Comparison of the singlet oxygen ene reactions of cyclic versus acyclic β,γ-unsaturated ketones: An experimental and computational study Dedicated to Professor Waldemar Adam on the occasion of his 75th birthday

Article abstract of DOI:10.1016/j.tetlet.2013.03.099

The photooxygenation of two β,γ-unsaturated ketones was studied by experimental and computational methods: 5-methyl-hex-4-en-2-one (1) and cyclohex-3-en-1-one (6) as model compounds for acyclic versus cyclic deconjugated enones. The open-chain substrate delivered a 1:1 mixture of regioisomers 2a,b following the established cis-selectivity model whereas the cyclic substrate reacts with ¹O₂ to give preferentially the conjugated product 7. This effect is in agreement with the mechanistic two-stage no-intermediate model and on a computational level corresponds to a regioselectivity control following the steepest decent pathway from the corresponding transition stages in a valley ridge potential energy surface region.

Full text of DOI:10.1016/j.tetlet.2013.03.099

Tetrahedron Letters 54 (2013) 2938–2941
Contents lists available at SciVerse ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Comparison of the singlet oxygen ene reactions of cyclic versus acyclic
b,c-unsaturated ketones: an experimental and computational study
⇑
Axel G. Griesbeck , Bernd Goldfuss, Matthias Leven, Alan de Kiff
Department of Chemistry, University of Cologne, Greinstr. 4, 50939 Köln, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
The photooxygenation of two b,c-unsaturated ketones was studied by experimental and computational
Received 26 January 2013
Revised 14 March 2013
Accepted 22 March 2013
Available online 6 April 2013
methods: 5-methyl-hex-4-en-2-one (1) and cyclohex-3-en-1-one (6) as model compounds for acyclic
versus cyclic deconjugated enones. The open-chain substrate delivered a 1:1 mixture of regioisomers
2a,b following the established cis-selectivity model whereas the cyclic substrate reacts with 1O2 to give
preferentially the conjugated product 7. This effect is in agreement with the mechanistic two-stage
no-intermediate model and on a computational level corresponds to a regioselectivity control following
the steepest decent pathway from the corresponding transition stages in a valley ridge potential energy
surface region.
Dedicated to Professor Waldemar Adam on
the occasion of his 75th birthday
Ó 2013 Elsevier Ltd. All rights reserved.
Keywords:
Ene reaction
Singlet oxygen
Deconjugated ketone substrates
Computational analysis
The singlet oxygen (1O2) ene reaction was described for the first
time by Günther O. Schenck (therefore often named Schenck reac-
tion) in 1943.1 In this reaction, 1O2 attacks one carbon of a carbon–
carbon double bond with transfer of an allylic hydrogen with
simultaneous allylic shift of the double bond.2 Alternatively, an
allylic silyl group from a silyl enolether in case of the silyl-ene
reaction can be transferred. As products of this reaction, allylic
hydroperoxides are formed. Since the first examples of 1O2 ene
reactions were reported, this transformation has found numerous
applications in modern organic synthesis.3 Furthermore, the mech-
anistic course of this reaction has been a matter of debate for
several decades and still is not completely understood. This situa-
tion is largely due to several selectivity features that could not be
incorporated into one mechanistic picture (vide infra). Several
mechanisms have been postulated for this reaction with concerted
or ‘two-stage no-intermediate’ mechanisms,4 as well as two-step
processes involving 1,4-biradical,5 1,4-zwitterion,6 perepoxide,7
or dioxetane intermediates.8 Scheme 1 represents a typical text-
book presentation involving a reversible primary interaction fol-
lowed by a short-lived intermediate that collapses to the product.
Orfanopoulos and Stephenson performed classical and elegant
inter and intramolecular isotope effect experiments with isotopi-
cally labeled tetramethylethylenes that provided strong evidence
for the perepoxide intermediate.9 Also, the small negative activation
enthalpies and highly negative activation entropies observed for the
singlet oxygen ene reaction from kinetic measurements have shown
that the reaction of 1O2 with electron-rich olefins proceeds
103 times slower than the diffusion rate which accounts for the
presence of non-productive encounters between 1O2 and the alkene
favoring the participation of a reversibly formed exciplex as
intermediate.10
The regiochemistry of the ene reaction between 1O2 and
substrates with multiple sites for allylic hydrogen transfer was
extensively studied and several general effects can predict the
regioselective introduction of the hydroperoxy group:2 (a) the
cis-effect11 (syn-effect). In the reaction of 1O2 with trisubstituted
alkenes12 or enol ethers,13 the allylic hydrogen atoms on the
more substituted side of the double bond are more reactive for
H-abstraction by 1O2 (see Scheme 2 indicated by hydrogens
O2
O
1O2
O
HOO
H
H
Scheme 1. Ene reaction with singlet oxygen––the mechanistic basis.
CH2
CH3
H
OOH
1O2
H
+
R
R
R
OOH
⇑
Corresponding author. Tel.: +49 221 470 3083; fax: +49 221 470 5057.
E-mail address: griesbeck@uni-koeln.de (A.G. Griesbeck).
Scheme 2. Singlet oxygen ene reaction––the cis effect.
0040-4039/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.tetlet.2013.03.099

A. G. Griesbeck et al. / Tetrahedron Letters 54 (2013) 2938–2941
2939
indicated in blue); (b) the gem-effect,14 that leads to highly selec-
tive abstraction of an allylic hydrogen atom from a substituent in
Apart from discussions in the literature about the first interac-
tions of singlet oxygen and the alkene substrate on the reaction
coordinate, there is one characteristic profile for the abstraction
of competing hydrogen atoms in allylic positions3–5 which is also
found in the case of the unsaturated ketones 1 and 6. The reaction
coordinate is running down from the substrates or the preformed
adducts and splits into two pathways which are leading to the
products of type A and B (Fig. 1). This splitting does not involve
any activation barriers so that regioselectivity is induced by the
steepest descent pathways, resulting from the local shape of poten-
tial energy surfaces and related dynamic effects.4
a-position of an a,b-unsaturated carbonyl compound; (c) the
large-group effect,15 that leads to selective (moderate) abstraction
of an allylic hydrogen from the substituent geminal to a large
group. With the exception of hydroxyl groups, substituents in
allylic positions do not exhibit appreciable effects on the regiose-
lectivity of the 1O2 ene reaction and ca. 1:1 mixtures of secondary
and tertiary allylic hydroperoxides are formed (Scheme 2). We
have recently reported on the change in regioselectivity that can
be observed for photooxygenations in micro-emulsions16 and in
hydrogels.17
This energy profiles were analyzed by scans of the potential en-
ergy surface starting from the TS-structures leading to A and B and
along the dissociation of the cleaved carbon–hydrogen bond.
Results and discussion
In the area of open-chain enone photooxygenation, two trends
are in agreement with the regioselectivity rule of thumb described
above: low regioselectivity (consequence of the cis-effect) in
O
1O2
H
O
H
the photooxygenation of b,c-unsaturated ketone 1 (1:1 mixture
5
0
of regioisomeric hydroperoxides 2a and 2b)16 and the high
regioselectivity (resulting from the gem-effect) in the photooxy-
O
O
9
genation of
4 and its isomeric peroxyacetal 5 (Scheme 3)).18
When applying the cyclic b, -unsaturated ketone 6 as substrate,
a,b-unsaturated ketone 3 (mixture of hydroperoxide
4
0
c
the conjugated product 7 is formed as the major product in polar
and nonpolar solvents (methylene chloride, acetone, carbon tetra-
chloride) with tetraphenyl-porphyrine (TPP) as singlet oxygen sen-
sitizer. Allylic hydroperoxide 7 is not described in the chemical
literature, only mentioned in two patents where 7 is produced by
thermal methods.19 From 1H and 13C NMR the structure of 7 was
unambiguously determined. Reduction of the product mixture
with Me2S resulted in 4-hydroxycyclohexenone, well known from
the Kornblum–DeLaMare rearrangement of 1,3-cyclohexadiene-
endoperoxide.20 The mechanistic question remains whether and
why the cis-effect breaks down for 6 or a thermodynamic control
exists that prefers the formation of the more stable conjugated en-
one as a major product.
3
0
O
2
0
8
OH
O
1
0
O
O
B
- 2 . 5
7
- 2 . 0
0
- 2
- 1 . 5
- 1
OH
- 1 . 0
0
- 0 . 5
1
A
0
. 0
2
0
. 5
3
Computational studies
1O2
The mechanistic pathways of singlet oxygen (1 g–1O2) ene reac-
D
5
0
O
O
O
tions employing 1 and 6 were also investigated by quantum chem-
ical methods in order to rationalize the origin of regioselectivities.
All structure geometry optimizations were carried out with the
density functional theory using the non empirical TPSS functional
(meta GGA) developed by Staroverov et al.,21 combined with the
contracted TZVP basis set from Ahlrich et al.,22 implemented in
the program package TURBOMOLE6.3.23 The grid size for numerical
integration was set to m5.
O
H
4
0
H
10
3
0
OH
2
0
O
O
OOH
1O2
+
2b
1
0
B
O
OOH
O O
O
O
2a
2b
1
54 : 46
0
O
OH
O
O
OOH O
- 2
- 2
.
5
1O2
- 1
- 2
.
0
OH
O
- 1
.
5
0
- 1
.
0
2a
A
> 95%
1
5
4
3
- 0
.
5
0
.
0
2
HOO
0
.
5
1O2
3
HOO
O
+
O
Figure 1. Reaction path along the singlet oxygen ene reaction with b,c-unsaturated
7
8
70 : 30
6
cyclic ketone 6 and acyclic ketone 1 TPSS/TZVP. Regioselectivities only result from
the descent in the region around the TS 9 and 10 toward the products A and B. (The
attack of singlet oxygen on compound 6 leads to identical transition structures for
both possible faces.)
Scheme 3. Ene reaction with compounds 1 (acyclic deconjugated enone), 3 (acyclic
Michael system), and 6 (cyclic deconjugated enone).

2940
A. G. Griesbeck et al. / Tetrahedron Letters 54 (2013) 2938–2941
Perepoxide like structures could only be found for an exo aligned
attack of singlet oxygen on compound 6 (see Supplementary infor-
mation for details). Other perepoxide structures were found to be
no stationary points on the potential energy surfaces for both sub-
strates. There is a thermodynamic preference for the Michael-type
products in both cases. Product 2a is favored by 3.1 kcal/mol and
product 7 by 2.7 kcal/mol (including zero point corrections) but
the experiment shows that thermodynamic effects are not respon-
sible for regioselectivity. Since the regioselectivities are also not in-
duced by any activation barriers, another model is needed to
explain the regioselectivities which elucidates the control via the
descent of the particular pathways. The only stationary point in
the regions where regioselectivity is created is the transition state
(9 and 10 in Fig. 2). It is apparent that the real minimum reaction
path is split in a valley ridge inflection point close to the transition
state from recent studies.4 Scans following the steepest descents
from the stationary points 9 and 10 in the direction of both possi-
ble products of type A and B are able to represent the both reaction
pathways.
Figure 3. Transition structures 9 and 10 with depicted atom distances between
oxygen and the abstracted hydrogen in allylic positions TPSS/TZVP.
This flat area which has to be crossed on the way to product 8
could diminish the likely hood for the generation of 8 and increase
the generation of 7 since the descent toward 7 is almost vertical in
the first steps.
A closer look at the transition structures 9 and 10 gives also an
advice which reaction path is more favored. The O–H distances of
the oxygen that abstracts the hydrogen atoms in the allylic posi-
tions of the unsaturated ketone shown in Figure 3 are very similar
for the acyclic system 10 (the bond lengths are differing only by
0.1 Å). Structure 9 for the cyclic transition state is of much more
asymmetric nature. The oxygen–hydrogen distances are differing
by 0.4 Å so that a transition state arises from the cyclic geometry
that is early for the generation of 7 and a late transition state for
the generation of 8.
Potential energy profiles in higher resolution of the area around
the transition states 9 and 10 are showing markedly different prop-
erties. The descent toward the thermodynamic favored products is
steeper in both cases but there is a flat plateau like region in the
pathway toward the unfavored product B for substrate 6 (Fig. 3).
O
H
The abstraction of the
a-hydrogen by the peroxy group in 9
O
H
(Fig. 3) appears to be more favored than the abstraction of the
0.0
-0.5
-1.0
-1.5
-2.0
a
O
competing b-hydrogens, as the longer O–H distance indicates a
9
stronger O–H-attraction along the negatively curved normal mode
of saddle point 9. In other words the system starts loosing energy
a
on the reaction coordinate with a still long O–H distance. Also the
lower selectivity observed in the formation of 2a and 2b, preferring
the thermodynamically disfavored product 2b, can be explained by
the less pronounced asymmetry of the O–H distances in TS 10.
O
O
OH
O
OH
Conclusion
O
8
7
The divergent regioselectivity effects experimentally deter-
mined for the ene substrates 1 and 6, respectively, are in agree-
ment with the mechanistic two-stage no-intermediate model and
on a computational level correspond to a control mechanism fol-
lowing the steepest decent pathway from the corresponding tran-
sition stages in a valley ridge potential energy surface region.
A
B
-0.3
-0.2
-0.1
0.0
0.1
0.2
0.3
Displacement [Å]
O
O
H
Acknowledgement
O
10
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
-1.2
H
This research was supported by the Deutsche Forschungs-
gemeinschaft (DFG).
Supplementary data
O
OH
O
Supplementary data (The detailed energy profiles for the reac-
tion of singlet oxygen with 1 and 6, respectively, are described.)
associated with this article can be found, in the online version, at
http://dx.doi.org/10.1016/j.tetlet.2013.03.099.
OH
O
O
2a
2b
References and notes
A
1. Schenck, G. O. Naturwissenschaften 1948, 35, 28–29.
B
2. (a) Clennan, E. L. Tetrahedron 2005, 61, 6665–6691; (b) Clennan, E. L.
Tetrahedron 2000, 56, 9151–9179.
3. (a) Prein, M.; Adam, W. Angew. Chem., Int. Ed. 1996, 35, 477–494; (b) Alsters, P.
L.; Jary, W.; Nardello-Rataj, V.; Aubry, J.-M. Org. Process Res. Dev. 2010, 14, 259–
262.
-0.3
-0.2
-0.1
0.0
0.1
0.2
Displacement [Å]
0.3
Figure 2. Scans following the steepest descent from transition structures 9 and 10
toward the product regions TPSS/TZVP (resolution higher than in Figure 1). The
pathway from 9 to 8 shows a markedly flat shaped region around the transition
state.
4. (a) Singleton, D. A.; Hang, C.; Szymanski, M. J.; Meyer, M. P.; Leach, A. G.;
Kuwata, K. T.; Chen, J. S.; Greer, A.; Foote, C. S.; Houk, K. N. J. Am. Chem. Soc.
2003, 125, 1319–1328; (b) Orfanopoulos, M.; Stephenson, L. M. J. Am. Chem. Soc.
1980, 102, 1417–1418; (c) Grdina, M. B.; Orfanopoulos, M.; Stephenson, L. M. J.

A. G. Griesbeck et al. / Tetrahedron Letters 54 (2013) 2938–2941
2941
Am. Chem. Soc. 1979, 101, 3111–3112; (d) Leach, A. G.; Houk, K. N. Chem.
Commun. 2002, 1243–1255; (e) Singleton, D. A.; Hang, C.; Szymanski, M. J.;
Greenwald, E. E. J. Am. Chem. Soc. 2003, 125, 1176–1177; (f) Ess, D. H.; Wheeler,
S. E.; Iafe, R. G.; Xu, L.; Celebi-Olcum, N.; Houk, K. N. Angew. Chem., Int. Ed. 2008,
47, 7592–7601.
14. (a) Adam, W.; Griesbeck, A. G. Synthesis 1986, 1050–1052; (b) Orfanopoulos,
M.; Foote, C. S. Tetrahedron Lett. 1985, 26, 5991–5994.
15. (a) Clennan, E. L.; Chen, X. J. Am. Chem. Soc. 1989, 111, 5787–5792; (b)
Orfanopoulos, M.;Stratakis, M.;Elemis, Y. J. Am. Chem. Soc. 1990, 112, 6417–6419.
16. Griesbeck, A. G.; Cho, M. Tetrahedron Lett. 2009, 50, 121–123.
17. Griesbeck, A. G.; Uhlig, J.; Sottmann, T.; Belkoura, L.; Strey, R. E. Chem. Eur. J.
2012, 18, 16161–16165.
5. Maranzana, A.; Canepa, C.; Ghigo, G.; Tonachini, G. Eur. J. Org. Chem. 2005,
3643–3649.
6. Jefford, C. W.; Kohmoto, S.; Boukouvalas, J.; Burger, U. J. Am. Chem. Soc. 1983,
105, 6498–6500.
7. Poon, T. H. W.; Pringle, K.; Foote, C. S. J. Am. Chem. Soc. 1995, 117, 7611–7618.
8. Gollnick, K.; Kuhn, H.J. In Singlet Oxygen Wasserman, H.H.; Murray, R.W., Eds.;
Academic press: New York, 1979, pp 287–427.
9. Orfanopoulos, M.; Stephenson, L. M. J. Am. Chem. Soc. 1980, 102, 1417–1418.
10. Hurst, J. R.; Wilson, S. L.; Schuster, G. B. Tetrahedron 1985, 41, 2191–2197.
11. Orfanopoulos, M.; Gardina, M. B.; Stephenson, L. M. J. Am. Chem. Soc. 1979, 101,
275–276.
12. Rautenstrauch, V.; Thommen, W.; Schulte-Elte, K. H. Helv. Chim. Acta 1986, 69,
1638–1643.
18. These regioselectivity phenomena that are also very well in agreement with
this theoretical model are described elsewhere in detail: Griesbeck, A.G.;
Dogan, S.; Balci, M.; Leven, M.; Goldfuss, B.; de Kiff, A.; Kleczka, M.; Vollmer, M.
unpublished results.
19. (a) Nagahara, H.; Mori, S. Jpn. Kokai. Tokkyo. Koho. JP 04364138 A 19921216,
1992.; (b) Kosmacheva, T. G.; Agabekov, V. E.; Mitskevich, N. I. Vestsi Akad.
Navuk BSSR, Ser. Khim. Navuk 1981, 28–31.
20. Staben, S. T.; Linghu, X.; Toste, F. D. J. Am. Chem. Soc. 2006, 128, 12658–12659.
21. Staroverov, V. N.; Scuseria, G. E.; Tao, J.; Perdew, J. P. J. Chem. Phys. 2003, 119,
12129–12137.
22. Schäfer, A.; Horn, H.; Ahlrichs, R. J. Chem. Phys. 1992, 97, 2571–2577.
23. Ahlrichs, R.; Bär, M.; Häser, M.; Horn, H.; Kölmel, C. Chem. Phys. Lett. 1989, 162,
165–169.
13. (a) Rousseau, G.; Le Perchec, P.; Conia, J. M. Tetrahedron Lett. 1977, 18, 2517–
2520; (b) Jefford, C. W. Tetrahedron Lett. 1979, 20, 985–988.