1-Azido-1-Alkynes: Synthesis and spectroscopic characterization of azidoacetylene

Article abstract of DOI:10.1002/anie.201203626

Sleeping Beauty awakes: After 102 years of unsuccessful attempts to synthesize azidoacetylene, spectroscopic evidence for this compound has been shown. This highly explosive compound was synthesized by the treatment of ethynyliodonium salts with azide (QN₃=n-C₁₆H ₃₃Bu₃PN₃). Azidoacetylene can be trapped by a cycloaddition reaction to yield a stable triazole, otherwise cleavage to generate cyanocarbene dominates. Copyright

Full text of DOI:10.1002/anie.201203626

Angewandte
Chemie
DOI: 10.1002/anie.201203626
Ethynyl Azides
1-Azido-1-Alkynes: Synthesis and Spectroscopic Characterization of
Azidoacetylene**
Klaus Banert,* Renꢀ Arnold, Manfred Hagedorn, Philipp Thoss, and Alexander A. Auer*
Dedicated to Professor K. Barry Sharpless
In 1910, Forster and Newman had already studied the
addition of bromine to vinyl azide (2) in an aqueous solution
to prepare azidoacetylene (1a) by based-induced elimination
of two molecules of hydrogen bromide (Scheme 1). However,
owing to the explosion-like course of the addition reaction
and the high sensitivity of the product 3 to hydrolysis, 3 could
not be detected and aldehyde 5 was generated instead.[1]
A
hundred years later, it was shown that the transformation 2!
3 can be performed conveniently and quantitatively with
bromine in organic solvents at low temperature. Careful
hydrolysis of 3 led to an equilibrium mixture of a-azido
alcohol 4 and aldehyde 5,[2] but all efforts to convert 3 into the
alkyne 1a were in vain. Since the 1950s, many groups have
tried to generate 1-azido-1-alkynes, but these species have
remained elusive even until today.[3] Other attempts at
preparing these compounds were also unsuccessful or yielded
unwanted products.[4] For example, the reactions of (phenyl-
ethynyl)sodium or alkyl-1-ynyl lithiums and aromatic sulfonyl
azides exclusively led to 1,2,3-triazole derivatives instead of
ethynyl azides.[5] Furthermore, the transformation of azide 6
into the dinitriles 8, a reaction that can be performed under
a variety of conditions and in high yield, was postulated to
proceed via the short-lived alkyne 1b (Scheme 1).[6] However,
2H-azirine 7 was also predicted to be an intermediate in the
Scheme 1. Some of numerous unsuccessful attempts to prepare
ethynyl azides 1a and 1b.
transformation 6!8.[7] Recently, the highly unstable hetero-
cycle 7 was photochemically prepared from 6 in 90% yield,
characterized by its NMR data, and thermolyzed to give the
nitriles 8, nearly quantitatively.[8]
[*] Prof. Dr. K. Banert, R. Arnold, Dr. M. Hagedorn, P. Thoss
Technische Universitꢀt Chemnitz, Organische Chemie
Strasse der Nationen 62, 09111 Chemnitz (Germany)
E-mail: klaus.banert@chemie.tu-chemnitz.de
In 1983, Tanaka and Yamabe reported on the treatment of
chloride 9 with sodium azide in dimethyl sulfoxide, a reaction
that was claimed to afford in low yield (5–8%) the sulfox-
imine 17 and some other products but no dinitriles 8
(Scheme 2). The synthesis of 17 was rationalized to proceed
through the formation of short-lived azide 1b followed by loss
of dinitrogen and trapping of the nitrene 14 by the solvent.[9]
When the experiment starting with 9 was repeated recently, it
was shown that the actual product 18, resulting from the
reaction pathway via carbene 15b, had been erroneously
taken for 17.[8] Furthermore, previous ab initio studies dem-
onstrated that ethynylnitrenes such as 14 do not correspond to
a local minimum of energy and that cleavage of dinitrogen
from azide 1b should generate carbene 15b.[10,11] Not only the
carbene-interception product 18 but also the triazole 13b was
obtained when the reaction of 9 with lithium azide in dimethyl
sulfoxide was performed in the presence of cyclooctyne.[8]
This heterocyclic product is currently the most plausible
evidence for short-lived ethynyl azides. To the best of our
knowledge, any spectroscopic proof for the existence of these
azides is still missing.
Homepage: http://www.tu-chemnitz.de/chemie/org/index.html.en
Prof. Dr. A. A. Auer[+]
Max-Planck-Institut fꢁr Eisenforschung GmbH
Max-Planck-Strasse 1, 40237 Dꢁsseldorf (Germany)
E-mail: aauer@gwdg.de
[+] New address: Max-Planck-Institut fꢁr Chemische Energiekonver-
sion
Stiftstrasse 34–36, 45470 Mꢁlheim an der Ruhr (Germany)
[**] We gratefully acknowledge financial support from the Deutsche
Forschungsgemeinschaft (BA 903/12-1). We thank Dr. O. Jaurich
(Mettler Toledo) and Dr. S. Hemeltjen (TU Chemnitz) for their help
in the measurement of low-temperature IR spectra, as well as A.
Kꢀßner and Dr. A. Ihle for the realization of some experiments and
assistance with preparing the manuscript. Dedicated to Professor
K. Barry Sharpless, who inspired us to bring alkynes and azides
together. Reactions of Unsaturated Azides, Part 30; for Part 29, see:
K. Banert, C. Berndt, M. Hagedorn, H. Liu. T. Anacker, J. Friedrich, G.
Rauhut, Angew. Chem. 2012, 124, 4796–4800; Angew. Chem. Int. Ed.
2012, 51, 4718–4721.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201203626.
Angew. Chem. Int. Ed. 2012, 51, 7515 –7518
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
7515

.
Angewandte
Communications
methylidenecyclopropanes is well-known,[18]
the analogous reaction with alkynes has
seldom been investigated because the cor-
responding methylidenecyclopropenes are
highly unstable and could only be charac-
terized after interception by Diels–Alder
reaction.[19] Thus, it is not surprising that no
methylidenecyclopropenes were detected
from the reaction mixture of 10, QN3, and
alkynes. On the other hand, treatment of 10
with QN3 in the presence of ring-strained
cyclooctyne instead of open-chain acety-
lenes afforded not only a small yield of
19b (7%) but also the triazole 13b (21%),
which is the trapping product of azide 1b.
These results encouraged us to also
investigate the reaction of iodonium salts
20[20] with QN3 (Scheme 3). We assume that
the attack of azide at the terminal carbon
atom of 20 will be more rapid than the
analogous reaction of 10. Furthermore,
quantum chemical calculations,[10] in which
energy barriers for the cleavage of dinitro-
Scheme 2. Generation of phenylethynyl azide (1b). QN3 =hexadecyltributylphosphonium
azide.
Herein, we show that 1-azidoalkynes can be synthesized
more rapidly and under significantly milder reaction condi-
tions when ethynyliodonium salts instead of 1-chloroacety-
lenes are treated with azide salts. Moreover, better yields of
the trapping products are achieved, and it was possible to
detect azidoacetylene (1a) for the first time by using IR and
NMR spectroscopy. These results are strongly supported by
quantum chemical calculations.
The conversion of 9 with sodium azide in dimethyl
sulfoxide is very slow at ambient temperature and furnished
only low yields of 18 (up to 10%) and 13b (2–12%). Even
slower reactions were observed when the polar solvent
(DMSO or DMF) was diluted by the less-polar intercep-
tion-reagent cyclooctyne.[12] We assume that the nucleophilic
substitution 9!1b is a two-step addition–elimination process
with attack of azide at the carbon atom that is connected to
the chlorine atom. On the other hand, phenyliodonium salts,
such as 10,[13] are well-known to react with azide at the other
carbon atom to generate azidovinylidene intermediates such
as 11b after cleavage of iodobenzene.[14] Proof for short-lived
azidovinylidenes was presented by Stang and Kitamura, who
analyzed the subsequent reactions, which led to insertion
products.[15]
When we treated the salt 10 with a highly soluble azide
source such as QN3[16] (hexadecyltributylphosphonium azide)
in the presence of 2,3-dimethyl-2-butene, not only the
vinylidene-trapping product 12 (17% yield) but also the
known[17] cyanocyclopropane 16b (29%) were isolated
(Scheme 2). We explain the formation of 16b through
a Fritsch–Buttenberg–Wiechell-like rearrangement (FBW)
of 11b, cleavage of dinitrogen from resulting 1b, and
interception of carbene 15b. After we had repeated the
experiment with 3-hexyne instead of 2,3-dimethyl-2-butene,
we found exclusively the 3-cyanocyclopropene 19a (43%).
Whereas general trapping of vinylidenes by olefins to produce
Scheme 3. Synthesis and reactions of azidoacetylene (1a).
QN3 =hexadecyltributylphosphonium azide.
gen were studied for different ethynyl azides, predicted an
even greater stability of 1a than that of 1b, whereas
azidoacetylenes with donor groups, such as amino or ethyl-
sulfanyl, should cleave dinitrogen very easily.[21–23]
After treatment of the iodonium salt 20b with QN3 at
ꢀ408C in the presence of 2,3-dimethyl-2-butene, we isolated
the known[24] cyclopropane 16a (15% yield) and the insertion
product 21 (7%), which obviously result from carbene 15a
(Scheme 3). We could not detect the corresponding azidome-
thylidenecyclopropane, which may form by the trapping of
vinylidene 11a. Thus, we assume that the isomerization 11a!
1a is more rapid than the FBW-like rearrangement 11b!1b.
When the reaction of 20a with QN3 was performed at ꢀ40 to
ꢀ308C with cyclooctyne instead of 2,3-dimethyl-2-butene, we
exclusively isolated the triazole 13a (91%), which is a con-
sequence of the 1,3-dipolar cycloaddition of azide 1a to the
cycloalkyne. Treatment of 20a or 20b with QN3 in acetone,
dichloromethane, or chloroform (ꢀ408C/30–120 min) in the
7516
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 7515 –7518

Angewandte
Chemie
Table 1: Selected experimental and calculated spectroscopic data of azidoacetylene (1a) and [15N3]-1a.[a,b,c]
ꢀ1
ꢀ
ꢁ
~
1a
IR (CHCl3,
ꢀ208C):[c]
n=3312 (m, C H), 2188 (s, C C), 2148 (w), 2086 (s, N3), 1251 cm (w, N3)
1H NMR
d=3.38 (s) [d=3.7]
([D6]acetone):
2
13C NMR
d=55.81 (d, 1J=265.5 Hz, CH), 72.25 (d, J=58.3 Hz, CN3) [d=55.5 (1J=259 Hz), 76.5 (2J=59 Hz)]
([D6]acetone):
1H NMR (CD2Cl2): d=2.55 (s) [d=3.0]
13C NMR (CD2Cl2): d=54.52 (d, 1J=262.8 Hz, CH) 71.94 (d, 2J=57.5 Hz, CN3) [d=54.0 (1J=259 Hz), 75.0 (2J=59 Hz)]
[15N3]- 1H NMR (CD2Cl2): d=2.57 (dd, 4JHN =1.7 Hz, 3JHN =0.9 Hz)
1a
1
2
3
13C NMR (CD2Cl2): d=54.31 (ddd, JCH =264.9 Hz, JCN =5.7 Hz, JCN =2.9 Hz, CH), 71.69 (dd, 2JCH =58.8 Hz, 1JCN =15.4 Hz, CN3)
[1JCN =ꢀ17 Hz, 2JCNa =5.6 Hz, 2JCNb =2.6 Hz]
15N NMR (CD2Cl2): d=ꢀ314.6 (br d, JNN =15.5 Hz, Na), ꢀ151.8 (d, 1JNN =7.4 Hz, Ng), ꢀ139.9 (ddd,1JNaNb =15.5 Hz, 1JNbNg =7.4 Hz,
1
4JNH =1.7 Hz, Nb) [d=317.8 (1JNN =ꢀ15.5 Hz, 3JNH =ꢀ0.8 Hz, Na), ꢀ151.8 reference value (1JNN =ꢀ6 Hz, Ng), ꢀ133.8
(1JNaNb =ꢀ15.5 Hz, 1JNbNg =ꢀ6 Hz, 4JNH =1.9 Hz, Nb)]
[a] 1H and 13C NMR spectra were recorded at ꢀ408C and 400 and 100.6 MHz, respectively. 15N NMR spectra were recorded at ꢀ608C and 40.5 MHz,
and referenced to nitromethane (d=0). The sign of experimental J values was not determined. [b] All NMR data based on quantum chemical
calculations are given in square brackets. In the case of the 1H NMR and 13C NMR spectra, the reference of d values is determined with the help of
experimental and calculated chemical shifts of acetylene (see the Supporting Information). Calculated 15N NMR d values were referenced to the
experimental chemical shift of Ng. All results have been obtained from all-electron calculations and are based on geometries that have been obtained at
the CCSD(T)/cc-pVTZ level of theory. NMR chemical shifts have been calculated at the CCSD(T)/pz3d2f level of theory including zero-point vibrational
effects calculated at the MP2/cc-pVTZ (geometry and anharmonic force field) CCSD(T)/qz2p (chemical shifts) level of theory. Spin-spin coupling
constants have been calculated at the CCSD/unc-cc-pVTZ-J level of theory including FC, SD, PSO and DSO contributions (FC and SD terms have been
calculated using an unrelaxed UHF reference). [c] For IR data based on analytical CCSD(T)/cc-pVTZ normal mode calculations, see the Supporting
Information.
absence of any trapping reagent led to solutions of 1a after
careful[25] recondensation of the reaction mixtures under
reduced pressure (yield approximately 40%, as determined
methane was warmed to ambient temperature in the presence
of 2,3-dimethyl-2-butene, the expected compounds 16a
(44%) and 21 (27%) were formed. The same products were
also found with 35% and 15% yield, respectively, after
irradiation of a solution of 1a in dichloromethane for 3 hours
at ꢀ608C.
by 1H NMR spectroscopy). 15N-labeled azidoacetylene
([15N3]-1a) was also synthesized using Q15N3
and 20a.
[16,26]
We characterized 1a using IR, H NMR, 13C NMR, and
15N NMR spectroscopy, which provided conclusive proof of
the structure (Table 1). These data, especially the d and
J values from the NMR spectra were compared with the
corresponding values resulting from quantum chemical cal-
culations at the coupled-cluster level of theory. Experimental
and calculated data for acetylene were included for the
referencing of the chemical shifts in the case of calculated
1H NMR and 13C NMR spectra. Generally, good agreement
between the measured and the theoretical data was found.
1
In summary, we have characterized the first ethynyl azide
by IR, 1H NMR, 13C NMR, and 15N NMR spectroscopy. After
102 years of unsuccessful attempts, not only have the azide
adducts, for example, the cycloadducts 13 and the cyanocar-
bene-trapping products 16, 18, 19, and 21, been observed but
also the parent azide itself has been characterized. Currently,
we are investigating whether the properties of azidoacetylene
(1a) as an electron-rich alkyne can be utilized for electro-
ꢁ
philic addition or cycloaddition reactions at the C C bond.
ꢀ
The upfield shift for the signal corresponding to the C H
Received: May 10, 2012
Published online: July 6, 2012
carbon (d ca. 55) indicates that the azido group acts as
a p donor, through which 1a becomes an electron-rich alkyne.
1
2
The coupling constants J(13C,1H) and J(13C,1H) of 1a are
significantly larger than those of acetylene. This is typical for
terminal alkynes with oxygen or nitrogen donor substitu-
ents.[27] The IR spectrum of 1a shows expected bands
Keywords: carbenes · ethynyl azides · NMR spectroscopy ·
.
quantum chemistry · reaction mechanisms
ꢀ
ꢁ
corresponding to C H, C C, asymmetric N3, and symmetric
ꢁ
N3 stretching modes. However, the high extinction of the C C
[1] M. O. Forster, S. H. Newman, J. Chem. Soc. 1910, 97, 2570 – 2579.
[2] K. Banert, C. Berndt, S. Firdous, M. Hagedorn, Y.-H. Joo, T.
Rꢀffer, H. Lang, Angew. Chem. 2010, 122, 10404 – 10407; Angew.
Chem. Int. Ed. 2010, 49, 10206 – 10209.
stretching vibration is as strong as the azide band; this finding
is remarkable and possibly results from coupling of vibration
modes.
[3] Reviews on ethynyl azides: a) K. Banert in Organic Azides—
Syntheses and Applications (Eds.: S. Brꢁse, K. Banert), Wiley,
Chichester, 2010, pp. 115 – 166; b) K. Banert in Houben-Weyl,
Science of Synthesis, 5th ed., Vol. 24 (Ed.: A. de Meijere),
Thieme, Stuttgart, 2006, pp. 1061 – 1072; c) K. I. Booker-Milburn
in Comprehensive Organic Functional Group Tranformations,
Azidoacetylene (1a) decomposed with a half-life period
of approximately 17 hours in dichloromethane at ꢀ308C.
Nevertheless, it can be reacted intermolecularly, for example,
with cyclooctyne to furnish 13a (35% yield based on
precursor 20a). When a cold solution of 1a in dichloro-
Angew. Chem. Int. Ed. 2012, 51, 7515 –7518
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7517

.
Angewandte
Communications
Vol. 2 (Eds.: A. R. Katritzky, O. Meth-Cohn, C. W. Rees),
Pergamon, Oxford, 1995, pp. 1054 – 1055; d) G. Himbert in
Houben-Weyl, 4th ed., Vol. E15 (Eds.: H. Kropf, E. Schaumann),
Thieme, Stuttgart, 1993, p. 3455.
[18] a) D. P. Fox, P. J. Stang, Y. Apeloig, M. Karni, J. Am. Chem. Soc.
1986, 108, 750 – 756; b) M. Ochiai, T. Sueda, K. Uemura, Y.
Masaki, J. Org. Chem. 1995, 60, 2624 – 2626; c) H.-Y. Lee, Y.-H.
Lee, Synlett 2001, 1656 – 1658.
[4] a) J. H. Boyer, C. H. Mack, N. Goebel, L. R. Morgan, Jr., J. Org.
Chem. 1958, 23, 1051 – 1053; b) J. H. Boyer, R. Selvarajan,
Tetrahedron Lett. 1969, 10, 47 – 50; c) V. A. Garibina, A. A.
Leonov, A. V. Dogadina, B. I. Ionin, A. A. Petrov, J. Gen. Chem.
USSR (Engl. Transl.) 1985, 55, 1771 – 1781.
[5] a) R. Helwig, M. Hanack, Chem. Ber. 1985, 118, 1008 – 1021;
b) E. Robson, J. M. Tedder, B. Webster, J. Chem. Soc. 1963,
1863 – 1865; c) M. E. Meza-AviÇa, M. K. Patel, C. B. Lee, T. J.
Dietz, M. P. Croatt, Org. Lett. 2011, 13, 2984 – 2987.
[6] J. H. Boyer, R. Selvarajan, J. Am. Chem. Soc. 1969, 91, 6122 –
6126.
[19] P. J. Stang, M. G. Mangum, J. Am. Chem. Soc. 1975, 97, 3856.
[20] a) T. Kitamura, M. H. Morshed, S. Tsukada, Y. Miyazaki, N.
Iguchi, D. Inoue, J. Org. Chem. 2011, 76, 8117 – 8120; b) M.
Ochiai, T. Ito, Y. Takaoka, Y. Masaki, M. Kunishima, S. Tani, Y.
Nagao, J. Chem. Soc. Chem. Commun. 1990, 118 – 119; c) P. J.
Stang, A. M. Arif, C. M. Crittell, Angew. Chem. 1990, 102, 307 –
308; Angew. Chem. Int. Ed. Engl. 1990, 29, 287 – 288.
[21] In the light of this ab inito study, it seems unrealistic to utilize 1-
azido-2-dimethylaminoethyne as a hypergolic fuel for hypergolic
bipropellant mixtures.[22] Moreover, the report[23] on the syn-
thesis and isolation of (ethylsulfanyl)ethynyl azide is highly
questionable. In this case, it has also been shown that the
presented NMR data are incompatible with the structure,[10] and
several attempts to reproduce the generation of the latter azide
were unsuccessful.[3a,8]
[7] A. Hassner, R. J. Isbister, J. Am. Chem. Soc. 1969, 91, 6126 –
6128.
[8] K. Banert, M. Hagedorn, J. Wutke, P. Ecorchard, D. Schaarsch-
midt, H. Lang, Chem. Commun. 2010, 46, 4058 – 4060.
[9] a) R. Tanaka, K. Yamabe, J. Chem. Soc. Chem. Commun. 1983,
329 – 330; b) K. Yamabe, R. Tanaka, Sasebo Kogyo Koto Senmon
Gakko Kenkyu Hokoku 1985, 22, 119 – 123 [Chem. Abstr. 1987,
106, 49680t].
[10] E. Prochnow, A. A. Auer, K. Banert, J. Phys. Chem. A 2007, 111,
9945 – 9951.
[11] For other quantum chemical calculations to characterize ethynyl
azides, see: a) P. Politzer, P. Lane, P. Sjoberg, M. E. Grice, H.
Shechter, Struct. Chem. 1995, 6, 217 – 223; b) A. V. Golovin, V. V.
Takhistov, J. Mol. Struct. 2004, 701, 57 – 91.
[12] Synthesis of cyclooctyne: L. F. Tietze, T. Eicher, Reaktionen und
Synthesen im Organisch-Chemischen Praktikum und Forschung-
slaboratorium, 2nd ed., Thieme, Stuttgart, 1991, p. 40.
[13] A. J. Margida, G. F. Koser, J. Org. Chem. 1984, 49, 4703 – 4706.
[14] For reviews, see: a) P. J. Stang, V. V. Zhdankin, Chem. Rev. 1996,
96, 1123 – 1178; b) V. V. Zhdankin, P. J. Stang, Tetrahedron 1998,
54, 10927 – 10966; c) V. V. Zhdankin, P. J. Stang, Chem. Rev.
2008, 108, 5299 – 5358; d) P. J. Stang, J. Org. Chem. 2003, 68,
2997 – 3008; e) W. Kirmse, Angew. Chem. 1997, 109, 1212 – 1218;
Angew. Chem. Int. Ed. Engl. 1997, 36, 1164 – 1170.
[22] D. Sengupta (CFD Research Corporation, AL), US 2008/
0202655A1, 2008 [Chem. Abstr. 2008, 149, 310772].
[23] a) S. G. Dꢂyachkova, E. A. Nikitina, N. K. Gusarova, M. L.
Alꢂpert, B. A. Trofimox, Russ. Chem. Bull. 2001, 50, 751 – 752;
b) S. G. Dꢂyachkova, E. A. Nikitina, N. K. Gusarova, A. I.
Albanov, B. A. Trofimov, Russ. J. Gen. Chem. (Engl. Transl.)
2003, 73, 782 – 785.
[24] a) J. Martel, J. Tessier, J.-P. Demoute (Roussel Uclaf), FR
2479192, A1 19811002, 1981 [Chem. Abstr. 1982, 96, 85115];
b) R. A. Moss, F. Zheng, K. Krogh-Jespersen, Org. Lett. 2001, 3,
1439 – 1442.
[25] Caution! The azide 1a is a highly volatile and explosive
compound that can be safely used in cold solution. However,
recondensation under reduced pressure can lead to an explosion
(protective shield). The explosive properties of 1a are similar to
those of diazomethane. For potential hazards in handling
hydrazoic acid and organic azides, see: T. Keicher, S. Lçbbecke
in Organic Azides: Syntheses and Applications (Eds.: S. Brꢁse, K.
Banert), Wiley, Chichester, 2010, pp. 3 – 27.
[26] K. Banert, J. Lehmann, H. Quast, G. Meichsner, D. Regnat, B.
Seiferling, J. Chem. Soc. Perkin Trans. 2 2002, 126 – 134.
[27] A. Sebald, B. Wrackmeyer, Spectrochim. Acta Part A 1981, 37,
365 – 368.
[15] T. Kitamura, P. J. Stang, Tetrahedron Lett. 1988, 29, 1887 – 1890.
[16] K. Banert, Synthesis 2007, 3431 – 3446.
[17] R. A. Moss, G. Kmiecik-Ławrynowicz, D. P. Cox, Synth.
Commun. 1984, 14, 21 – 25.
7518
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 7515 –7518