Stereochemical deuterium labeling as mechanistic probe for differentiating the singlet- and triplet-diradical spin states in the rearrangement of the 2-spiroepoxy-1,3-cyclopentanediyl to oxetanes [11]

Full text of DOI:10.1021/ja000656s

6508
J. Am. Chem. Soc. 2000, 122, 6508-6509
Stereochemical Deuterium Labeling as Mechanistic
Probe for Differentiating the Singlet- and
Triplet-Diradical Spin States in the Rearrangement
of the 2-Spiroepoxy-1,3-cyclopentanediyl to Oxetanes
Manabu Abe,*,† Waldemar Adam,‡ Yasunori Ino,† and
Masatomo Nojima†
Department of Materials Chemistry,
Graduate School of Engineering
Osaka UniVersity, Suita 565-0871, Osaka, Japan
Insitut fu¨r Organische Chemie der UniVersita¨t Wu¨rzburg
Am Hubland,D-97074 Wu¨rzburg, Germany
ReceiVed February 23, 2000
Figure 1. Partial 1H NMR spectra (600 MHz) of the oxetane 2, formed
in the photolysis of (a) nondeuterated azoalkane 1, (b) anti(d2)-1, and (c)
syn(d2)-1.
Substituent effects on the spin multiplicity of diradicals and
their reactivity have been a topic of timely current interest.1,2 We
have recently observed that for the 2-spiroepoxy-1,3-diradical 1,3-
DR the singlet ground state is preferred by ∼1 kcal/mol.3 The
latter is conveniently generated by photodenitrogenation of
spiroepoxy azoalkane 1, which unexpectedly led to the oxetane
2 by selective CO-bond cleavage of the epoxide ring.3 Herein
unequivocal stereochemical evidence is presented by means of
deuterium-labeling experiments that the oxetane (d2)-2 formation
in the photodenitrogenation of the azoalkanes anti(d2)- or syn(d2)-1
depends dramatically on the irradiation conditions (Scheme 1):
The direct (singlet) process proceeds with complete retention of
the initial azoalkane configuration, the benzophenone-sensitized
(triplet) one suffers extensive loss of stereochemical memory.
These unprecedented facts are interpreted mechanistically in terms
of a singlet nitrogen-free S-1,3-DR diradical in the direct
photodenitrogenation versus the triplet T-1,3-DR in the ben-
zophenone-sensitized one as precursors to the oxetane product
2.
The direct (344 ( 4 nm) and benzophenone-sensitized (380 (
4 nm) photodenitrogenation of the anti(d2)-1 (λmax ) 343 nm, ꢀ
185) and of the syn(d2)-1 (λmax ) 352 nm, ꢀ 85) in C6D6 with a
500-W Xenon lamp (a monochromator was used for wavelength
selection) led to the oxetane (d2)-2 (Scheme 1 and Table 1), for
which we have assigned the configurations of all the hydrogen
atoms by means of C,H-COSY and NOE measurements (600-
MHz in C6D6; cf. Supporting Information). The spectral data
allowed unequivocally to distinguish between all of the four Ha-d
hydrogen atoms, as indicated in Figure 1 (spectrum a). The direct
irradiation of anti(d2)-1 (d content 88 ( 5%) gave quantitatively
the dideuterated oxetane exo(d2)-2, whose d content was deter-
mined directly on the photolysate by 1H NMR (600 MHz)
spectroscopy (Table 1, entry 1). As is evident from Figure 1
(spectrum b), only the Ha and Hd positions (trans to the CO bond)
contain deuterium atoms (90 ( 5% content). Alternatively, the
direct irradiation of the syn(d2)-1 diastereomer (d content 88 (
5%) displayed deuteration (85 ( 5%) only at the Hb and Hc
positions [Table 1, entry 2, and Figure 1 (spectrum c)]. Thus,
these deuterium distributions unequivocally manifest that the
exo(d2)-2 or endo(d2)-2 oxetanes are formed stereoselectively
[100% retention within the experimental error (( 5%)] in the
direct photolysis of the azoalkanes anti(d2)-1 or syn(d2)-1.
Scheme 1
For the benzophenone-sensitized photolysis (380 ( 4 nm),
extensively randomized distributions of the deuterium labels were
observed at the four Ha-d positions (Table 1, entries 3 and 4). In
the absence of benzophenone, the oxetane (d2)-2 was not observed
on irradiation at 380 nm; thus, the oxetane formed in the presence
of benzophenone must be derived from the triplet-excited azo
chromophore.4-6 In the sensitized photolysis of anti(d2)-1 (d
content 88 ( 5%), the Ha and Hd positions in the oxetane (d2)-2
were deuterated only to the extent of 57 ( 5% (64% retention
when corrected for the d content in 1),7 the remainder (25 ( 5%)
of the deuteration was found at the Hb and Hc positions (Table 1,
entry 3). The sensitized photodenitrogenation of syn(d2)-1 (d
content 88 ( 5%) gave similar results (Table 1, entry 4); namely,
the extent of retention (%) of configuration in the oxetane (d2)-2
was found to be 59%. A control experiment with labeled oxetane
exo(d2)-2 showed that no deuterium scrambling occurred under
the benzophenone-sensitized irradiation conditions (380 ( 4 nm).
The results clearly signify that the stereochemical course of the
oxetane (d2)-2 formation depends on the spin state of the n,π*-
excited azo chromophore.
The mechanism in Scheme 2 is proposed to account for the
spin-state-dependent formation of the oxetane (d2)-2. To simplify
the mechanistic discussion, only the photoreaction of the anti(d2)-1
azoalkane is presented. It is generally accepted that diazenyl
diradicals DZ are the first intermediates in photodenitrogenation
of azoalkanes.8 In our case, the anti(d2)-S-DZ species may be
(4) Triplet-triplet energy transfer from benzophenone (ET ∼69 kcal/mol,
ref 5) to diazabicyclo[2.2.1]hept-2-ene (DBH) derivatives (ET ∼60 kcal/mol,
ref 6) is an energetically favored process; see, Engel, P. S. Chem. ReV. 1980,
80, 99.
(5) Leigh, W. J.; Arnold, D. R. J. Chem. Soc., Chem. Commun. 1980, 406.
(6) Rau, H. Angew. Chem., Int. Ed. Engl. 1973, 12, 224.
(7) Retention (%) ) [d-content in the Ha or Hd positions/d content in (d2)-
1] × 100; thus, for complete randomization the retention value (%) would be
50%.
† Osaka University.
‡ Universita¨t Wu¨rzburg.
(1) (a) Skancke, A.; Hrovat, D. A.; Borden, W. T. J. Am. Chem. Soc. 1998,
120, 7079. (b) Adam, W.; Borden, W. T.; Burda, C.; Foster, H.; Heidenfelder,
T.; Heubes, M.; Hrovat, D. A.; Kita, F.; Lewis, S. B.; Scheutzow, D.; Wirz,
J. J. Am. Chem. Soc. 1998, 120, 593. (c) Abe, M.; Adam, W.; Heidenfelder,
T.; Nau, W. M.; Zhang, X. J. Am. Chem. Soc. 2000, 122, 2019.
(2) Berson, J. A. Acc. Chem. Res. 1997, 30, 238.
(8) (a) Adams, J. S.; Weisman, R. B.; Engel, P. S. J. Am. Chem. Soc. 1990,
112, 9115. (b) Adam, W.; Denninger, V.; Finzel, R.; Kita, F.; Platsch, H.;
Walter, H.; Zang, G. J. Am. Chem. Soc. 1992, 114, 5027. (c) Reyes, M. B.;
Carpenter, B. K. J. Am. Chem. Soc. 1998, 120, 1641. (d) Yamamoto, N.;
Olivucci, M.; Celani, P.; Bernardi, F.; Robb, M. A. J. Am. Chem. Soc. 1998,
120, 2391. (e) Diau, E. W.-G.; Abou-Zied, O. K.; Scala, A. A.; Zewail, A. H.
J. Am. Chem. Soc. 1998, 120, 3245.
(3) Abe, M.; Adam, W.; Nau, W. M. J. Am. Chem. Soc. 1998, 120, 11304.
10.1021/ja000656s CCC: $19.00 © 2000 American Chemical Society
Published on Web 06/23/2000

Communications to the Editor
J. Am. Chem. Soc., Vol. 122, No. 27, 2000 6509
Table 1. Stereoselectivities in the Direct and Benzophenone-Sensitized Photodenitrogenation of the Diastereomerically Deuterium-Labeled
anti(d2)- and syn(d2)-Azoalkane 1a
d content (%) in oxetane (d2)-2b
entry
(d2)-1c
irradiation conditionsd
convn. (%)
Ha
Hb
Hc
Hd
oxetane (d2)-2
retn. (%)e
1
2
3
4
anti
syn
anti
syn
direct hV
direct hV
Ph2CO/hV
Ph2CO/hV
73
44
25
10
90
0
57
30
0
85
24
50
0
83
25
53
90
0
57
30
exo
endo
exo/endo
endo/exo
100
100
64
59
1
a At 25 °C, 54 µmol (d2)-1 in 0.7 mL of C6D6. b Deuterium content (%) was determined by H NMR (600 MHz) peak areas; error (5% of the
stated value. c The deuterium content is 88 ( 5%, the remaining12% is hydrogen. d 344 ( 4 nm for the direct irradiation and 380 ( 4 nm for the
Ph2CO-sensitized photolysis were used in the photodenitrogenation (500-W Xenon lamp, focused by a monochromator). e Extent of retention (%)
of the initial azoalkane configuration in the oxetane (d2)-2; for complete deuterium randomization the value would be 50%.
Scheme 2
several options: The most likely one is denitrogenation to the
triplet 1,3-diradical anti(d2)-T-1,3-DR, subsequent epoxide ring-
opening and cyclization would afford the oxetane 2. Prerequisite
for oxetane formation is intersystem crossing (ISC) to the singlet
state of the precursor species. If the ISC process of the T-1,3-
DR to the S-1,3-DR were to dominate, retention of the initial
configuration would prevail in the oxetane products; however,
this is not the case since experimentally extensive randomization
was observed in the oxetane (Table 1). Evidently, the anti(d2)-
T-1,3-DR triplet diradical suffers fast CO-bond cleavage to
generate the T-1,4-DR diradical. If the latter had enough lifetime
to rotate about the CC bond, a 1:1 mixture of exo and endo
oxetanes (d2)-2 would result; however, the randomization is not
perfect, as judged by the fact that from anti(d2)-1 64% and from
the syn(d2)-1 59% retention have been observed experimentally
(Table 1, entries 3 and 4). This indicates some residual stereo-
chemical memory in the triplet-sensitized process, for which the
following mechanistic possibilities are feasible: (1) ISC from the
triplet diazenyl diradical T-DZ to the singlet state S-DZ competes
with the N2 extrusion; (2) ISC from the triplet 1,3-diradical T-1,3-
DR to the singlet state S-1,3-DR competes with the CO-bond
cleavage; (3) ISC from the triplet 1,4-diradical T-1,4-DR to the
singlet state S-1,4-DR competes with the CC-bond rotation as
indicated in structure T-1,4-DR (Scheme 2). At this point it is
difficult to diagnose these alternatives experimentally.
In summary, our deuterium-labeling study has demonstrated
that the stereochemistry of oxetane (d2)-2 formation in the
photodenitrogenation of diazene (d2)-1 depends on the spin state
of the intermediary 2-spiroepoxy-1,3-cyclopentanediyl diradical
1,3-DR. The singlet diradical S-1,3-DR is exclusively generated
in the direct photolysis and rearranges stereoselectively to the
oxetane (d2)-2 with complete retention of the initial configuration.
Alternatively, in the Ph2CO-sensitized photodenitrogenation, the
triplet diradical T-1,3-DR intervenes, which extensively random-
izes its deuterium label in the oxetane (d2)-2 product through the
relatively long-lived triplet 1,4-diradical T-1,4-DR.
also the initial intermediate from the singlet-excited azo chro-
mophore S[anti(d2)-1]*. If epoxide CO bond cleavage of the
anti(d2)-S-DZ species were faster than denitrogenation, the
nitrogen-containing 1,6-diradical N,O-DR would result. Subse-
quent N2 loss to the nitrogen-free S-1,4-DR diradical and closure
to the oxetane 2 cannot apply, because in that case an extensively
randomized deuterium distribution would be expected. The
alternative option of the SH2 denitrogenation9 of N,O-DR would
afford endo(d2)-2, which again contradicts the experimental facts,
that is, 100% retention in the direct photolysis (Table 1, entries
1 and 2). Consequently, the singlet diazenyl diradical anti(d2)-
S-DZ expels preferentially N2 to produce the anti(d2)-S-1,3-DR
singlet diradical, which by suprafacial oxygen [1,2] shift10,11
generates the observed exo(d2)-2 oxetane with complete retention
of the initial configuration (stereochemical memory).
In sharp contrast to the singlet pathway (direct irradiation), an
extensively randomized distribution of the deuterium atoms was
observed in the oxetane (d2)-2 for the triplet process (Ph2CO
sensitization) of anti(d2)-1 and syn(d2)-1 (Table 1, entries 3 and
4). From the triplet-excited azoalkane [anti(d2)-1]*, again first
the diazenyl diradical anti(d2)-T-DZ is generated, which has
Acknowledgment. The work in Osaka was supported in part by a
grant-in-aid for Scientific Research on Priority Areas “Molecular Physical
Chemistry” from the Ministry of Education, Science, Sports, and Culture
of Japan, the work in Wu¨rzburg by the Volkswagen Foundation, Deutsche
Forschunggemeinschaft, and Founds der Chemischen Industrie. We thank
Mrs. T. Muneishi at Analytical Division of Faculty of Engineering, Osaka
University, for measuring the 600-MHz NMR spectra. M.A. is grateful
to the Alexander-von-Humboldt Foundation for a postdoctoral fellowship
(1997-1998).
T
(9) (a) Roth, W. R.; Martin, M. Tetrahedron Lett. 1967, 4695. (b) Adam,
W.; Garc´ıa, H.; Mart´ı, V.; Moorthy, J. N. J. Am. Chem. Soc. 1999, 121, 9475.
(10) The unique CO bond cleavage may be rationalized by a stereoelectronic
effect, in which the radical p orbitals and the epoxide CO bond are arranged
in parallel, as suggested by semiempirical (AM1) calculation (ref 3); see, Giese,
B.; Dupuis, J.; Gro¨ninger, K.; Hasskerl, T.; Nix, M.; Witzel, T. In Substituent
Effects in Radical Chemistry; Viehe, H. G.; Janousek, Z.; Mere´nyi, R., Eds.;
NATO Advanced Study Institute Series C189; D. Reidel: Dordrecht, 1986;
pp 283-296.
Supporting Information Available: Synthetic scheme for the azo-
alkanes anti(d2)-1 and syn(d2)-1; NOE data for the assignment of Ha-d
protons (PDF). This material is available free of charge via the Internet
at http://pubs.acs.org.
(11) The observed suprafacial process (retention of the configuration) is
energetically favored, since a highly strained transition state would be expected
for the antarafacial shift.
JA000656S