Synthesis of carbamoyl azides from primary amines and carbon dioxide under mild conditions

Article abstract of DOI:10.1021/jo702506v

(Chemical Equation Presented) Treatment of amines under a carbon dioxide atmosphere with tetramethylphenylguanidine (PhTMG) and diphenylphosphoryl azide (DPPA) in acetonitrile below 0°C provides carbamoyl azides in high to excellent yields. In addition, e

Full text of DOI:10.1021/jo702506v

SCHEME 1
Synthesis of Carbamoyl Azides from Primary
Amines and Carbon Dioxide under Mild
Conditions
Eduardo Garc´ıa-Egido, Miryam Ferna´ndez-Sua´rez, and
Luis Mun˜oz*
Departamento de Qu´ımica Orga´nica, UniVersidade de Vigo,
E-36310 Vigo, Spain
lmunoz@uVigo.es
via a Curtius rearrangement,4 as well as by radical azidonation
of aldehydes.5 In addition, they have also been prepared by
degradation of tertiary amines.6
ReceiVed NoVember 21, 2007
In this paper, we present a new method for the preparation
of carbamoyl azides under mild conditions through the one-pot
carbamoylation of primary amines using carbon dioxide, a
guanidine as a base and diphenyl phosphoryl azide (DPPA)
(Scheme 1).
It is well-known that an amine and carbon dioxide are in
equilibrium with a carbamate anion in a basic medium. If this
equilibrium is sufficiently shifted toward the carbamate, the
anion can be alkylated in good yield with several electrophiles.7
One approach to shift the equilibrium is the use of a pentaalkyl
guanidine as the base. Guanidines are non-nucleophilic and
relatively strong bases.8 Their basicity has been explained on
the basis of the exceptional stability of the guanidinium cation.
This cation has a highly delocalized charge, which has allowed
its use for the stabilization of polydentate anions.9 It is accepted
that in the aforementioned equilibrium, guanidinium cations
stabilize the carbamate anions effectively, thus increasing their
concentration in solution.
We chose 1,1,3,3-tetramethyl-2-phenylguanidine (PhTMG)
as the base in the reaction because we found that the structure
of the guanidine has a negligible effect on the outcome of the
reaction. In addition, PhTMG could be easily prepared in high
yield by slightly modifying the procedure of Bredereck.10
The reaction was carried out in acetonitrile. Several studies
indicated that guanidines have a higher basicity in acetonitrile
than in other solvents.11 Furthermore, since the anions are only
weakly solvated by acetonitrile, the stabilization of the carbam-
ate anion by the guanidinium cation is favored.12 In addition,
Treatment of amines under a carbon dioxide atmosphere with
tetramethylphenylguanidine (PhTMG) and diphenylphos-
phoryl azide (DPPA) in acetonitrile below 0 °C provides
carbamoyl azides in high to excellent yields. In addition,
epimerization is not observed when optically pure R-amino
esters are used as substrates.
Caution: Carbamoyl azides are potentially explosiVe com-
pounds. Reactions must be carried out behind a safety shield.
Knowledge of the carbamoyl azide functionality dates back
to the end of the last century. However, the use of these
compounds as useful synthetic intermediates has been limited,
mainly because of their presumed instability and the harsh
conditions necessary for their preparation.1 In fact, some
carbamoyl azides, as other azides, are considered to be powerful
explosives. In addition, carbamoyl azides are sensitive to heat
and to both acidic and basic conditions. As a consequence,
examples of the use of carbamoyl azides in synthesis have been
very scarce over the past few years.2 However, one of the most
outstanding examples is the synthesis of biotin carried out by
De Clerck.3
There are three standard procedures for carbamoyl azide
synthesis: (a) diazotization of semicarbazides and related
compounds; (b) treatment of isocyanates with hydrazoic acid;
(c) reaction of carbamoyl chlorides with sodium azide. However,
all of these methods have several drawbacks, especially with
sensitive substrates. First, yields are poor to moderate in most
cases. Second, the reagents are toxic and explosive (hydrazoic
acid). In addition, several intermediates, such as carbamoyl
chlorides and isocyanates, are rather unstable toward nucleo-
philic attack.
(4) Affandi, H.; Bayquen, A. V.; Read, R. W. Tetrahedron Lett. 1994,
35, 2729.
(5) (a) Marinescu, L.; Thinggaard, J.; Thomsen, I. B.; Bols, M. J. Org.
Chem. 2003, 68, 9453. (b) Marinescu, L. G.; Pedersen, C. M.; Bols, M.
Tetrahedron 2005, 61, 123. (c) Pedersen, C. M.; Marinescu, L. G.; Bols,
M. Org. Biomol. Chem. 2005, 3, 816.
(6) Gumaste, V. K.; Deshmukh, A. R. A. S. Tetrahedron Lett. 2004, 45,
6571.
(7) (a) McGhee, W.; Riley, D. J. Org. Chem. 1995, 60, 6205. (b)
McGhee, W.; Riley, D.; Christ, K.; Pan, Y.; Parnas, B. J. Org. Chem. 1995,
60, 2820.
(8) Barton, D. H. R.; Elliot, J. D.; Ge´ro, S. D. J. Chem. Soc., Perkin
Trans. 1 1982, 2085.
(9) Boyle, P. H.; Convery, M. A.; Davis, A. P.; Hosken, G. D.; Murray,
B. A. J. Chem. Soc., Chem. Commun. 1992, 239.
(10) Bredereck, H.; Bredereck, K. Chem. Ber. 1961, 94, 2278.
(11) Leffek, K. T.; Pruszynski, P.; Thanapaalasingham, K. Can. J. Chem.
1989, 67, 590.
More efficient preparations of carbamoyl azides have been
achieved from carboxylic acids and carboxylic acid chlorides
(1) Lieber, E.; Minnis, R. L.; Rao, C. N. R. Chem. ReV. 1965, 65, 377.
(2) (a) Lwowski, W.; de Mauriac, R. A.; Thompson, M.; Wilde, R. E.;
Chen, S.-Y. J. Org. Chem. 1975, 40, 2608. (b) Del Signore, G.; Fioravanti,
S.; Pellacani, L.; Tardella, P. A. Tetrahedron 2001, 57, 4623.
(12) Liotta, C. L.; Grisdale, E. E.; Hopkins, H. P., Jr. Tetrahedron Lett.
1975, 48, 4205.
(3) Deroose, F. D.; De Clercq, P. J. Org. Chem. 1995, 60, 321.
10.1021/jo702506v CCC: $40.75 © 2008 American Chemical Society
Published on Web 03/04/2008
J. Org. Chem. 2008, 73, 2909-2911
2909

TABLE 1. Yields of Carbamoyl Azides from Various Amines
SCHEME 2
entry
amine (1)
2
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
pentane-1,5-diamine (a)
benzylamine (b)
2-(1-cyclohexenyl)ethylamine (c)
cyclohexylamine (d)
homoveratrylamine (e)
allylamine (f)
69
74
84
76
76
71
80
90
88
88
86
79
90
81
84
82
results of this analysis confirm the carbamoyl azide structure
for all the compounds and show the typical structural measure-
ments for the carbamoyl azide functionality. The azide is nearly
linear with an insignificant distortion in the lattice.
isobutylamine (g)
HCl‚L-Val-OMe (h)
HCl‚D-Val-OMe (i)
HCl‚L-Glu(OMe)-OMe (j)
HCl‚L-Trp-OMe (k)
pTsOH‚LAsp(OBn)-OBn (l)
HCl‚L-Ile-OMe (m)
TFA‚L-Met-OMe (n)
TFA‚L-Phe-OMe (o)
HCl‚D-Phe-OMe (p)
When the reaction was carried out with L-histidine methyl
ester 1q as the substrate (Scheme 2) the carbamoyl azide was
not isolated. Instead, the major compound from the reaction
mixture was the cyclic urea 4, presumably obtained by intramo-
lecular attack of the imidazole nitrogen on either the carbamoyl
azide or one of its precursors (see below). Compound 4 was
characterized by spectroscopic methods and by single-crystal
X-ray analysis.
Despite claims in the literature of carbamoyl azide instability,
we did not observe decomposition of the compounds obtained
here. Indeed, byproducts were not observed in pure carbamoyl
azides stored for several months at low temperature in a freezer
(-20 °C).
In order to assess the scope of the reaction, the possibility of
epimerization under the reaction conditions was examined for
enolizable R-protons. Thus, the carbamoyl azides of several
amino esters were submitted to GC analysis on a chiral support
in order to check their enantiomeric purity. Unfortunately,
derivatization proved necessary because the chromatogram
showed several peaks, probably due to thermal instability under
GC conditions. Therefore, the carbamoyl azides of both enan-
tiomers of valine methyl ester and phenylalanine methyl ester
were treated with dimethylamine in acetonitrile for 10 min in
order to form the dimethylurea. Each derivative showed only a
single peak when submitted to GC. In addition, co-injection of
enantiomers showed different retention times for each, proving
that racemization (>99.5 ee) did not take place during either
the synthesis of the carbamoyl azide or its derivatization.
Moreover, the absence of epimerization could also be deduced
by NMR spectra of the carbamoyl azide of isoleucine, since
only a single set of peaks corresponding to a single diastereomer
could be observed. This result is in agreement with the
previously reported non racemization in the DPPA mediated
coupling of acylamino acids with amino acids.13
the high solubility of carbon dioxide in acetonitrile also
facilitates the reaction.
A few problems had to be overcome in order to obtain optimal
yields. Initially, the reaction was performed at 0 °C by bubbling
carbon dioxide into dry acetonitrile at atmospheric pressure until
saturation was achieved. The amine, the guanidine, and the
DPPA were then added sequentially. Although the carbamoyl
azide was formed in rather good yield, in some experiments a
substantial amount of the symmetrical urea appeared as a
byproduct. The characterization of this byproduct, which has
1H NMR and 13C NMR spectra that are almost identical to the
carbamoyl azide spectra but had lost the intense IR azide band
around 2150 cm-1, was only possible by MS. This result
indicated that carefully designed experimental conditions had
to be used in order to avoid undesired side-reactions and to
improve the yield of the carbamoyl azide.
First, the need for anhydrous solvent and carbon dioxide was
established as we confirmed that the presence of small amounts
of water led to lower yields, presumably due to hydrolysis of
the carbamoyl azide.
The appearance of the urea was attributed to the reaction
between either the formed carbamoyl azide or a previous
intermediate and the free amine in the reaction medium. In an
effort to minimize the amount of urea, several modifications
were made: the initial temperature of the reaction was kept
below -35 °C, the order of addition of reagents was reversed,
the amine was added slowly, and excess sodium azide was added
to the reaction medium.
When the reaction was carried out using the same conditions
on dibutylamine, a completely different outcome was obtained.
In this case, the mixed anhydride 5 was isolated in high yield
(98%) as the only product. This compound was stable and could
be easily purified and characterized. When the reaction mixture
was heated under reflux overnight, a moderate yield (53%) of
carbamoyl azide 6 was obtained (Scheme 3). Similar results
were obtained on applying the same procedure to other
secondary amines, with the yields of carbamoyl azides being
low to moderate. These yields were not significantly improved
on using harsher reaction conditions - longer times, adding
more azide, raising the temperature, and/or adding a Lewis acid.
Another significant difference with the behavior of primary
amines is that urea was not detected when the reaction was
carried out with secondary amines.
The results of the reactions on a 1 mmol scale for several
simple primary amines are given in Table 1 (entries 1-7). Yields
correspond to isolated and purified products. The yields for the
carbamoyl azides vary between 69 and 84%, whereas the amount
of isolated urea was kept below 2% (except for 1c, 8%). Other
products were not detected in these experiments.
Entries 8-16 represent the reaction results starting from
amino ester salts of proteinogenic amino acids. Yields of isolated
carbamoyl azides are somewhat higher. The yield of urea is
also slightly higher although, in most cases, this was kept well
below 5%. The reason why yields from simple amines are to
some extent lower than from amino esters is unclear. We
attributed several low yields for small amines to losses in
isolation due to the volatility of the carbamoyl azides of low
molecular weight amines.
Compound 2e was isolated as a crystalline solid and this
allowed a single-crystal X-ray analysis to be carried out. The
(13) Shioiri, T.; Ninomiya, K.; Yamada, S. J. Am. Chem. Soc. 1972, 94,
6203.
2910 J. Org. Chem., Vol. 73, No. 7, 2008

SCHEME 3
1,5-diamine, after purification by flash chromatography (EtOAc/
hexanes from 1:9 to 1:1), the product was obtained as a white
solid (0.160 g, 69% yield): Rf 0.60 (hex/EtOAc 1:1); mp
1
115.8 °C dec; H NMR (CDCl3, 400.13 MHz) (two rotamers,
ratio 93:7) δ 5.12 (br s, 2H), 3.25 (q, 3.7H, J ) 6.5 Hz), 3.13
(q, 0.3H, J ) 6.5 Hz), 1.55 (q, 4H, J ) 7.0 Hz), 1.36 (m, 2H);
13C NMR (CDCl3, 100.62 MHz) (major rotamer) δ 156.3, 40.4,
28.7, 23.1; FTIR (KBr) νmax 3256 (br), 2178, 2142, 1686, 1668,
1539, 1258, 1225 (br) cm-1; HRMS (FAB+, mNBA matrix)
calcd for C7H13N8O2 241.1161, found 241.1154.
The different outcome of the reaction with primary and
secondary amines suggests different mechanisms. Although the
overall process is not well understood, the first step should
involve nucleophilic attack of the carbamate anion to the
phosphorus of DPPA to yield a mixed anhydride in the same
way postulated for the conversion of carboxylates to acyl azides.
In fact, this intermediate has been isolated for secondary amines
although it could not be detected for primary ones. This mixed
anhydride may evolve through different pathways with a variety
of intermediates depending on the substrate. A possible explana-
tion for the differences between primary and secondary amines
in the experimental outcome of the reaction is that the mixed
anhydride can evolve easily to an isocyanate, for primary
amines, in the basic reaction medium. This isocyanate can then
react with the nucleophiles present in the medium to yield either
the carbamoyl azide or the symmetrical urea. Since the formation
of the isocyanate is impossible for secondary amines, the
substitution on the carbonyl carbon of the mixed anhydride must
take place through a different mechanism.
Dibutylcarbamic (Diphenylphosphoric) Anhydride (5).
DPPA (0.20 mL, 0.93 mmol) was added to an ice-cooled
solution of dibutylamine (0.168 mL, 1.0 mmol) and PhTMG
(0.268 g, 1.4 mmol) in acetonitrile. CO2 was bubbled through
the solution for 30 min until saturation was achieved. The
mixture was stirred for 5 h under a carbon dioxide atmosphere,
allowing the temperature to rise to room temperature. The crude
product was dissolved in EtOAc and washed with water and
with 5% HCl. The organic phase was dried over anhydrous Na2-
SO4, filtered, and concentrated under reduced pressure. Com-
pound 5 was obtained as a yellowish oil (0.369 g, 98% yield):
1
Rf 0.33 (EtOAc/hexanes 1:5); H NMR (CDCl3) δ 7.38-7.20
(m, 10H), 3.25 (t, 2H, J ) 7.6 Hz), 3.08 (t, 2H, J ) 7.6 Hz),
1.55 (m, 2H), 1.43-1.28 (m, 4H), 1.14 (m, 2H, J ) 7.4 Hz),
0.93 (t, 3H, J ) 7.3 Hz), 0.85 (t, 3H, J ) 7.3 Hz); 13C NMR
(CDCl3) δ 150.4, 150.4, 147.8, 129.7, 125.6, 120.3, 120.3, 48.0,
47.9, 30.5, 29.5, 19.8, 19.7, 13.7, 13.6; 31P NMR (CDCl3) δ
-1.84; LRMS (EI) m/z (relative intensity), 405 (M+, 2), 349
(39), 312 (48), 263 (88), 156 (80), 99 (79), 84 (100); HRMS
(EI) calcd for C21H28NO5P (M+) 405.1705, found 405.1716.
In summary, a reliable method for the preparation of
carbamoyl azides under very mild conditions is presented. The
reaction provides very good yields of isolated products from a
broad range of primary amines. Thus, the methodology fulfils
the requirements for the preparation of carbamoyl azides either
as intermediates or as final targets in organic synthesis.
Dibutylcarbamoyl Azide (6). Dibutylamine (0.168 mL, 1.0
mmol) was submitted to the same procedure as described for
the preparation of 5. The mixture was heated at 60 °C for 14 h
under a carbon dioxide atmosphere. The reaction mixture was
dissolved in EtOAc and sequentially washed with water and
5% HCl. The organic phase was dried over anhydrous Na2SO4,
filtered, and concentrated under reduced pressure. The crude
product was purified by flash chromatography (EtOAc/hexanes
1:12) to give 6 as a colorless oil (0.095 g, 53% yield): Rf 0.66
(EtOAc/hexanes 1:9); 1H NMR (CDCl3) δ 3.28 (t, 2H, J ) 7.6
Hz), 3.16 (t, 2H, J ) 7.5 Hz), 1.67-1.46 (m, 4H), 1.34-1.26
(m, 4H), 0.93 (t, 6H, J ) 7.3 Hz); 13C NMR (CDCl3) δ 156.3,
48.7, 47.2, 30.7, 29.8, 20.1, 19.8, 13.8, 13.7; HRMS (FAB+,
mNBA matrix) calcd for C9H19N4O (M + 1)+ 199.1559, found
199.1566.
Experimental Section
General Procedure for the Synthesis of Carbamoyl Azides
from Primary Amines. A mixture of NaN3 (0.200 g, 3.08
mmol, 302 mol %) and DPPA (0.280 g, 0.22 mL, 1.00 mmol)
in dry acetonitrile (10 mL) was cooled in a dry ice/acetone bath
at -41 °C. Carbon dioxide was slowly bubbled through the
mixture until saturation was achieved. A solution of the amine
(1.06 mmol, 106 mol %) and PhTMG (0.220 g, 1.15 mmol,
112 mol %) in dry acetonitrile (15 mL) was added dropwise
(1.5-2.0 h). Once the addition was finished, the stream of CO2
was stopped, and the mixture was stirred under a carbon dioxide
atmosphere, allowing the temperature to rise to room temper-
ature overnight (14-17 h). The mixture was dissolved in EtOAc
(150 mL), and the solution was washed with water (3 × 15
mL) and with 5% aq HCl (3 × 15 mL). The resulting organic
phase was dried with anhydrous Na2SO4, filtered, and concen-
trated under reduced pressure. The residue was purified by flash
chromatography.
Acknowledgment. Financial support (projects PGIDT99-
PXI30105A and BQU2002-02807) from the Xunta de Galicia
and Ministerio de Educacio´n y Ciencia is gratefully acknowl-
edged. E.G.-E. acknowledges the Xunta de Galicia for a
fellowship.
Supporting Information Available: Experimental procedures
and characterization data. This material is available free of charge
via the Internet at http://pubs.acs.org.
Pentane-1,5-diyldicarbamoyl Azide (2a). Following the
general procedure, 2.04 mmol of DPPA and 2.30 mmol of
PhTMG were used. Thus, starting from 1.00 mmol of pentane-
JO702506V
J. Org. Chem, Vol. 73, No. 7, 2008 2911