Attempted isolation and characterization of diazirinone (N2CO)

Article abstract of DOI:10.1021/jo9026462

Chemical Equation Presented Stimulated and intrigued by the report of the synthesis of diazirinone (1), a metastable adduct of N₂ and CO, we carried out further experimental and theoretical studies aimed at the detailed spectroscopic characterization of this species. Our attempts to generate and detect diazirinone (1) in either the condensed phase (using matrix isolation spectroscopy) or in the gas phase (using millimeter-wave rotational spectroscopy), however, have been unsuccessful. Trapping the volatile products produced from the reaction of 3-chloro-3-(p-nitrophenoxy)diazirine (5) with tetrabutylammonium fluoride (TBAF) under matrix-isolation conditions affords chlorofluorodiazirine (8) and carbon monoxide but fails to provide evidence for diazirinone (1). Moreover, sophisticated ab initio calculations of the structure and fundamental vibrational frequencies of diazirinone (1) produce an estimate for the fundamental band origin of the C=O stretch (2046 cm^-1) that is ca. 100 cm^-1: lower in frequency than the experimental value previously attributed to this band. This discrepancy lies well outside any expected solvent shift or calculation error at this level of theory. In an effort to reconcile our findings with the earlier reports concerning diazirinone (1), we reconsidered the infrared spectral evidence upon which the original claim of diazirinone synthesis was based. New experiments demonstrate that these spectra may be explained and reproduced with a combination of solution-phase and gas-phase absorptions of CO, without recourse to invoke diazirinone (1).

Full text of DOI:10.1021/jo9026462

pubs.acs.org/joc
Attempted Isolation and Characterization of Diazirinone (N2CO)
Christopher J. Shaffer,† Brian J. Esselman,† Robert J. McMahon,*,† John F. Stanton,*,‡
and R. Claude Woods*,†
†Department of Chemistry, University of Wisconsin, 1101 University Avenue, Madison, Wisconsin
53706-1322, and ‡Institute for Theoretical Chemistry, Department of Chemistry and Biochemistry,
University of Texas, 1 University Station A5300, Austin, Texas 78712-0165
mcmahon@chem.wisc.edu; jfstanton@mail.utexas.edu; rcwoods@wisc.edu
Received December 14, 2009
Stimulated and intrigued by the report of the synthesis of diazirinone (1), a metastable adduct of N2
and CO, we carried out further experimental and theoretical studies aimed at the detailed spectro-
scopic characterization of this species. Our attempts to generate and detect diazirinone (1) in either
the condensed phase (using matrix isolation spectroscopy) or in the gas phase (using millimeter-wave
rotational spectroscopy), however, have been unsuccessful. Trapping the volatile products produced
from the reaction of 3-chloro-3-(p-nitrophenoxy)diazirine (5) with tetrabutylammonium fluoride
(TBAF) under matrix-isolation conditions affords chlorofluorodiazirine (8) and carbon monoxide
but fails to provide evidence for diazirinone (1). Moreover, sophisticated ab initio calculations of the
structure and fundamental vibrational frequencies of diazirinone (1) produce an estimate for the
fundamental band origin of the CdO stretch (2046 cm-1) that is ca. 100 cm-1 lower in frequency than
the experimental value previously attributed to this band. This discrepancy lies well outside any
expected solvent shift or calculation error at this level of theory. In an effort to reconcile our findings
with the earlier reports concerning diazirinone (1), we reconsidered the infrared spectral evidence
upon which the original claim of diazirinone synthesis was based. New experiments demonstrate that
these spectra may be explained and reproduced with a combination of solution-phase and gas-phase
absorptions of CO, without recourse to invoke diazirinone (1).
Introduction
7-norbornadienone,4 and tetranitrogen,5 provide opportu-
nities to probe the limits of chemical (in)stability. Each of
these molecules may undergo a highly exothermic decompo-
sition to its respective stable constituents, presenting a
potentially severe synthetic obstacle to isolation and char-
acterization. Comparison of diazirinone (1) and cyclopro-
penone provides insight concerning the nature of weakly
aromatic systems.6,7 The anticipated polarity of diazirinone
(1) makes it an interesting target to investigate via rotational
Diazirinone (1), the C2v conjugate of molecular nitrogen
and carbon monoxide, is an intriguing species, inasmuch
as it is a mixed dimer of the two most strongly bonded
diatomic molecules. Diazirinone (1), along with similar high-
energy, metastable compounds1 such as ethylenedione,2,3
(1) Hoffmann, R.; Hopf, H. Angew. Chem., Int. Ed. 2008, 47, 4474–4481.
€
(2) Schroder, D.; Heinemann, C.; Schwarz, H.; Harvey, J. N.; Dua, S.;
Blanksby, S. J.; Bowie, J. H. Chem.;Eur. J. 1998, 4, 2550–2557.
(3) Berson, J. A.; Birney, D. M.; Dailey, W. P.; Liebman, J. F. Mol.
Struct. Energ. 1988, 6, 391–441. , 445-446.
(4) LeBlanc, B. F.; Sheridan, R. S. J. Am. Chem. Soc. 1985, 107, 4554–
4556.
(5) Cacace, F.; de Petris, G.; Troiani, A. Science 2002, 295, 480–481.
(6) Korkin, A. A.; Balkova, A.; Bartlett, R. J.; Boyd, R. J.; Schleyer,
P. v. R. J. Phys. Chem. 1996, 100, 5702–5714.
(7) Skancke, A.; Liebman, J. F. J. Org. Chem. 1999, 64, 6361–6365.
DOI: 10.1021/jo9026462
r
Published on Web 02/12/2010
J. Org. Chem. 2010, 75, 1815–1821 1815
2010 American Chemical Society

Shaffer et al.
JOCFeatured Article
SCHEME 1a
SCHEME 2
aComputed energies (kcal/mol) of N2CO isomers,6,8,9 relative to dia-
zirinone (1).
spectroscopy, which could provide insights into the structure
and bonding of this peculiar molecule. The laboratory rota-
tional spectrum of 1 would also provide the basis to search
for diazirinone as a possible component of the interstellar
medium via radioastronomy. These considerations motiva-
ted us to initiate the studies described herein.
matrix-isolated N2CO isomers (NCNO, CNNO, NCON).12
De Petris et al. generated an N2CO isomer in a neutrali-
zation-reionization mass spectrometry (NRMS) experi-
ment.9 Careful analysis suggests that the isomer generated
is triplet NNCO (33), and persuasive arguments rule out both
diazirinone (1) and nitrosyl cyanide (2). Concurrently,
diazirinone (1) was reported as an unanticipated product
obtained during the preparation of chlorofluorodiazirine.10
When 3-chloro-3-(p-nitrophenoxy)diazirine (5) was treated
with molten tetrabutylammonium fluoride (TBAF), the
expected products, chlorofluorodiazirine (8) and p-nitrophe-
nol, were obtained only as minor products, while p-fluoro-
nitrobenzene (7) was obtained as a major product, along
with a transient species exhibiting an infrared absorption at
2150 cm-1 (Scheme 2). As the reaction mixture warmed
toward room temperature, the absorption at 2150 cm-1
decayed over 5-10 min, while a near-simultaneous increase
of CO absorptions at 2116 and 2169 cm-1 was observed. The
mechanism shown in Scheme 2 was used to rationalize the
experimental observations, and it was concluded that the
transient species produced was diazirinone (1). Subsequent
studies to prepare other dihalodiazirines consistently
produced p-fluoronitrobenzene (7), from which it was
assumed that diazirinone was also produced. Diazirinone
Diazirinone (1) is the lowest energy isomer on the
N2CO potential energy surface, other than CO þ N2
(Scheme 1).3,6,8-11 Nitrosyl cyanide (2), which is an isolable
species, lies ca. 15 kcal/mol higher in energy.8,11 The open-
chain isomer NNCO (3) is not bound on the singlet PES but
is a low-lying minimum on the triplet PES. The enthalpy
barrier for the dissociation of diazirinone (1) to N2 and CO is
ca. 27 kcal/mol, while the exothermicity of the reaction is ca.
100 kcal/mol.8,11 The barrier is consistent with the rate of
thermal decomposition attributed to 1.10 Korkin et al. noted
an interesting difference between the N2CO PES and the
isoelectronic PES of N4.6 In the case of N2CO, the tetra-
hedron-like isomer 4 is much higher in energy (115 kcal/mol)
than diazirinone (1), which was rationalized in terms of the
inherent formal charges. In the case of N4, the tetrahedral
structure is 10 kcal/mol lower in energy than the distorted
structure that is analogous to diazirinone.
In terms of experimental studies, Maier and co-
workers reported the photochemistry and spectroscopy of
(8) Korkin, A. A.; Schleyer, P. v. R.; Boyd, R. J. Chem. Phys. Lett. 1994,
227, 312–320.
(9) de Petris, G.; Cacace, F.; Cipollini, R.; Cartoni, A.; Rosi, M.; Troiani,
A. Angew. Chem., Int. Ed. 2005, 44, 462–465.
(10) Moss, R. A.; Chu, G.; Sauers, R. R. J. Am. Chem. Soc. 2005, 127,
2408–2409.
(11) Zhu, R. S.; Lin, M. C. J. Phys. Chem. A 2007, 111, 6766–6771.
(12) Maier, G.; Reisenauer, H. P.; Eckwert, J.; Naumann, M.; De Marco,
M. Angew. Chem., Int. Ed. Engl. 1997, 36, 1707–1709.
1816 J. Org. Chem. Vol. 75, No. 6, 2010

Shaffer et al.
JOCFeatured Article
FIGURE 1. Infrared spectrum (Ar, 10 K) of the volatile components from the reaction of p-nitrophenoxychlorodiazirine (5) and Bu4NþF- in
acetonitrile. Top: spectrum of volatile products that passed directly through a gas trap at 77 K without condensing. Bottom: spectrum obtained
after warming the trap to 263 K. All peaks can be assigned to either CO (2140 cm-1) or acetonitrile. Both plots are set to the same scale but are
offset by 2.5 absorbance units to allow for comparison.
was not isolated as a pure compound in any of these
studies.10,13-19
Utilizing a procedure similar to that described in the litera-
ture, 3-chloro-3-(p-nitrophenoxy)diazirine (5) was dissolved
in acetonitrile and added to molten Bu4NþF- at -10 °C.10
The reaction mixture bubbled steadily for 30 min before
being closed off to obtain the infrared spectrum of the
matrix. The spectrum exhibited a single absorption feature
at 2140 cm-1, which is consistent with the literature value for
matrix isolated CO (Figure 1).20,21 When the gas trap was
allowed to warm from 77 to 263 K, neither increased
absorption by CO nor a new absorption attributable to
diazirinone was observed (Figure 1). These observations
suggest that diazirinone had not condensed in the gas trap
at 77 K, as it neither deposited on the matrix nor decomposed
to produce more CO. Since we did not observe a peak at
2150 cm-1, we were unable to confirm that diazirinone was a
persistent species or an observable intermediate.
Examination of the isolable reaction products, obtained
upon treatment of 3-chloro-3-(p-nitrophenoxy)diazirine (5)
with Bu4NþF-, by 1H and 19F NMR spectroscopy confirms
the formation of p-nitrophenol, p-fluoronitrobenzene (7),
and 3-fluoro-3-(p-nitrophenoxy)diazirine (9). The only
points of uncertainty, therefore, appear to involve (i) the
detection and characterization of diazirinone (1) and (ii) the
suitability of utilizing p-fluoronitrobenzene (7) as a surrogate
for inferring the formation of diazirinone (1) (Scheme 2).
Gas- and Solution-Phase Infrared Spectroscopy of Carbon
Monoxide. In view of our inability to isolate diazirinone or
trap it under matrix isolation conditions, as well as the signi-
ficant discrepancy between our predicted band origin of the
CdO fundamental stretching vibration of diazirinone
(2046 cm-1) (see below) and the IR peak previously attrib-
uted to this species (2150 cm-1),10 we have confronted the
Results and Discussion
Attempted Synthesis and Isolation of Diazirinone (1). Our
primary goal, to obtain the gas-phase millimeter-wave rota-
tional spectrum of diazirinone (1), could be straightforwardly
pursued bypreparing a sampleof1, either as a neat liquid or as
a solution in a high-boiling solvent, from which 1 could be
distilled. The reported rapid thermal decomposition of dia-
zirinone (ca. 10 min at -10 °C) suggested that isolation of
diazirinone (1) would likely require vacuum transfer of the
metastable species to a cold trap. Several attempts to distill
diazirinone (1) from the reaction mixture of 3-chloro-3-(p-
nitrophenoxy)diazirine (5) and tetrabutylammonium fluo-
ride, using a series of low volatility solvents (acetonitrile,
diglyme, and butyl diglyme), into an LN2-cooled trap, failed
to yield any condensed product. Our choice of solvent was
influenced by a number of requirements: The solvent needed
to be fluid at temperatures between -20 and þ25 °C, to have a
low vapor pressure at those temperatures (preventing distilla-
tion into the matrix or gas trap), and to solubilize Bu4NþF-.
In view of the failures to capture diazirinone in an LN2
trap, we attempted the deposition of diazirinone into an
argon matrix at 25 K, transferred from the same reaction
mixtures. A long-path gas trap (20 cm), cooled to 77 K, was
placed between the reaction solution and the matrix window.
(13) Chu, G.; Moss, R. A.; Sauers, R. R. J. Am. Chem. Soc. 2005, 127,
14206–14207.
(14) Moss, R. A. Acc. Chem. Res. 2006, 39, 267–272.
(15) Moss, R. A.; Tian, J.; Sauers, R. R.; Skalit, C.; Krogh-Jespersen, K.
Org. Lett. 2007, 9, 4053–4056.
(16) Moss, R. A.; Tian, J.; Sauers, R. R.; Ess, D. H.; Houk, K. N.;
Krogh-Jespersen, K. J. Am. Chem. Soc. 2007, 129, 5167–5174.
(17) Moss, R. A.; Tian, J.; Chu, G.; Sauers, R. R.; Krogh-Jespersen, K.
Pure Appl. Chem. 2007, 79, 993–1001.
(18) Moss, R. A.; Wang, L.; Krogh-Jespersen, K. J. Am. Chem. Soc. 2009,
131, 2128–2130.
(19) Moss, R. A. J. Phys. Org. Chem. 2009, early view.
(20) Ball, D. W. A Bibliography of Matrix Isolation Spectroscopy:
1954-1985; Rice University Press: Houston, 1988.
(21) Ochsner, D. W.; Ball, D. W.; Kafafi, Z. H. A Bibliography of
Matrix Isolation Spectroscopy: 1985-1997; Naval Research Laboratory:
Washington, DC, 1998.
J. Org. Chem. Vol. 75, No. 6, 2010 1817

Shaffer et al.
JOCFeatured Article
FIGURE 2. Infrared spectrum of carbon monoxide dissolved in acetonitrile (1 cm quartz cell). Top: νCO observed at 2140 cm-1 (4 min after
saturation; warming from -35 °C). The progressive five lower plots, taken at 2 min intervals, show the decrease in intensity of solution-phase
CO (2140 cm-1) as CO degasses from solution upon warming.
need to reconsider the assignment of spectral features in the
reported IR spectrum. We suggest that the two features (2116
and 2169 cm-1) in Figure 1 of ref 10, previously attributed to
solution-phase CO are, in fact, the maximum intensity
positions of the P and R branches of gas-phase CO, respec-
tively. Furthermore, we suggest that the peak at 2150 cm-1
that was assigned to the diazirinone (1) CdO stretch is
actually the absorption of CO in a condensed phase. The
presence of gas-phase carbon monoxide is strongly indicated
by the presence of partially resolved spectroscopic features in
the range 2060-2130 cm-1, which correspond to individual
rotational lines. Such extensive rotational structure is obser-
ved only in the gas phase, while in the condensed phase (either
in solution or in a matrix), the CO absorption collapses due to
the stronger intermolecular interactions.
(0.005-0.008 M at 1 atm CO),25-27 we attempted to saturate
a solution of acetonitrile with CO at -35 °C for 30 min. To
further compensate for low solubility, we employed a long
path length (1.00 cm) quartz cuvette. The IR spectral window
that is accessible with this cell is only 2080-2200 cm-1, due
to the high optical density resulting from the combined
absorptions of the cuvette and acetonitrile. When the cuvette
was placed in the spectrometer chamber, a structureless
absorption centered at 2140 cm-1 was observed (Figure 2;
top). Mimicking the disappearance of the 2150 cm-1 peak
previously assigned to the decomposition of diazirinone, the
intensity of this peak decreased to near zero over 10 min as
the CO degassed from the gradually warming solution
(Figure 2). In the spectrum previously attributed to diazir-
inone (1), the optical density of the sample was very high,
corresponding to a transmittance value of 1.36-1.78% in the
region of the CdO stretching absorption.10 Although it is
possible to measure intense absorption features under these
conditions, the resolution at this saturation level starts to
degrade and peaks begin to broaden, creating uncertainty
To confirm this interpretation, we dissolved CO in ace-
tonitrile and obtained an infrared spectrum. Literature
values for carbon monoxide in acetonitrile place the funda-
mental frequency at 2140 cm-1, but these reports are vague
on solution conditions.22-24 In view of the low solubility
(22) Szanyi, J.; Paffett, M. T. J. Chem. Soc., Faraday Trans. 1996, 92,
5165–5175.
(23) Christensen, P. A.; Hamnett, A.; Muir, A. V. G.; Freeman, N. A.
J. Electroanal. Chem. 1990, 288, 197–215.
(25) Lopez-Castillo, Z. K.; Aki, S. N. V. K.; Stadtherr, M. A.; Brennecke,
J. F. Ind. Eng. Chem. Res. 2006, 45, 5351–5360.
(26) Fujita, E.; Creutz, C.; Sutin, N.; Szalda, D. J. J. Am. Chem. Soc.
1991, 113, 343–353.
(24) Anderson, M. R.; Blackwood, D.; Pons, S. J. Electroanal. Chem.
1988, 256, 387–395.
(27) Szalda, D. J.; Fujita, E.; Creutz, C. Inorg. Chem. 1989, 28, 1446–
1450.
1818 J. Org. Chem. Vol. 75, No. 6, 2010

Shaffer et al.
JOCFeatured Article
FIGURE 3. Infrared spectrum of gas phase CO exhibiting features at 2118 and 2172 cm-1
.
about the frequency measurement and subsequent assign-
ment. Thus, we interpret the absorption attributed to dia-
zirinone (1) at 2150 cm-1 as that of CO broadened by
saturation; this feature is commensurate with the absorption
TABLE 1. Computed Fundamental Vibrational Frequencies for Dia-
zirinone (1)a
state
b
symmetry
frequency
intensity
mode description
ν1
a1
a1
a1
b1
b2
b2
a1
2046
1325
903
565
961
316.7
0.2
5.2
28.7
11.3
11.9
120.7
CO stretch
NN stretch
CN symmetric stretch
out-of-plane bend
OCN scissor
of CO in acetonitrile at 2140 cm-1
.
ν2
ν3
ν4
ν5
ν6
In order to confirm our assignment of the 2116 and 2169
cm-1 features as gas phase CO absorptions, we collected an
independent IR spectrum of gas-phase CO at 2-cm-1 resolu-
tion. The spectrum exhibits features at 2118 and 2172 cm-1
(Figure 3). At this resolution, the appearance of the spectrum
is extremely similar to the published one,10 in that partially
resolved peaks were observed in the P-branch region but not
in the R-branch region. This behavior is due to the closer
spacing of the R-branch rotational lines relative to those in
the P-branch. When the spectral resolution was increased to
0.060 cm-1, all of the rotational lines in both branches were
fully resolved.
Computational Studies. Our ab initio calculations estimate
the fundamental CdO stretching frequency, ν1, of diaziri-
none (1) to be 2046 cm-1 (Table 1). This value derives from a
high-level calculation that incorporates a sophisticated treat-
ment of electron correlation, a large basis set, and treatment
of vibrational anharmonicity using second-order vibrational
perturbation theory. The discrepancy between computed
(2046 cm-1) and experimental (2150 cm-1) values is trou-
bling in both magnitude and sign. A calculation of this type,
for a molecule of this size, is expected to reproduce the
experimental value to within 25 cm-1 (see below). Moreover,
computed values are typically higher in frequency than
experimental values. Thus, the present calculations cast
considerable doubt on the assignment of the experimentally
observed IR band at 2150 cm-1 to diazirinone (1).
529
1860
NCN scissor
b
2ν5
aCCSD(T)/ANO2; frequency (cm-1). bStrong Fermi resonance; see
text.
associated cubic force constant φ155 (-307 cm-1) is quite
large. (The resonance has been treated here in the usual way,
in which the associated fundamental and overtone levels are
computed by diagonalization of a 2 ꢀ 2 effective Hamil-
tonian.) An accuracy of 25 cm-1 is typically associated with
vibrational frequencies calculated at this level of theory,28-31
but the involvement of the Fermi resonance introduces
somewhat greater uncertainty. By conservative estimate,
the calculated value of 2046 cm-1 might be in error by as
much as 50 cm-1 32,33
The computed dipole moment of
.
diazirinone (1) is 1.05 D, a value that is more than sufficient
to permit the measurement of its rotational spectrum. The
rotational constants A0, B0, and C0 predicted from the VPT2
model are 41360, 8400, and 6970 MHz, where all values are
rounded to the nearest 10 MHz.
Analysis. In light of the evidence described above, we
surmise that carbon monoxide dissolved in a condensed
phase accounts for the broad peak at 2150 cm-1 previously
(28) Jochnowitz, E. B.; Zhang, X.; Nimlos, M. R.; Varner, M. E.;
Stanton, J. F.; Ellison, G. B. J. Phys. Chem. A 2005, 109, 3812–3821.
(29) Zhang, X.; Nimlos, M. R.; Ellison, G. B.; Varner, M. E.; Stanton,
J. F. J. Chem. Phys. 2006, 124, 084305.
(30) Zhang, X.; Nimlos, M. R.; Ellison, G. B.; Varner, M. E.; Stanton,
J. F. J. Chem. Phys. 2007, 126, 174308.
(31) Stanton, J. F.; Flowers, B. A.; Matthews, D. A.; Ware, A. F.; Ellison,
G. B. J. Mol. Spectrosc. 2008, 251, 384–393.
(32) Konen, I. M.; Li, E. X. J.; Lester, M. I.; Vazquez, J.; Stanton, J. F.
J. Chem. Phys. 2006, 125, 074310.
(33) Vazquez, J.; Stanton, J. F. Mol. Phys. 2007, 105, 101–109.
In terms of the detailed vibrational spectroscopy of diazir-
inone (1), there is a strong Fermi resonance between the
fundamental CdO stretching frequency, ν1, and the over-
tone of ν5, the latter being best described as a wagging of the
CO bond perpendicular to the plane of the ring. Although
the zeroth-order energy difference between ν1 and 2ν5 is
not that small (186 cm-1; Table 1), the magnitude of the
J. Org. Chem. Vol. 75, No. 6, 2010 1819

Shaffer et al.
JOCFeatured Article
SCHEME 3
diazirinone (2150 cm-1) is not consistent with the value
predicted (2046 cm-1) by a sophisticated level of quantum
theory. (iv) The high optical density (small %T) of the
previously reported experimental spectrum does not allow
a precise determination of the peak position. This spectrum
may be interpreted in terms of the broad, featureless
absorption of CO in a condensed phase. (v) The evolution
of the broad IR absorption at 2150 cm-1 to give absorp-
tions at 2116 and 2169 cm-1 may be rationalized in terms
of the degassing of CO from the condensed phase into the
gas phase. We therefore conclude that diazirinone (1), if
formed at all, has a much shorter lifetime than originally
suggested.10,13-18
Experimental Section
General Methods. The matrix isolation apparatus and general
techniques for its use have been described previously.35,36
Computational Methods. Geometries, anharmonic force field
corrections, vibrational frequencies, and rotational constants
were computed for diazirinone (1; C2v) using the coupled-cluster
singles and doubles model, together with a perturbative treat-
ment of triple excitations [CCSD(T)],37 as implemented in the
CFOUR software package.38 The basis set used was the atomic
assigned to diazirinone (1). Although we cannot exclude the
mechanism proposed for the formation of diazirinone
(Scheme 3, path a), our observations require a much shorter
lifetime for diazirinone than originally suggested. Two plau-
sible explanations are that diazirinone is formed, but is
unstable to the reaction conditions, or that N2 and CO are
produced without the intervention of diazirinone (Scheme 3,
path b). The reaction of 3-chloro-3-( p-nitrophenoxy)diazirine
(5) with Bu4NþF- undoubtedly affords p-fluoronitrobenzene
(7), nitrogen, and carbon monoxide.34 At the concentrations
typically employed in these experiments, the decomposition of
diazirine 5 would create a supersaturated solution of carbon
monoxide. As the reaction mixture warms from -15 °C, CO
will degas from solution, flooding the IR sample compartment
with gas-phase CO. This rationalization is consistent with the
previously reported IR experiments, in which the initially
observed band at 2150 cm-1 gives way to the spectrum of
natural orbital basis ANO2, which is based on Taylor and
€
39
Almlof’s natural atomic orbitals, truncated to 5s4p3d2f1g
on each atom. Geometries were further optimized using analytic
gradients,40,41 and harmonic frequencies obtained with analytic
second derivatives.42 Following this, the cubic and quartic force
constants were calculated by numerical differentiation of ana-
lytic second derivatives calculated at displaced points, following
the approach of Stanton et al.43 Second-order vibrational
perturbation theory (VPT2) was then used to compute the
fundamental vibrational frequencies and ground state rota-
tional constants.44 All calculations were done in the frozen-core
approximation with the CFOUR program system.
Attempted Matrix Isolation Trapping of Diazirinone (1). The
literature procedure for the reaction of tetrabutylammonium
fluoride and 3-chloro-3-(p-nitrophenoxy)diazirine (5) was
modified slightly to allow for matrix isolation. Molten tetra-
butylammonium fluoride was prepared in a 25 mL round-
bottom flask, equipped with a side arm, by warming tetrabutyl-
ammonium fluoride trihydrate (1.0 g, 3.2 mmol) to 50 °C under
vacuum overnight (0.2 mmHg). The flask was then connected to
a long path gas trap (20 cm) that could be shut off on both ends,
which then fed into the deposition chamber of the matrix
gas-phase CO (P and R branches at 2116 and 2169 cm-1
,
respectively). When we prepared a solution of carbon mono-
xide in acetonitrile at -35 °C, an absorption at 2140 cm-1 was
observed. Convincingly, the sample degasses carbon mono-
xide over the same time period (10 min) as the previously
reported decomposition of “diazirinone” at 2150 cm-1 and the
unresolved P and R branches simultaneously appear at nearly
identical positions to those reported in Ref. 10.
Summary. Diazirinone (1) was previously reported as a
metastable product of the reaction of 3-chloro-3-(p-nitro-
phenoxy)diazirine (5) with Bu4NþF-. The assignment was
based on the decay of an infrared absorption at 2150 cm-1
and the increase of two peaks for carbon monoxide at 2116
and 2169 cm-1. Our attempts to isolate diazirinone, how-
ever, led us to question this assignment: (i) Trapping of the
volatile reaction products under matrix-isolation conditions
(codeposition with argon at 10 K) led only to the observation
of CO. (ii) No evidence was obtained to suggest that dia-
zirinone was condensed in a vacuum trap cooled to 77 K. (iii)
The frequency of the experimental CdO stretch attributed to
(35) McMahon, R. J.; Chapman, O. L.; Hayes, R. A.; Hess, T. C.;
Krimmer, H. P. J. Am. Chem. Soc. 1985, 107, 7597–7606.
(36) Seburg, R. A.; McMahon, R. J. J. Am. Chem. Soc. 1992, 114, 7183–
7189.
(37) Raghavachari, K.; Trucks, G. W.; Pople, J. A.; Head-Gordon, M.
Chem. Phys. Lett. 1989, 157, 479–483.
(38) Stanton, J. F.; Gauss, J.; Harding, M. E.; Szalay, P. G. with
contributions from Auer, A. A.; Bartlett, R. J.; Benedikt, U.; Berger, C.;
Bernholdt, D. E.; Bomble, Y. J.; Christiansen, O.; Heckert, M.; Heun, O.;
ꢀ
Huber, C.; Jagau, T.-C.; Jonsson, D.; J. Juselius, Klein, K.; Lauderdale,
W. J.; Matthews, D. A.; Metzroth, T.; D. P. O’Neill, Price, D. R.; Prochnow,
ꢀ
E.; , Ruud, K.; Schiffmann, F.; Stopkowicz, S.; J. Vazquez, Wang, F.; Watts,
J. D. and the integral packages: MOLECULE (J. Almlof and P. R. Taylor),
€
PROPS (P. R. Taylor), ABACUS (T. Helgaker, H. J. Aa. Jensen, P.
€
Jorgensen, and J. Olsen), and ECP routines by A. V. Mitin, C. van Wullen.
www.cfour.de.
(39) Almlof, J.; Taylor, P. R. J. Chem. Phys. 1987, 86, 4070–4077.
(40) Scuseria, G. E. J. Chem. Phys. 1991, 94, 442–447.
(41) Lee, T. J.; Rendell, A. P. J. Chem. Phys. 1991, 94, 6229–6236.
(42) Gauss, J.; Stanton, J. F. Chem. Phys. Lett. 1997, 276, 70–77.
(43) Stanton, J. F.; Lopreore, C. L.; Gauss, J. J. Chem. Phys. 1998, 108,
7190–7196.
(34) Based on the typical amount of 3-chloro-3-(p-nitrophenoxy)diazirine
(5) used (35 mg) and the relative product yields, the theoretical concentration
of carbon monoxide would be 0.586 M, far exceeding the room-temperature
solubility in acetonitrile at 1 atm (0.005 M).
(44) Mills, I. M. In Molecular Spectroscopy: Modern Research; Rao, K.
N., Mathews, C. W., Eds.; Academic Press: New York, 1972; p 115.
1820 J. Org. Chem. Vol. 75, No. 6, 2010

Shaffer et al.
JOCFeatured Article
isolation apparatus. The system was evacuated and the matrix
window cooled to 25 K. The gas trap was cooled in an LN2 bath
(77 K). 3-Chloro-3-(p-nitrophenoxy)diazirine (5) (50 mg, 0.23
mmol) was dissolved in 1 mL of solvent (typically CH3CN or
CHCl3). This solution was added through the rubber-septum
sealed side arm of the 25 mL round-bottom flask with stirring at
-10 °C. The side arm was closed, and the reaction bubbled steadily
for 30 min while the temperature was maintained at -10 °C, before
being closed off to allow infrared characterization of the matrix.
Product Analysis. The reaction residual reaction mixture from
the matrix isolation experiment was dissolved in 5 mL of CH2Cl2
and washed twice with 5 mL of water before drying with Na2SO4
and concentrated by rotary evaporation. This crude mixture
contained the expected 3-fluoro-3-(p-nitrophenoxy)diazirine
(9), p-fluoronitrobenzene (7), and p-nitrophenol as primary
products.10 Flash column chromatography (10% EtOAc in
hexanes) of the concentrate resulted in an inseparable mixture
of 3-fluoro-3-(p-nitrophenoxy)diazirine (9) and p-fluoronitro-
benzene (7); these species were identified by NMR spectroscopy.
3-Fluoro-3-(p-nitrophenoxy)diazirine (9) exhibited spectro-
scopic data in excellent agreement with that reported pre-
viously:10 1H NMR (δ, CDCl3) 7.38 (m, 2H), 8.27 (m, 2H);
13C NMR (δ, CDCl3) 86.8 (diazirine C, d, JC-F=275 Hz giving
p-Fluoronitrobenzene (7) was identified by comparison to
commercially available material: 1H NMR (δ, CDCl3) 7.18 (m,
2H), 8.21 (m, 2H); 19F NMR (δ, CDCl3 referenced using unified
scale based on proton TMS reference) -106.5.
An EtOAc flush of the column afforded spectroscopically
pure p-nitrophenol: 1H NMR (δ, CDCl3) 6.90 (m, 2H), 8.13 (m,
2H); HRMS (EIþ) calcd for [C6H5NO3]þ requires m/z 139.0264,
found 139.0261.
IR Spectrum of CO: Gas Phase. A balloon of carbon mono-
xide (Matheson) was emptied into the cavity of the infrared
spectrometer and sealed from atmosphere. Thirty-two scans
were taken at a lower resolution setting (1.928 cm-1) before
background subtraction and identification of the P and R
branches at 2116 and 2169 cm-1
.
IR Spectrum of CO: Acetonitrile Solution. A Hellma 260 QS
282 1.000 cm path length quartz cell was filled with acetonitrile.
A balloon of CO (Matheson) was filled at atmospheric pressure
and bubbled through the acetonitrile in the cuvette for 30 min at
-35 °C. The cell was immediately placed on a stand in the IR
chamber and measured approximately every two min over the
course of 16 min. The observed transition at 2140 cm-1 neared
complete decay over 16 min and was absent from the spectrum
of pure acetonitrile.
rise to resonances at 85.4 and 88.2), 117.4, 126.1, 144.7, 156.2; 19
F
NMR (δ, CDCl3, referenced to o-difluorobenzene) -119.0;10 (δ,
CDCl3, referenced using unified scale based on proton TMS
reference): -123.5.45 Attempts to identify 3-fluoro-3-(p-nitro-
phenoxy)diazirine (9) using EIþ and ESIþ mass spectrometry
were unsuccessful. There is little literature precedent for exact
mass identification of halodiazirines, as these molecules
commonly undergo complex fragmentation pathways once
ionized.46 Attempts to interpret fragmentation data were
further complicated by contamination with p-fluoronitro-
benzene (7).
Acknowledgment. We gratefully acknowledge financial
support from the National Science Foundation (NSF-
0715305 (R.J.M.) and NSF-0710146 (J.F.S.)). We thank
Mitch Daane for meaningful discussions and valuable
insight.
Note Added after ASAP Publication. This paper was pub-
lished on the Friday, February 12, 2010, with errors in Table 1
and some of the text. The corrected version was reposted on
February 18, 2010.
(45) Harris, R. K.; Becker, E. D.; Cabral De Menezes, S. M.; Goodfellow,
R.; Granger, P. Pure Appl. Chem. 2001, 73, 1795–1818.
(46) Burkholder, C. D.; Jones, W. E.; Wasson, J. S.; Liu, M. T. H. J. Am.
Chem. Soc. 1980, 102, 2847–2848.
Supporting Information Available: NMR spectra of diazir-
ine 5 and its reaction products. This material is available free of
charge via the Internet at http://pubs.acs.org.
J. Org. Chem. Vol. 75, No. 6, 2010 1821