A. Vidal et al. / Tetrahedron Letters 54 (2013) 2536–2537
2537
Table 1
Calculated energies of different steps in the photodimerization of 5-phenyluracil (1)
and 6-phenyluracil (2)a
Substrate
1
Geometry
Transition stateb
Biradicalb (T1)
Product
syn
anti
ꢀ4985.53
ꢀ4985.48
0.05
ꢀ5009.91
ꢀ5008.48
1.43
ꢀ5012.61
ꢀ5019.84
7.23
D
E
2
syn
anti
ꢀ4991.53
ꢀ4991.65
0.12
ꢀ5024.53
ꢀ5023.83
0.70
ꢀ5022.08
ꢀ5028.32
6.24
D
E
a
Total energy is given in kcal molꢀ1. The lower energy at each step is in bold.
The value given is for the most energetically stable approach (transition state)
b
or conformer (biradical).
Figure 2. ORTEP plots (30% probability) of the X-ray structures of the photodimers,
showing the HH cis–syn–cis configuration of (left) and the HH cis–anti–cis
configuration of 4 (right). Solvent molecules have been removed for clarity.
3
the origins of the stereoselectivity may lie in self-association
effects in the solution, via some kind of stacking phenomenon;
such an effect was previously suggested to operate in the photodi-
merization of tetramethyluracil.15
In summary, we have prepared and characterized two new
pyrimidinedione photodimers. The molecular structure of one of
these, compound 3, is of particular interest being the most-easily
synthesized cis–syn–cis adduct to date and offering a multiple
hydrogen bond donor/acceptor functional group montage. Since
the pendant aryl groups may act as chromophores and points of
attachment to larger scaffolds, this system may inspire new inves-
tigations of cyclobutane photodimers as templates for molecular
recognition.
References and notes
Figure 3. Views of the cis–syn–cis structure of compound 3 from above (left) and
side-on (right), highlighting the orientation of the heterocyclic systems. Phenyl
rings are removed for clarity.
1. Cadet, J.; Vigny, P. In The Photochemistry of Nucleic Acids; Morrison, H., Ed.;
Wiley: New York, 1990.
2. Shetlar, M. D.; Basus, V. J. Photochem. Photobiol. 2011, 87, 82–102.
3. (a) Fenick, D. J.; Carr, H. S.; Falvey, D. E. J. Org. Chem. 1995, 60, 624–631; (b)
Kosmoski, J. V.; Smerdon, M. J. Biochemistry 1999, 38, 9485–9494; (c) Carell, T.;
Epple, R.; Gramlich, V. Helv. Chim. Acta 1997, 80, 2191–2203.
4. Skene, W. G.; Berl, V.; Risler, H.; Khoury, R.; Lehn, J.-M. Org. Biol. Chem. 2006, 4,
3652–3663.
5. (a) Hirst, S. C.; Hamilton, A. D. Tetrahedron Lett. 1990, 31, 2401–2404; (b) Park,
T. K.; Schroeder, J.; Rebek, J., Jr. J. Am. Chem. Soc. 1991, 113, 5125–5127; (c)
Takase, M.; Inouye, M. J. Org. Chem. 2003, 68, 1134–1137.
6. Pauvert, M.; Laine, P.; Jonas, M.; Wiest, O. J. Org. Chem. 2004, 69, 543–548.
7. Khattak, M. N.; Wang, S. Y. Tetrahedron 1972, 28, 945–957.
mixtures.13 The puckered cyclobutane has a mean torsion angle of
21° and the heterocycles are flattened half-chairs, twisted by about
30° from superposition (Fig. 3, left). This makes for a closer align-
0
ment of the hydrogen-bond donor–acceptor pairs N1–H/C2 @O,
0
0
/C2@O N3 –H, and N3–H/C4 @O of neighboring heterocycles
(Fig. 3, right).
The photosensitizing effect of acetone implicates the triplet
excited state of 1 or 2, and formation of the first bond should occur
between the non-arylated positions to provide a stabilized biradi-
cal intermediate, leading to a HH regioisomer in each case. How-
ever it was not clear why one adduct should be anti while the
other was syn. We performed semi-empirical calculations to verify
the feasibility of formation of the observed structures.14 For each
substrate 1 and 2 the transition state (TS), biradical intermediate
and dimer product energies for both syn and anti structures were
determined (Table 1). For 1, virtually no difference was apparent
between the TS energies of the syn and anti approaches, although
the syn biradical was 1.43 kcal molꢀ1 more stable. While the anti
product is more stable than the syn, there is no energy barrier to
the formation of the latter. For 2, even less difference was evident
between the energies of the syn and anti biradicals, with syn being
favored by only 0.70 kcal molꢀ1 and the anti TS being marginally
more favorable. Here again, there is no clear-cut argument in favor
of either approach mode; the only significant energy difference is
the greater stability of the anti dimer product. It thus appears that
a reaction pathway to either syn or anti geometry is available, so
8. Shim, S. C.; Lee, S. H. Photochem. Photobiol. 1979, 29, 1035–1038.
9. (a) Birnbaum, G. I. Acta Crystallogr., B. 1972, 28, 1248–1254; (b) Birnbaum, G. I.;
Dunston, J. M.; Szabo, A. G. Tetrahedron Lett. 1971, 947–950.
10. Gauzy, C.; Saby, B.; Pereira, E.; Faure, S.; Aitken, D. J. Synlett 2006, 1394–1398.
11. Compound 3: White solid, mp >360 °C (DMSO). 1H NMR (300 MHz DMSO-d6) d
4.39 (2H, s), 7.21 (10H, m), 7.99 (2H, s), 10.27 (2H, s). 13C NMR (75 MHz DMSO-
d6) d 58.2, 54.5, 127.5, 127.8, 128.0, 135.8, 152.4, 169.4. Compound 4: White
solid, mp 279 °C (DMF). 1H NMR (300 MHz DMSO-d6) d 4.19 (2H, s), 7.407.50
(10H, m), 7.87 (2H, s), 10.48 (2H, s). 13C NMR (75 MHz DMSO-d6) d 39.1, 64.8,
127.0, 128.0, 128.4, 137.4, 151.5, 167.7.
12. CCDC 916146 and CCDC 916145 contain the crystallographic data for
compounds 3 and 4, respectively. These data can be obtained free of charge
13. The cis–syn–cis uracil photodimer has been characterized: see (a) Adman, E.;
Gordon, M. P.; Jensen, L. H. Chem. Commun. 1968, 1019–1020; (b) Adman, E.;
Jensen, L. H. Acta Crystallogr., B. 1970, 26, 1326–1334.
14. Semi-empirical calculations were performed with Hyperchem 5.1 using the
PM3 method with the Unrestricted Hartree-Fock wave function. The Polak-
Ribiere algorithm with conjugate gradient calculations was adopted for
geometry optimizations. For the dimerization process, one molecule was
considered in the first excited triplet state, and the other was considered in the
singlet ground state. Transition states were searched by quadratic synchronous
transit algorithm. Radical intermediates were treated in the first triplet state.
15. Otten, J. G.; Yeh, C. S.; Byrn, S.; Morrison, H. J. Am. Chem. Soc. 1977, 99, 6353–
6359.