Stepwise cycloreversion of oxetane radical cations with initial C-O bond cleavage

Article abstract of DOI:10.1021/ja025697y

2,4,6-Triaryl(thia)pyrylium salts have been used as electron-transfer photosensitizers for the cycloreversion of the oxetane ring system. The radical cation of 2,3-diphenyl-4-hydroxymethyloxetane (1) undergoes stepwise splitting via initial O-C₂ cleavage. Spin and charge in the resulting intermediate are located in the oxygen and carbon atoms, respectively. Subsequent intramolecular nucleophilic attack produces 2,3-diphenyl-4-hydroxytetrahydrofuran (4a). Formation of this product occurs in the submicrosecond time scale, competing with C₃-C₄ cleavage to the detectable (λ_max = 470 nm) trans-stilbene radical cation. Copyright

Full text of DOI:10.1021/ja025697y

Published on Web 05/15/2002
Stepwise Cycloreversion of Oxetane Radical Cations with Initial C-O Bond
Cleavage
Miguel A. Miranda* and M. Angeles Izquierdo
Departamento de Qu´ımica/Instituto de Tecnolog´ıa Qu´ımica UPV- CSIC, UniVersidad Polite´cnica de Valencia,
Camino Vera s/n, Apdo. 22012, 46022, Valencia, Spain
Received January 24, 2002
Scheme 1
Cycloreversion (CR) of oxetanes by photoinduced electron
transfer (PET) has attracted considerable interest, as this process
appears to be involved in the enzymatic repair of the (6-4)
photoproducts of the DNA dipyrimidine sites by photolyase.1
Although photoreducing sensitization is the mode usually operating
in biological systems,2 the photooxidative approach is considered
to be relevant for DNA repair therapies.3 Further, CR of oxetane
radical cations is known to occur in the gas phase (MS), where the
molecular ions are hardly detected due to their instability under
the ionization conditions.4
Chart 1
Nonconcerted, two-step mechanisms have been proposed for the
splitting of oxetane radical cations based on the loss of stereo-
chemistry of the products obtained under PET conditions.5 Although
a concerted route cannot be ruled out from these results, the stepwise
mechanism appears more likely due to the presence of a heteroatom
(Scheme 1).6
Theoretical calculations on the cationic ET cycloreversion of
oxetanes point to an initial C-C bond breaking (Scheme 1, pathway
b1) according to free energy changes estimated for the gas-phase
reaction.4,6 By contrast, the anionic ET cycloreversion of (6-4)
photoproducts of DNA dipyrimidine sites appears to start with C-O
cleavage.6 This possibility (Scheme 1, pathway a1) should also be
considered in the cationic CR.
Formation of the neutral carbonyl/olefin units requires a back
electron transfer (BET) step. In nonconcerted mechanisms, BET
could take place either prior to (Scheme 1, pathways a2 and b2) or
after the second bond breaking process (Scheme 1, pathway d).
Time-resolved studies have provided no direct evidence for the
concerted versus stepwise nature of ring splitting.7
Thus, the mechanism of PET cycloreversion of oxetanes is still
unclear. In particular, the following points have to be clarified: (a)
the involvement of a concerted or stepwise mechanism, (b) the
nature of the first bond breaking, and (c) the location of spin and
charge in the intermediates.
Intramolecular nucleophilic trapping of the carbocationic site has
been reported as an elegant tool for elucidating the intermediates
involved in several PET reactions, such as the photooxygenation
of diphenylalkenes8 or the denitrogenation of bicyclic azoalkanes.9
In the present work, a similar strategy has been employed to gain
conclusive mechanistic insight into the CR of oxetane radical
cations. Thus, the hydroxymethyl-substituted oxetane 1 has been
submitted to steady-state and time-resolved photolysis studies, using
triaryl(thia)pyrylium salts 2a-c10 as electron-transfer sensitizers
(Chart 1). The results clearly show that the process occurs through
a nonconcerted mechanism involving initial cleavage of the C-O
bond (Scheme 1, pathway a1 + a3 + d).
Compound 1 was prepared by Paterno-Bu¨chi photoaddition of
cinnamyl alcohol to benzaldehyde.11 Irradiation of 1 in the presence
of 2a-c produced stilbene 3, 2,5-dihydroxy-1,4-dioxane (the dimer
of hydroxyacetaldehyde), and a new product whose structure was
safely assigned as 4a, based on NMR and MS data.12 As expected
from involvement of the triplet sensitizer,7 the highest conversion
was achieved with 2b (Table 1), where intersystem crossing is most
efficient.
From the 1H and 13C NMR data of 4a (including coupling
constants and NOE effects) it is clear that the initial trans-diphenyl
arrangement of 1 is also present in the photoproduct; thus
isomerization to 4b via C2-C3 rotation (Scheme 2) does not take
place to a significant extent.
Isolation of 3 indicates that CR involves cleavage of the O-C2
and C3-C4 bonds. Not even traces of cinnamyl alcohol or
benzaldehyde (arising from C2-C3/C4-O splitting) were detected.
Formation of 4a is explained as outlined in Scheme 2. Initial
cleavage along pathway a1 gives rise to a distonic radical cation,
where spin and charge are located in the oxygen and the C2 atoms,
respectively. This intermediate can undergo either further C-C
scission (pathway a3), to give 3, or intramolecular nucleophilic
attack leading to 4a. The latter process must occur somewhat slower,
as indicated by the 3/4a product ratio (see Table 1), which is
consistently about 2.7:1. At the present stage, it is not possible to
exclude the possibility that cleavage of the O-C2 and C3-C4 bonds
takes place to some extent via a concerted mechanism (Scheme 1,
pathway c), which might compete with the nonconcerted one.
In principle, ring splitting of 1+• to give 3 might also have
occurred via initial C3-C4 bond cleavage. However, in this case,
intramolecular nucleophilic trapping should lead to 5 (Scheme 2).
* Corresponding author. E-mail: mmiranda@qim.upv.es.
9
6532
J. AM. CHEM. SOC. 2002, 124, 6532-6533
10.1021/ja025697y CCC: $22.00 © 2002 American Chemical Society

C O M M U N I C A T I O N S
Table 1. Product Yields in the PET CR of 1 Sensitized by 2a-ca
c
c
c
sensitizerb
unreacted 1
3
4a
2a
2b
2c
35
52
48
73
32
17
27
16
a Oxetane 1, 6 × 10-2 M; sensitizer, 10-3 M; CDCl3; argon; Luzchem
multilamp photoreactor, 8W lamps (4×) with emission maximum at 350
nm; irradiation time 20 min. b Control samples without sensitizer were
irradiated in parallel to ensure that reaction occurs only in the presence of
1
sensitizer. c Reaction followed by H NMR.
Scheme 2
Figure 1. Transient spectrum obtained from LFP (λ ) 355 nm) of 1 (1.25
× 10-3 M) and 2b (1.2 × 10-4 M) in acetonitrile under nitrogen. Spectra
recorded 1, 2, and 3.5 µs after the laser pulse. Inset: Growth and decay of
the 470 nm band.
Acknowledgment. Financial support by the Spanish MCYT
(Grant BQU2001-2725) is gratefully acknowledged. We also thank
Dr. S. Gil for the NOE experiments.
Supporting Information Available: Table with relevant NMR data
of 4a; detailed reaction mechanism; 1H NMR monitoring of the reaction
of 1 photosensitized by 2b; transient absorption spectra obtained after
LFP of 1 + 2b, 1′ + 2b, and 1 + 2a; and growth and decay of the
relevant bands at 470, 550, and 580 nm (PDF). This material is available
free of charge via the Internet at http://pubs.acs.org.
As this product was not detected in the reaction mixture, C3-C4
cleavage of the oxetane radical cation (pathway b1) can be safely
ruled out.
References
(1) (a) Prakash, G.; Falvey D. E. J. Am. Chem. Soc. 1995, 117, 11375-11376.
(b) Kim, S. T.; Malhotra, K.; Smith, C. A.; Taylor, J. S.; Sancar, A. J.
Biol. Chem. 1994, 269, 8535-8540. (c) Zhao, X.; Liu, J.; Hsu, D. S.;
Zhao, S; Taylor, J. S.; Sancar, A. J. Biol. Chem. 1997, 272, 32580-
32590.
Previously, we have shown7 that laser flash photolysis (LFP) of
2,3-diphenyl-4-methyloxetane (1′) at 355 nm, in the presence of
pyrylium salts, leads to 3+• (absorption maxima at 470 nm).13 This
indicates that BET, under the employed reaction conditions, must
occur in the last step (Scheme 1, pathway d); hence biradicals are
not direct precursors of the neutral carbonyl/olefin pair. We have
now applied the time-resolved LFP techniques to the PET cyclo-
reversion of 1, in order gain further insight into the reaction
mechanism.
(2) (a) Heelis, P. F.; Hartman, R. F.; Rose, S. D. J. Photochem. Photobiol. A
1996, 95, 89-98. (b) Scannell, M. P.; Fenick, D. J.; Yeh, S. R.; Falvey,
D. E. J. Am. Chem. Soc. 1997, 119, 1971-1977. (c) Yeh, S. R.; Falvey,
D. E. J. Am. Chem. Soc. 1991, 113, 8557-8558.
(3) (a) Dandliker, P. J.; Holmlin, R. E.; Barton, J. K. Science 1997, 275,
1465-1468. (b) Pac, C.; Miyamoto, I.; Masaki, Y.; Ferusho, S.; Yanagida,
S.; Ohno, T.; Yoshimura, A. Photochem. Photobiol. 1990, 52, 973-979.
(4) (4) (a) Jones, G., II; McDonnell, L. P. Org. Mass Spectrom. 1975, 10,
1-4. (b) Virtanen, P. O. I.; Karjalainen, A.; Ruotsalainen, H. Suomen
Kemi. B 1970, 43, 219-223.
(5) Nakabayashi, K.; Kojima, J. I.; Tanabe, K.; Yasuda, M.; Shima, K. Bull.
Chem. Soc. Jpn. 1989, 62, 96-101.
Thus, LFP (355 nm) of 1 in the presence of 2a-c gave rise to
an intense signal centered around 470 nm assigned to 3+•. This is
shown in Figure 1 for the case of 2b. Besides, some contribution
of the (thia)pyrylium triplet absorption is also present in the transient
spectra between 470 and 560 nm. Another band peaking at 550
nm was also clearly observed in the LFP experiments performed
with 2a and 2c as photosensitizers. This band was safely assigned
to the pyranyl radical resulting from reduction of 2a,c.14 As the
thiapyranyl radical does not absorb in the range 500-600 nm,15 it
does not appear in Figure 1. It was remarkable that formation of
3+• was not “instantaneous”. The inset of Figure 1 shows a growth
of the 470 nm band in the submicrosecond time scale, corresponding
to generation of 3+• from its undetectable precursor, the radical
cation from cleavage of the O-C2 bond. The rate constant estimated
for this growth was 107 s-1. These data, together with the 3/4a
product ratio, allow a rough estimation of the rate constant of the
intramolecular nucleophilic attack, which must be about 2.7 × 106
(6) Wang, Y.; Gaspar, P. P.; Taylor, J. S. J. Am. Chem. Soc. 2000, 122, 5510-
5519.
(7) Miranda, M. A.; Izquierdo, M. A.; Galindo, F. Org. Lett. 2001, 3, 1965-
1967.
(8) Jiang, Z. Q.; Foote,C. S. Tetrahedron Lett. 1983, 24, 461-464.
(9) Adam, W.; Sendelbach, J. J. Org. Chem. 1993, 58, 5316-5322.
(10) (a) Miranda, M. A.; Garc´ıa, H. Chem. ReV. 1994, 94, 1063-1089. (b)
Martiny, M.; Steckhan, E.; Esch T. Chem. Ber. 1993, 126, 1671-1682.
(11) Similar reactions have been previously reported between benzaldehyde
and â-methylstyrene or a silyl derivative of cinnamyl alcohol: Fleming,
S. A.; Gao, J. J. Tetrahedron Lett. 1997, 38, 5407-5410.
(12) Molecular ion: m/z 240. Important fragment ions: m/z 180, 134, 107,
92.
(13) (a) Spada, L. T.; Lewis, F. D.; Bedell, A. M.; Dykstra, R. E.; Elbert, J.
E.; Gould, I. R.; Farid, S. J. Am. Chem. Soc. 1990, 112, 8055-8064. (b)
Akaba, R.; Sakuragi, H.; Tokumaru, K. Chem. Phys. Lett. 1990, 174, 80-
84.
(14) (a) Kuriyama, Y.; Arai, T.; Sakuragi, H.; Tokumaru, K. Chem. Lett. 1988,
1193-1196. (b) Akaba, R.; Oshima, K.; Kawai, Y.; Obuchi, Y.; Negishi,
A.; Sakuragi, H.; Tokumaru, K. Tetrahedron Lett. 1991, 32, 109-112.
(15) Akaba, R.; Kamata, M.; Kuriyama, Y.; Sakuragi, H.; Tokumaru, K. Chem.
Lett. 1993, 1157-1158.
s-1
.
Summarizing, the results obtained here in the steady-state and
time-resolved experiments support stepwise CR of 1+• via initial
O-C2 cleavage. Current studies are aimed at disclosing the
generality of this finding.
JA025697Y
9
J. AM. CHEM. SOC. VOL. 124, NO. 23, 2002 6533