that precursor was inconvenient and explosive. We reported
that the reaction of p-nitrophenoxychlorodiazirine (6) with
tetrabutylammonium (TBA) fluoride provided diazirine 1 and
diazirinone.4 Now we describe a related, but more efficient
and reproducible, preparation of 1 (cf. Scheme 1).
Chlorofluorodiazirine was identified by its UV spectrum
in pentane (see Supporting Information, Figure S-1) which
displayed absorption peaks at 356, 351, 344, 339, 334, 329,
324, and 320 nm, in excellent agreement with the published
(gas phase) spectrum.9 Diazirine 1 also afforded a 19F NMR
resonance at δ -104.8 (CFCl3, CDCl3).
LFP6 of diazirine 1 (A356 ) 0.5 in pentane) affords a weak
absorption at 368 nm (Figure 1) which we assign to ClCF.
Scheme 1
N-Methanesulfonyloxy-O-phenylisourea (7)11 was oxidized
to phenoxychlorodiazirine (8)12 by NaOCl in methanol. The
diazirine, purified by flash chromatography, was obtained
in 84% yield.13 Diazirine 8 was converted to phenoxyfluo-
rodiazirine (9) by exchange with “molten” TBAF,14 and the
fluorodiazirine was dinitrated to 2,4-dinitrophenoxyfluorodi-
azirine (10) with nitronium tetrafluoroborate in nitromethane.
Diazirine 10 was slowly added to a solution of 3-fold excess
anhydrous TBACl in HMPA under a vacuum at 25 °C.
Chlorofluorodiazirine 1 distilled out and was trapped in 1
mL of pentane or CDCl3 at 77 K. After dilution with an
additional 1 mL of solvent, we obtained solutions of 1 with
A ) 1.6-2.0 at 356 nm.
The diazirine exchange reaction which converts 10 to 1
most likely proceeds via a double SN2′ mechanism,1 in which
an initial SN2′ attack of Cl- at diazirine N leads to the loss
of 2,4-dinitrophenoxide and the generation of a N-chloro-
isodiazirine. The latter is then converted to 1 by a second
SN2′ attack of Cl- at carbon. (The process is analogous to
the conversion of 2,4-dinitrophenoxychlorodiazirine to dichlo-
rodiazirine by TBACl.6) A competitive ipso attack of Cl-
on the 2,4-dinitrophenoxy moiety of 10 presumably leads
to diazirinone (11), which subsequently fragments to nitrogen
and carbon monoxide (within 5 min at rt).4
Figure 1. LFP UV spectrum of ClCF under nitrogen. The spectrum
is unchanged under oxygen.
The spectrum is identical under nitrogen or oxygen. There
is no evidence for rapid reaction of ClCF with oxygen, in
contrast to the behavior of CCl2.6 The 368 nm absorption
accords with the reported UV specrum of ClCF at 340-390
nm obtained by photolysis of CH2ClF or CD2ClF in Ar
matrices at 14 K.15,16
The computed absorption of ClCF is at 365 nm (Table
1), in excellent agreement with experiment. For CCl2, ClCF,
and CF2, the percent contribution of the HOMO-LUMO
excitation to the lowest carbene excited state is nearly
constant at 75%. The doubly occupied in-plane σ lone pair
and the formally vacant 2p orbitals on C are the primary
contributors to these two molecular orbitals, so that a σ f
p designation for the excitation appears appropriate. Note
that the computed oscillator strength (f) of ClCF is 3 times
that of CCl2. Given that the observed absorbance of ClCF is
very weak (Figure 1), it is perhaps not surprising that we
were unable to observe the absorbance of CCl2 under
comparable LFP conditions.6
(9) Mitsch, R. A.; Neuvar, E. W.; Koshar, R. J.; Dybvig, D. H. J.
Heterocycl. Chem. 1965, 2, 371.
(10) Mitsch, R. A. J. Org. Chem. 1968, 33, 1847.
(15) Smith, C. E.; Milligan, D. E.; Jacox, M. E. J. Chem. Phys. 1971,
54, 2780.
(11) Fede´, J.-M.; Jockusch, S.; Lin, N.; Moss, R. A.; Turro, N. J. Org.
Lett. 2003, 5, 5027.
(12) Moss, R. A.; Perez, L. A.; Wlostowska, J.; Guo, W.; Krogh-
Jespersen, K. J. Org. Chem. 1982, 47, 4177.
(13) See the Supporting Information for experimental details.
(14) (a) Cox, D. P.; Moss, R. A.; Terpinski, J. J. Am. Chem. Soc. 1983,
105, 6513. (b) Moss, R. A.; Terpinski, J.; Cox, D. P.; Denney, D. Z.; Krogh-
Jespersen, K. J. Am. Chem. Soc. 1985, 107, 2743.
(16) A reviewer has noted that the ClCF absorption is surprisingly narrow
for a HOMO to LUMO carbene band. At its base, it spans ∼100 nm. In
comparison, the HOMO to LUMO absorption of MeCCl in pentane solution
spans ∼200 nm (Moss, R. A.; Tian, J.; Sauers, R. R.; Krogh-Jespersen, K.
J. Am. Chem. Soc. 2007, 129, 10019). There are very few HOMO-LUMO
bands of carbenes in solution to afford further comparisons. Note that we
do not observe diazirine bleaching in Figure 1; only about 0.1% of the
diazirine is destroyed per laser pulse.
4054
Org. Lett., Vol. 9, No. 20, 2007