Triplet-sensitised di-π-methane rearrangement of N-substituted 2-azabarrelenones

Article abstract of DOI:10.1039/c8cc08704k

When irradiated at λ = 366 nm or at λ = 420 nm in the presence of an appropriate sensitiser the title compounds underwent a di-π-methane rearrangement which led to the formation of tricyclic azasemibullvalenones (2a,2a¹,2b,4a-tetrahydroazacyclo

Full text of DOI:10.1039/c8cc08704k

ChemComm
View Article Online
View Journal | View Issue
COMMUNICATION
Triplet-sensitised di-p-methane rearrangement
of N-substituted 2-azabarrelenones†
Cite this: Chem. Commun., 2019,
55, 302
Andreas Troster and Thorsten Bach
*
¨
Received 31st October 2018,
Accepted 30th November 2018
DOI: 10.1039/c8cc08704k
rsc.li/chemcomm
When irradiated at k = 366 nm or at k = 420 nm in the presence of an
It was found that catalytic amounts (50 mol%) of thioxanthone
appropriate sensitiser the title compounds underwent a di-p-methane (4) or xanthone (5) trigger a high-yielding di-p-methane rearrange-
rearrangement which led to the formation of tricyclic azasemibull- ment reaction of compounds 1 that leads to otherwise inaccessible
valenones (2a,2a1,2b,4a-tetrahydroazacyclopropa[cd]pentalenones) in compounds with a unique tricyclic skeleton. The reaction can be
yields of 63–87%.
performed with visible light (l = 420 nm) or by UV-A irradiation
(l = 366 nm) depending on the sensitiser.
2-Azabarrelenones (1)1 represent a compound class in which
The synthesis of the starting materials was performed in
one of the ethyne bridges in barrelene (2) is replaced by an analogy to a previously described method1a and is shown in
amide (–NRCO–) fragment (Fig. 1). The photochemistry Scheme 1 for the N-benzyl derivative 1a (for the synthesis
of barrelene and its substituted analogues was explored by of other 2-azabarrelenones, see the ESI†). The Diels–Alder
Zimmerman and co-workers who discovered the di-p-methane reaction of N-benzylpyridone (6a) and maleic anhydride (7)
rearrangement of barrelene to semibullvalene.2 The reaction produced an intermediate 8a (diastereomeric ratio d.r. = 93/7)
is typically performed in solvent mixtures containing acetone to which water was added for hydrolysis. The mixture was
and was shown to occur by triplet sensitisation.3,4 The first refluxed for four hours which delivered the corresponding
di-p-methane rearrangement of a benzo-2-azabarrelenone was dicarboxylic acid 9a. The latter compound was purified by
reported for compound 3 by Paquette and Meisinger in 1970.5 recrystallization to yield exclusively a single diastereoisomer
Additional work was performed on the reaction of related (presumably the endo-product7). In the synthesis of some other
benzo-2-azabarrelenones which included variations of the 2-azabarrelenones (vide infra) a mixture of diastereoisomers
substitution pattern at the arene and at the b,g-unsaturated was taken into the final Kolbe decarboxylation step. The
lactam.6 However, the photochemistry of the parent 2-azabarre- electrolysis was performed in an undivided cell (Winkler platinum
lenones has to the best of our knowledge not yet been studied electrode) at a constant voltage of 30 V (py = pyridine).8
and we disclose in this paper our preliminary work on the topic.
Initial irradiation experiments were performed with thio-
xanthone (4, triplet energy ET = 272 kJ molꢀ1 9
as sensitiser
)
(Table 1, entry 1) at l = 420 nm. The reaction in dichloromethane
Fig. 1 Structures of 2-azabarrelenone (1), barrelene (2), benzo-2-
azabarrelenone 3, thioxanthone (4), and xanthone (5).
¨
Department Chemie and Catalysis Research Center (CRC), Technische Universitat
¨
Munchen, 85747 Garching, Germany. E-mail: thorsten.bach@ch.tum.de;
Fax: +49 89 28913315; Tel: +49 89 28913330
† Electronic supplementary information (ESI) available: Synthetic procedures and
full characterization for all starting materials (1) and products (10, 11, 16). See
DOI: 10.1039/c8cc08704k
Scheme 1 Three-step synthesis of 2-azabarrelenone 1a from pyridone
6a and maleic anhydride (7).
302 | Chem. Commun., 2019, 55, 302--305
This journal is ©The Royal Society of Chemistry 2019

View Article Online
Communication
ChemComm
Table 2 Triplet sensitised di-p-methane rearrangement of various N-
substituted 2-azabarrelenones 1
remained incomplete and the products could not be obtained
in sufficient purity to determine the reaction yield. The desired
transformation was cleaner in acetonitrile and delivered the
tricyclic lactams 10a and 11a with an azasemibullvalenone
skeleton (entry 2). Compound 10a prevailed and the regioiso-
meric ratio (r.r.) remained essentially constant throughout the
optimization experiments (r.r. = 10a/11a = 74/26 to 61/39). While
it had been observed with benzo-2-azabarrelenone 3 that acidic
conditions (70% HOAc in acetone) led to an improved regio-
selectivity6a there was no reaction under our conditions if HOAc
was employed as a co-solvent (entry 3). Trifluorotoluene as
solvent led to an improved chemoselectivity (72% combined
yield) and a complete conversion was achieved after 11 hours
(entry 4).
Entrya
1
Substrate
R
tb [h]
10c [%]
11c [%]
1b
8
33
32d
2
3
1c
4.5
3
46
45
19
25
1d
It was studied whether other sensitisers result in a higher
turnover. A known iridium complex (ET = 252 kJ mol
to promote the desired reaction at l = 420 nm (entry 5) while
xanthone (ET = 310 kJ molꢀ1 12
turned out to be an efficient
sensitiser at l = 366 nm. At a loading of 50 mol%, the reaction
was complete after five hours (entry 6, 83% yield). A further
decrease of the sensitiser concentration resulted in longer
reaction times (entries 7 and 8). As a control, it was probed
whether the reaction proceeded under the conditions of
entry 8 in the absence of a sensitiser but no reaction
was observed (entry 9). The conversion of substrate 1a was
)
ꢀ1 11 failed
4
1e
1f
6
49e
71f
27e
16g
)
5
7
a
The reactions were performed in a merry-go-round apparatus10
b
employing 16 fluorescent lamps as irradiation source. Reaction time
required for full conversion. Yield of isolated product. Minor
c
d
e
f
impurities could not be fully removed. d.r. = 53/47. d.r. = 78/22.
g
Single diastereoisomer.
complete for entries 2, 4, 6–8 (Table 1) but a major by-product of a 2a,2a1,2b,4a-tetrahydro-2-azacyclopropa[cd]pentalen-1(2H)-
(vide infra) was isolated apart from the desired di-p-methane one (10a) have not been described. Although thioxanthone (4)
rearrangement products. The two diastereoisomers 10a and 11a allows for the di-p-methane rearrangement of 1a to be con-
could be readily separated and represent compound classes ducted with visible light, it turned out that the reaction was
which have not yet been extensively investigated. Analogues slow with other 2-azabarrelenones and a full conversion could
of compound 11a have been prepared previously by rearrange- not be achieved within a reasonable period of time. A more
ment reactions of Dewar benzene13 and benzvalene14 or by rapid reaction was achieved with xanthone (5) which is why the
cyclopropanation.15 Compounds with the tricyclic core skeleton conditions of entry 6 (Table 1) were used for the reaction of
other substrates 1 (Table 2).
The N-substituted 2-azabarrelenones were prepared from the
respective N-substituted pyridones as described in Scheme 1 for
compound 1a. The reaction of N-phenyl derivative 1b (entry 1)
Table 1 Optimisation of the reaction conditions for the di-p-methane
rearrangement of N-benzyl-2-azabarrelenone (1a)
was slower and less clean than the reaction of compound 1a.
The reaction mixture turned yellow upon irradiation and only
regioisomer 10b could be isolated in pure form. The other
isomer 11b was still contaminated by minor impurities. The
reactions of N-phenylethyl and N-para-methoxybenzyl (PMB)
Entrya l [nm] Sens. [mol%] t [h] Solvent 10ab [%] 11ab [%]
2-azabarrelenones 1c (entry 2) and 1d (entry 3) were slightly
faster than the reaction of 1a and equally chemoselective
(65–70% yield). The r.r. remained in the region which had been
earlier recorded for products 10a and 11a. The chiral substrate
1e (entry 4) delivered products 10e and 11e as a mixture of two
diastereoisomers (76% yield). While the diastereoselectivity
remained low in this instance, the axially chiral compound 1f
(entry 5, 87% yield) exhibited a significant preference for
a single diastereoisomer in its di-p-methane rearrangement
reaction. The isolated major regioisomer 10f showed a d.r.
of 78/22, the minor regioisomer 11f was isolated as a single
diastereoisomer. However, there was evidence from analysis of
the crude 1H NMR spectrum that the other diastereoisomer was
also formed and that the d.r. was similar to the d.r. of 10f.
1
2
3
4
5
6
7
8
9
420
420
420
420
420
366
366
366
366
4 (50)
4 (50)
4 (50)
4 (50)
[Ir]f (2)
5 (50)
5 (25)
5 (10)
—
20
20
20
11
20
5
7
12
12
CH2Cl2 n.d.c
n.d.c
17
—
MeCN
HOAcd
PhCF3
MeCN
PhCF3
PhCF3
PhCF3
PhCF3
48
e
e
—
44
28
e
e
—
53
47
45
—
30
30
24
e
e
—
—
a
The reactions were performed under the indicated conditions in a
merry-go-round apparatus10 employing 16 fluorescent lamps as irradiation
source. Yield of isolated product. The reaction was incomplete and a
b
c
d
yield could not be determined (n.d.). A 70% (v/v) solution of acetic
e
f
acid in MeCN was used. No reaction was observed. [4,40-Bis(t-butyl)-
2,20-bipyridine]bis{3,5-difluoro-2-[5-(trifluoromethyl)-2-pyridin-yl]}phenyl-
iridium hexafluorophosphate was used as the iridium catalyst [Ir].
This journal is ©The Royal Society of Chemistry 2019
Chem. Commun., 2019, 55, 302--305 | 303

View Article Online
ChemComm
Communication
Fig. 2 Structure of product diastereoisomer 11f and of the monodeuter-
ated compounds 1a-d1, 10a-d1, and 11a-d1.
The relative configuration of compound 11f was elucidated
by NOESY experiments. Strong contacts between the tert-butyl
protons and the protons at C4a and C2a1 suggest a relative
configuration (Fig. 2) in which the chirality axis between the
nitrogen atom and the aryl ring has the configuration Ra*
relative to the (2aR*,2a1S*,2bR*,4aS*)-configured tetrahydro-1-
azacyclopropa[cd]pentalen-2(1H)-one skeleton.16 Analogously,
the major diastereoisomer 10f0 was shown to exhibit a Sa*
configuration at the chirality axis relative to the configuration
of the (2aR*,2a1R*,2bR*,4aS*)-tetrahydro-2-azacyclopropa[cd]-
pentalen-1(2H)-one core. In this case, the other diastereoisomer
10f00 could also be analyzed spectroscopically and showed
complementary NOESY signals to 10f0 (see ESI†).
Scheme 2 Suggested reaction course for the triplet-sensitised
di-p-methane rearrangement of 2-azabarrelenones (1).
Scheme 3 Mechanistic hypothesis for the formation of products 16 via
3-azatetracyclo[4.2.0.02,8.05,7]octan-4-one 14.
Circumstantial evidence for the intermediacy of 1,3-
diradicals 12 and 13 is based on the isolation of the above-
mentioned side products to which structure 16 was assigned
(Scheme 3). The bicyclic lactams were obtained in varying
amounts of 5–20% in the course of the sensitised di-p-methane
rearrangement reactions (Table 2). A straightforward explanation
for their formation rests of the intermediacy of strained 3-aza-
tetracyclo[4.2.0.02,8.05,7]octan-4-ones18 14, which result from
carbon–carbon bond formation in 1,3-diradicals 12 and 13. They
represent intramolecular [2+2] photocycloaddition19,20 products
with a high ring strain, and a facile fragmentation might occur
under slightly acidic conditions. A possible reaction pathway
includes intermediate 15, which is attacked by water to form
products 16. Although the reaction is depicted as a formal SN2
type reaction, it can also be conceived to proceed via an acylimi-
nium ion which provides product 16 along an SN1 type pathway.21
Epimerization via the respective aldehyde should be possible22
and compound 16 represents the more stable diastereoisomer.
In summary, it has been shown that parent 2-azabarre-
lenones 1 undergo a di-p-methane rearrangement that leads
to the yet unexplored tricyclic lactams 10 and 11. The factors
that govern the regioselectivity and diastereoselectivity of the
reaction are not completely understood and require additional
experiments. Further transformations of the products are currently
studied in particular with regard to a cleavage of the bond between
carbon atoms 2a1 and 2b.
In order to obtain some information on the mechanistic
course of the rearrangement the monodeuterated compound
1a-d1 was prepared from the 3-deuterated pyridone 6a-d1 (85%
deuterium incorporation). Upon sensitised irradiation, the
rearrangement products 10a-d1 and 11a-d1 were obtained in a
similar ratio (77% yield, r.r. = 65/35) as the undeuterated
compounds 10a and 11a. The degree of deuterium incorpora-
tion was unchanged and the deuterium atom was shown to be
attached to carbon C4a in 10a-d1 and to C2a in 11a-d1 proving
that the C–C bond in a-position to the carbonyl group was not
cleaved in the rearrangement process. An oxadi-p-methane
rearrangement pathway3a,b in which the carbonyl group under-
goes 1,2-migration can consequently be ruled out.
Due to their low absorbance at l 4 250 nm it was not
possible to obtain luminescence spectra of compounds 1. Still,
from the slow but successful reaction with thioxanthone the
triplet energy of these compounds can be estimated to be in the
range of ca. 260–275 kJ molꢀ1. In analogy to the mechanism
of the di-p-methane rearrangement of barrelene (2)2,3 it is
proposed that the excited triplet state 1* decays to two regio-
isomeric 1,3-diradical intermediates 12 and 13 (Scheme 2). The
subsequent cleavage of the marked cyclopropane bond (in red)
determines the stereoselectivity. If there is no chiral informa-
tion the indicated bond is broken at the same rate as its mirror
carbon–carbon bond leading to the racemic products 10a–d
and 11a–d. The a-phenylethyl group exerts little selectivity
which is why the two diastereoisomers of 10e and 11e are
formed in almost equal quantities.
Financial support by the Deutsche Forschungsgemeinschaft
(DFG) is gratefully acknowledged (Ba1372/20, GRK 1626).
AT thanks the research training group (Graduiertenkolleg)
1626 ‘‘Chemical Photocatalysis’’ for a scholarship.
A notable diastereoselectivity was observed with the axially
chiral17 tert-butylphenyl compound 1e. In this case, assuming
the chirality axis to be Sa* for 12f and Ra* for 13f, the preferential
formation of products 10f0 and 11f is linked to a selective bond
cleavage at the indicated positions in intermediates 12f and 13f.
The cleaved bond is always trans to the large tert-butyl group.
Conflicts of interest
There are no conflicts to declare.
304 | Chem. Commun., 2019, 55, 302--305
This journal is ©The Royal Society of Chemistry 2019

View Article Online
Communication
ChemComm
9 A. Ilyer, A. Clay, S. Jokusch and J. Sivaguru, J. Phys. Org. Chem., 2017,
30, e3738.
10 S. Poplata and T. Bach, J. Am. Chem. Soc., 2018, 140, 3228.
11 A. Singh, K. Teegardin, M. Kelly, K. S. Prasad, S. Krishnan and
J. D. Weaver, J. Organomet. Chem., 2015, 776, 51.
12 S. L. Murov, I. Carmichael and G. L. Hug, Handbook of Photochemistry,
2nd edn, Dekker, New York, 1993, p. 91.
13 (a) L. A. Paquette, Tetrahedron Lett., 1968, 9, 2133; (b) L. A. Paquette
and G. R. Krow, J. Am. Chem. Soc., 1969, 91, 6107.
Notes and references
1 (a) R. Gompper and A. Schmidt, Angew. Chem., Int. Ed., 1980,
19, 463; (b) K. Somekawa, R. Imai, R. Furukido and S. Kumamoto,
Bull. Chem. Soc. Jpn., 1981, 54, 1112; (c) K. Matsumoto, K. Hamada,
T. Uchida and H. Yoshida, Heterocycles, 1989, 29, 21.
2 (a) H. E. Zimmerman and G. L. Grunewald, J. Am. Chem. Soc., 1966,
88, 183; (b) H. E. Zimmerman, R. W. Binkley, R. S. Givens, G. L.
Grunewald and M. A. Sherwin, J. Am. Chem. Soc., 1969, 91, 3316.
3 Reviews: (a) M. G. Banwell and D. J.-Y. D. Bon in Molecular
Rearrangements in Organic Synthesis, ed. C. M. Rojas, Wiley, Hoboken,
2015, p. 261; (b) E. Riguet and N. Hoffmann in Comprehensive
Organic Synthesis, ed. P. Knochel and G. A. Molander, Elsevier,
Amsterdam, 2nd edn, 2014, vol. 5, p. 200; (c) T. Tsuno in Handbook
of Synthetic Photochemistry, ed. A. Albini and M. Fagnoni, Wiley-VCH,
Weinheim, 2010, p. 95; (d) H. E. Zimmerman and D. Armesto,
Chem. Rev., 1996, 96, 3065; (e) S. S. Hixson, P. S. Mariano and
H. E. Zimmerman, Chem. Rev., 1973, 73, 531.
¨
14 M. Christl, E. Brunn and F. Lanzendorfer, J. Am. Chem. Soc., 1984,
106, 373.
15 Y. Ohfune, T. Demura, S. Iwama, H. Matsuda, K. Namba,
K. Shimamoto and T. Shinada, Tetrahedron Lett., 2003, 44, 5431.
16 All compounds were obtained as racemic mixtures. The relative
configuration is indicated by the R*/S* nomenclature: E. L. Eliel and
S. H. Wilen, Stereochemistry of Organic Compounds, Wiley, New York,
1994, p. 1206.
17 For recent work on photochemical atropselective reactions, see:
(a) E. Kumarasamy, A. J.-L. Ayitou, N. Vallavoju, R. Raghunathan,
A. Iyer, A. Clay, S. K. Kandappa and J. Sivaguru, Acc. Chem. Res.,
2016, 49, 2713; (b) E. Kumarasamy, R. Raghunathan, M. P. Sibi and
J. Sivaguru, Chem. Rev., 2015, 115, 11239.
4 For recent synthetic and theoretical studies on the di-p-methane
rearrangement, see: (a) O. Yahiaoui, L. F. Pasteka, B. Judeel and
T. Fallon, Angew. Chem., Int. Ed., 2018, 57, 2570; (b) X. Li, T. Liao and
L. W. Chung, J. Am. Chem. Soc., 2017, 139, 16438; (c) H. Ihmels and J. Luo,
Z. Phys. Chem., 2014, 228, 229.
ˇ
18 A. I. Meyers and P. Singh, J. Org. Chem., 1970, 35, 3022.
5 L. A. Paquette and R. H. Meisinger, Tetrahedron Lett., 1970, 11,
1479.
¨
19 Review: S. Poplata, A. Troster, Y.-Q. Zou and T. Bach, Chem. Rev.,
2016, 116, 9748.
6 (a) M. Kuzuya, M. Ishikawa, T. Okuda and H. Hart, Tetrahedron Lett.,
1979, 20, 523; (b) M. Kuzuya, E. Mano, M. Ishikawa, T. Okuda and
H. Hart, Tetrahedron Lett., 1981, 22, 1613.
7 (a) H. Tomisawa and H. Hongo, Chem. Pharm. Bull., 1970, 18, 925;
(b) H. Hongo, Chem. Pharm. Bull., 1972, 20, 226; (c) H. Tomisawa,
H. Hongo, H. Kato, K. Sato and R. Fujita, Heterocycles, 1981, 16,
1947.
8 (a) H. Westberg and H. Dauben, Tetrahedron Lett., 1968, 9,
5123; (b) A. P. Marchand and R. W. Allen, J. Org. Chem., 1975,
40, 2551; (c) R. Gompper and K. H. Etzbach, Angew. Chem.,
20 For related [2+2] photocycloaddition reactions of dihydrobarrelenes,
see: (a) H. Prinzbach, W. Eberbach and G. Philippossian, Angew.
Chem., Int. Ed., 1968, 7, 888; (b) R. S. H. Liu, Tetrahedron Lett.,
1969, 1409; (c) S. F. Nelsen and J. P. Gillespie, Tetrahedron Lett.,
1969, 5059; (d) H. Prinzbach, H. Fritz, H. Hagemann, D. Hunkler,
S. Kagabu and G. Philippossian, Chem. Ber., 1974, 107, 1971; (e) P. G.
Gassman and T. H. Johnson, J. Am. Chem. Soc., 1975, 98, 861;
( f ) R. Gleiter and F. Ohlbach, J. Org. Chem., 1996, 61, 4929.
21 For acid-catalyzed ring opening reactions of donor–acceptor
substituted cyclopropanes, see: (a) J. W. Herndon and S. U. Tumer,
J. Org. Chem., 1991, 56, 286; (b) L. K. A. Pilsl, T. Ertl and O. Reiser, Org.
Lett., 2017, 19, 2754.
´
Int. Ed., 1978, 17, 603; (d) P. Camps, T. Gomez, A. Otermin,
M. Font-Bardia, C. Estarellas and F. J. Luque, Chem. – Eur. J.,
2015, 21, 14036.
22 M. Isobe, N. Fukami and T. Goto, Chem. Lett., 1985, 71.
This journal is ©The Royal Society of Chemistry 2019
Chem. Commun., 2019, 55, 302--305 | 305