 2D
NMR analysis, and Prof. C. Toniolo and Prof. F. For-
maggio (University of Padova) for useful discussions.
4-ter t-Bu tyl-4-m eth yl-oxazolidin e-2,5-dion e, 4e. Yield 75%,
mp 138-140 °C (sublimes, crystallized from dichloromethane/
pentane). 1H NMR (CDCl3, δ, ppm): 6.62 (1H, br s), 1.54 (3H,
s), 1.08 (9H, s). 13C NMR (CDCl3, δ, ppm): 171.63 (C), 152.69
(C), 69.04 (C), 37.26 (C), 24.46 (CH3), 19.51 (CH3). IR (neat,
cm-1): 3250, 1836, 1783. Anal. Calcd for C8H13NO3: C, 56.13;
H, 7.65; N, 8.18. Found: C, 56.22; H, 7.60; N, 8.11.
Su p p or tin g In for m a tion Ava ila ble: Kinetic profiles
1
(Figures S-1-S-5), selected H and 13C NMR spectra (Figures
S-6 and S-7), NOESY spectrum (Figure S-8), and HMBC
P r oced u r e for th e Kin etic Stu d y of TF F H-Med ia ted
F lu or in a tion Rea ction of Z-Am in o Acid s 1a -e. In a screw-
capped 5 mm NMR tube, 0.05 mmol of Z-amino acid 1 was
dissolved in 0.500 mL of CD2Cl2, together with 2 µL of 1,2-
dichoroethane as an internal standard. TFFH (19.88 mg, 0.075
mmol) and, after 5 min, dry pyridine (5 µL, 0.049 mmol) were
correlations (Figure S-9) of in situ formed derivatives 4d and
7 in CD2Cl2 and selected H NMR chemical shifts (Table S-1).
This material is available free of charge via the Internet at
http://pubs.acs.org.
1
J O0009393