Laser flash and dual wavelength photolysis of 3,4-diaza-2,2-dimethoxy-1-oxa[4.5]spirooct-3-ene. Migration of hydrogen and carbon in cyclobutylidene and in the excited state of its precursor

Full text of DOI:10.1021/ja961933r

J. Am. Chem. Soc. 1997, 119, 3191-3192
3191
irradiation (Rayonet) of 1b9 gave primarily diazocyclobutane
(2),10 a little of the corresponding azine, and traces of alkenes,11
two-color photolysis (SS, 250 and 300 nm) of 1b gave a mixture
of cyclobutene (CB) and methylenecyclopropane (MC).12 In
the presence of tetramethylethylene (TME), carbene 4 was
trapped as cycloadduct 5.13 Together, the LFP and SS results
permitted estimates of rate constants for the 1,2-rearrangements
of 4 and the determination of the product ratio (CB/MC) from
carbene and excited state sources.
LFP (308 nm) of either 1a or 1b in the presence of pyridine
gives an intense spectrum of the pyridinium ylide of 4. The
similarity of transient spectra and lifetime from an oxadiazoline
and a diazirine was established by LFP of 7 (308 nm) and 8
(351 nm).14-17 Presumably a two-photon or a multiphoton
Laser Flash and Dual Wavelength Photolysis of
3,4-Diaza-2,2-dimethoxy-1-oxa[4.5]spirooct-3-ene.
Migration of Hydrogen and Carbon in
Cyclobutylidene and in the Excited State of Its
Precursor
John Paul Pezacki,*,1 David L. Pole, and John Warkentin*
Department of Chemistry, McMaster UniVersity
1080 Main Street West
Hamilton, Ontario, Canada L8S4M1
Tongqian Chen,2 Francis Ford, John P. Toscano,3
Jennifer Fell,4 and Matthew S. Platz*
Department of Chemistry, The Ohio State UniVersity
120 West 18th AVenue, Columbus, Ohio 43210
ReceiVed June 10, 1996
The most common photochemical sources of carbenes are
diazirines and diazo compounds5 but such nitrogenous precur-
sors can form the same products from either the carbenes or
the excited states of the precursors, albeit in different ratios.6
Consequently, the observed product ratios are not necessarily
indicative of carbene chemistry.6 Dialkyldiazo compounds
themselves can be generated by direct7 (300 nm) or sensitized8
photolysis of oxadiazolines. At 25 °C, the diazoalkanes are
thermally and photochemically (300 nm) fairly stable, affording
azines slowly together with only traces of coproducts associated
with intramolecular rearrangement of corresponding dialkyl-
carbenes.
We are pleased to report that laser flash photolysis (LFP,
308 nm) of 3,4-diaza-2-methoxy-2-methyl-1-oxa[4.5]spirooct-
3-ene (1a) in the presence of pyridine gave the pyridinium ylide
of carbene 4 (λmax) 350-360 nm), permitting estimates of the
lifetime of 4. Moreover, while 300 nm steady state (SS)
process is responsible for the generation of 4 from 1a in the
laser beam. Analysis of the data14,19 from different pyridine
concentrations gave the lifetimes (τ) of 4 in CF2ClCFCl2,
cyclohexane, or cyclohexane-d12, as 4-20 ns ((20%) and, in
acetonitrile, 0.4-2 ns ((20%), assuming18 that kpyr ) 1-5 ×
109 M-1 s-1. Stern-Volmer (LFP) experiments14 reveal that
carbene 4 reacts with TME and pyridine with the same rate
constant within experimental error. As τ is identical in C6H12
and C6D12, we conclude that the lifetime of 4 in alkane solvent
is controlled by intramolecular processes and that the sum of
the rate constants for 1,2-H and 1,2-C migrations, kH + kC )
0.5 - 2.5 × 108 s-1 at ambient temperature.
(9) Oxadiazolines 1a and 1b show the same photochemistry except that
1a gives methyl acetate and 1b gives dimethyl carbonate. The advantage
in using 1b is that the 1H-NMR spectrum is not complicated by dia-
stereotopicity of the cyclobutyl hydrogens.
(10) (a) Brinker, U. H.; Boxberger, M. Angew. Chem., Int. Ed. Engl.
1984, 23, 974. (b) Applequist D. E.; McGreer D. E. J. Am. Chem. Soc.
1960, 82, 1965.
(11) Photolysis with 300 nm light alone (Pyrex, Rayonet) gave solutions
of diazocyclobutane that were persistent for many hours, slowly affording
primarily cyclobutanone azine and only traces (GC) of alkenes. Similar
irradiation of 2-methoxy-2,5,5-trimethyl-∆3-1,3,4-oxadiazoline gave, even
after prolonged exposure, only trace amounts of propene from facile 1,2-H
migration in dimethylcarbene. In an apparent conflict with the LFP, 308
nm light converts oxadiazoline to carbene whereas in SS experiments 300
and 250 nm light are required simultaneously in order to produce carbene
efficiently. The LFP method, however, is much more sensitive than chemical
analysis, and it can involve multiphoton processes which are not possible
in SS photolysis.
(1) NSERC Scholar 1996.
(2) Present address: Department of Chemistry, Colorado State University,
Fort Collins, CO.
(3) Present address: Department of Chemistry, Johns Hopkins University,
Baltimore, MD.
(4) Present address: Department of Chemistry, University of California,
Berkeley, CA.
(5) (a) Kirmse, W. Carbene Chemistry, 2nd ed.; Academic Press: New
York, 1971. (b) Baron, W. J.; DeCamp, M. R.; Hendrick, M. L.; Jones,
M., Jr.; Levin, R. H.; Sohn, M. B. In Carbenes; Jones, M., Jr., Moss, R.
A., Eds.; Wiley: New York, 1973; Vol. I, p 1. (c) Liu, M. T. H., Ed.
Chemistry of Diazirines; CRC Press: Boca Raton, FL, 1987. (d) Zollinger,
H. Diazo Chemistry; VCH Publishers: New York, 1995; Vol. 2. (e) Moss,
R. A. In AdVances in Carbene Chemistry; Brinker, U., Ed.; JAI Press:
Greenwich, CT, 1994; Vol. 1, p 59. (f) Jackson, J. E.; Platz, M. S. In
AdVances in Carbene Chemistry; Brinker, U., Ed.; JAI Press: Greenwich,
CT, 1994; Vol. 1, p 89. (g) Nickon, A.; Huang, F.-C.; Weglein, R.; Matsuo,
K.; Yagi, H. J. Am. Chem. Soc. 1974, 96, 5264. (h) Kyba, E. P.; Hudson,
C. W. J. Org. Chem. 1977, 42, 1935. (i) Kyba, E. P.; John, A. M. J. Am.
Chem. Soc. 1977, 99, 8329. (j) Press, L. S.; Shechter, H. J. Am. Chem.
Soc. 1979, 101, 509. (k) Freeman, P. K.; Hardy, T. A.; Baleat, J. R.; Wescott,
L. D., Jr. J. Org. Chem. 1977, 42, 3356. (l) Seghers, L.; Shechter, H.
Tetrahedron Lett. 1976, 23, 1943.
(12) Photolyses were for 2-3 h in degassed and sealed quartz NMR
tubes in a Rayonet chamber holding 14 × 250 nm lamps and 2 × 300 nm
lamps. A mixture of CB and MC (2-5% yield, relative to internal standard,
remainder mostly diazocyclobutane) was isolated by preparative GC and
identified by 1H-NMR and GC-MS. 1H-NMR (C6D6, 200 MHz) showed δ
5.46 (quintet, J ) 2.1 Hz, 2H) and 0.88 (t, J ) 2.1 Hz, 4H) from MC and
δ 5.91 and 2.44 (singlets; unresolved coupling) from CB. GC-MS (oven
30 °C) showed a broad peak at ca. 2 min retention time (before solvent)
with M ) 54 (C4H6) and a base peak of mass 39 (M - 15).
1
(13) Adduct 5 was isolated by preparative GC. H-NMR (microprobe,
(6) For discussions of the problem and for examples of excited state
rearrangements leading to products expected from carbenes, see: (a) ref
5e. (b) Modarelli, D. A.; Morgan, S.; Platz, M. S. J. Am. Chem. Soc. 1992,
114, 7034. (c) Moss, R. A. Pure Appl. Chem. 1995, 67, 741. (d) Platz, M.
S.; White, W. R., III; Modarelli, D. A. Res. Chem. Intermed. 1994, 20,
175. (e) Chen, N.; Jones, M., Jr.; Platz, M. S. J. Am. Chem. Soc. 1991,
113, 4981. (f) White, W. R., III; Platz, M. S. J. Org. Chem. 1992, 57, 2841.
(g) Moss, R. A.; Ho, G.-H. J. Phys. Org. Chem. 1993, 6, 126. (h) Moss, R.
A.; Liu, W.; Krogh-Jespersen, K. J. Phys. Chem. 1993, 97, 13413. (i) Chen,
N.; Jones, M., Jr. Tetrahedron Lett. 1989, 1, 305. (j) Eaton, P. E.; Appell,
R. B. J. Am. Chem. Soc. 1990, 112, 4055. (k) Eaton, P. E.; Hoffmann,
H.-L. J. Am. Chem. Soc. 1987, 109, 5285. (l) Wierlacher, S.; Sander, W.;
Liu, M. T. H. J. Am. Chem. Soc. 1993, 115, 8943.
C6D6, 500 MHz) δ: 1.87 (m, 6H), 0.88 (s, 12H). Single peak in the GC-
MS trace (oven 30 °C) at retention time ca. 12 min, mass 138 (M, C10H18)
and base peak 123 (M - 15).
(14) Platz, M. S.; Modarelli, D. A.; Morgan, S.; White, W. R.; Mullins,
M.; Celebi, S.; Toscano, J. P. Prog. React. Kinet. 1994, 19, 93.
(15) (a) Modarelli, D. A.; Platz, M. S. J. Am. Chem. Soc. 1991, 113,
8985. (b) Modarelli, D. A.; Morgan, S.; Platz, M. S.; J. Am. Chem. Soc.
1992, 114, 7034. (c) Platz, M. S.; White, W. R., III; Modarelli, D. A.; Celebi,
S. A. Res. Chem. Intermed. 1994, 175.
(16) Jackson, J. E.; Platz, M. S. In AdVances in Carbene Chemistry;
Brinker, U., Ed.; JAI Press, Greenwich, CT, 1994; Vol. 1, p 89.
(17) Although methoxymethylcarbene might also be generated from 1a,
it is known to react slowly with pyridine (kpyr ) 6.6 × 105 M-1 s-1) to
form an ylide with λmax ) 380 nm.18
(18) Ge, C. S.; Jane, E. G.; Jefferson, E. A.; Liu, W.: Moss, R. A.;
Włostowska, J.; Xue, S. J. Chem. Soc., Chem. Commun. 1994, 1479.
(19) (a) Frey, H. M. AdV. Photochem. 1964, 4, 225. (b) Frey, H. M. J.
Am. Chem. Soc. 1962, 84, 2293.
(7) (a) Majchrzak, M. W.; Be´khazi, M.; Tse-Sheepy, I.; Warkentin, J. J.
Org. Chem. 1989, 54, 1842. (b) Majchrzak, M. W.; Jefferson, E. A.;
Warkentin, J. J. Am. Chem. Soc. 1990, 112, 1842. (c) Jefferson, E. A.;
Warkentin, J. J. Org. Chem. 1994, 59, 455.
(8) Adam, W.; Finzel, R. Tetrahedron Lett. 1990, 31, 863.
S0002-7863(96)01933-6 CCC: $14.00 © 1997 American Chemical Society

3192 J. Am. Chem. Soc., Vol. 119, No. 13, 1997
Communications to the Editor
Scheme 1
Figure 1. Percent yield of 5 vs TME concentration in cyclohexane-
d12 (O), in acetonitrile-d3 (b), and in neat TME (9). Curve fitting of
the data in cyclohexane solutions is shown by the solid line and in
acetonitrile solutions by the dashed line. The plot shows a leveling off
at ∼21% indicating that an unquenchable reaction is occurring. The
inset shows double reciprocal plots for quenching of 4 by TME in
cyclohexane-d12 (O) and in acetonitrile-d3 (b).14,19 The slope/intercept
ratio gives kqτ in each solvent.
Simultaneous steady state photolysis of 1b, with 250 and 300
nm wavelengths at ambient temperature, gave CB and MC
which is consistent with the original report20 of thermal
decomposition of the sodium salt of cyclobutanone tosylhydra-
zone and with the results of dehalogenation of gem-dihalo-
cyclobutanes.21 Two-color photolysis of 1b in neat TME gave
4,4,5,5-tetramethyl[2.3]spirohexane (5)13 (Scheme 1). Diazo-
cyclobutane is known to undergo [3 + 2] cycloaddition reactions
with alkenes to give pyrazolines10a which can subsequently lose
N2 thermally10b or photochemically.10a Therefore precursor 1b
was photolyzed in neat TME with 300 nm light alone to
determine whether 5 was the indirect result of [3 + 2]
cycloaddition of 2 to TME or the direct result of carbene
of the absolute rate constant for 1,2-C migration would then be
2.5 × 108 s-1 and 5.0 × 107 s-1, at ∼25 °C, respectively.
Two-color photolyses of 1b in acetonitrile gave changes in
MC/CB ratios, as a function of [TME], analogous to those
obtained in cyclohexane solvent. Thus, for 4, solvent effects
on 1,2-H and 1,2-C migrations are similar. Yields of 5 as a
function of [TME] increased more rapidly in cyclohexane
compared with acetonitrile solutions (Figure 1). Double
reciprocal plots14,19,25 (Figure 1, inset) gave the ratio intercept/
slope which is equal to kTMEτ for quenching. For cyclohexane,
the ratio was 5.2 M-1 meaning that kTME ) 0.26-1.3 × 109
M-1 s-1 at ∼25 °C (τ ) 4-20 ns). For acetonitrile, the ratio
was 1.3 M-1 indicating that 4 has a much shorter lifetime in
that solvent (0.3-1 ns). Both results are consistent with the
data obtained by quenching with pyridine (above). More facile
1,2-H and 1,2-C migrations in a polar solvent indicates that
corresponding transition states from 4 are polar. 1,2-Hydrogen
migration in carbenes appears to be accelerated in other cases
by polar solvents.26 1,2-Carbon migration is special for 4, which
rearranges through a dipolar, nonclassical transition structure
according to Schoeller27 and Sulzbach et al.28
1
trapping. The resulting mixtures, analyzed by H-NMR and
GC-MS, revealed that 5 was not formed with 300 nm light alone,
excluding diazo compound 2 as the source of adduct 5.
Photolyses of 1b (250 and 300 nm) in cyclohexane solutions
with different concentrations of TME22 showed that yields of
CB and MC decreased as a function of increasing [TME] and
then leveled off, indicating that part of the rearrangement
reactions cannot be quenched by a carbene trap. Yields of 5
showed a similar saturation, at a maximum near 21% relative
to CB and MC, with increasing [TME]. On the basis of the
observed ratio of MC/CB (5.5:1) in cyclohexane, upper limits
at ∼25 °C would be kH ) 4 × 107 s-1 and kC ) 2 × 108 s-1
.
Corresponding lower limits would be kH ) 8 × 106 s-1 and kC
) 4 × 107 s-1
. However, the ratio of 1,2-H and 1,2-C
migrations was also observed to change as a function of [TME].
The data suggest that ca. 21% of the observed products are the
result of the intramolecular rearrangement of 4 and that the
remainder results from migrations in the excited state (3) of
diazo precursor 2.23 Since TME at g3 M captured all of 4, the
limiting ratio CB/MC reflects the partitioning of 3 between 1,2-C
and 1,2-H migration. At [TME] ) 0, when the ratio reflects a
composite of excited state and carbene rearrangements, the value
was ca. 5.5 in cyclohexane, while at high [TME] it was 3.6. In
order to change the ratio from 3.6 (excited state alone) to 5.5
(composite) with a 21% contribution from 4, it is clear that most
or all of 4 must rearrange to MC.24 Given that the lifetime (τ)
of 4 was measured as 4-20 ns in cyclohexane-d12 (above),
based18 on kpyr ) 1-5 × 109 M-1 s-1 , upper and lower limits
We are currently investigating other carbenes and other alkene
traps in order to determine product distributions and pathways
to product formation.
Acknowledgment. J.P.P. and J.W. thank Dr. Don Hughes and Mr.
Brian Sayer for help with NMR spectroscopy, Prof. W. J. Leigh for
advice on Rayonet experiments, and the NSERC of Canada for
continued support through an operating grant. M.S.P. gratefully
acknowledges the support of the NSF (CHE-8814950). J.T gratefully
acknowledges NIH postdoctoral fellowship support.
Supporting Information Available: Figures showing transient
absorption vs wavelength, [Py] vs Ay, and 1/[Py] vs 1/Ay and table
giving yields and ratios of products of quenching of 4 by TME (3
pages). See any current masthead page for ordering and Internet access
instructions.
JA961933R
(20) Friedman, L.; Shechter, H. J. Am. Chem. Soc. 1960, 82, 1002.
(21) Brinker, U. H.; Schenker, G. J. Chem. Soc., Chem. Commun. 1982,
679.
(22) A merry-go-round Rayonet apparatus ensured that all solutions were
photolyzed under identical conditions. Yields at low conversion were
obtained by 500 MHz 1H NMR spectroscopy with suppression of the TME
singlet as well as digital filtering of that singlet outside the spectral window.
Integrations of unobscured signals were compared with that of internal
standard hexamethyldisilane.
(23) Another interpretation would ascribe these results to the formation
of a carbene-olefin complex which can rearrange to CB and MC or collapse
to form adduct. See: Tomioka, H.; Hayashi, N.; Izawa, Y.; Liu, M. T. H.
J. Am. Chem. Soc. 1984, 106, 454. Bonneau, R.; Liu, M. T. H.; Kim, K.
C.; Goodman, J. L. J. Am. Chem. Soc. 1996, 118, 3829 and references
therein.
(24) While the data require that the 21% contribution from 4 be in the
form of MC, it is not possible to exclude CB formation entirely, because
of experimental errors. If excited state chemistry is not involved but 4 forms
a complex with TME,23 then the ratio MC:CB without added TME
represents the product distribution from the free carbene. In either case,
the data suggest that 1,2-C migration is highly favored.
(25) Curvature in the plots of quenching data in both solvents, as a result
of more than one adduct forming intermediate, could introduce additional
errors in kqτ.
(26) Sugiyama, M. H.; Celebi, S.; Platz, M. S. J. Am. Chem. Soc. 1992,
114, 966.
(27) Schoeller, W. W. J. Am. Chem. Soc. 1979, 101, 4811.
(28) Sulzbach, M.; Platz, M. S.; Schaefer, H. F., III.; Hadad, C. J. Am.
Chem. Soc. In press.