The photochemical C2-C6 cyclization of enyne-allenes: Interception of the fulvene diradical with a radical clock ring opening

Article abstract of DOI:10.1021/jo900439x

(Chemical Equation Presented) The mechanism of the photochemical C ²-C⁶ cyclization of enyne-allenes has been studied through radical clock, intramolecular kinetic isotope effect, and laser flash photolysis (LFP) experiments as well

Full text of DOI:10.1021/jo900439x

pubs.acs.org/joc
The Photochemical C2-C6 Cyclization of Enyne-Allenes: Interception of
the Fulvene Diradical with a Radical Clock Ring Opening
Gotz Bucher,*,†,§ Atul A. Mahajan,‡ and Michael Schmittel*,‡
€
†
€
€
Lehrstuhl fu€r Organische Chemie II, Ruhr-Universitat Bochum,Universitatsstrasse 150, D-44801 Bochum,
Germany, and ‡Organische Chemie I, FB 8 (Chemie-Biologie), Universita€t Siegen, Adolf-Reichwein-Strasse,
D-57068 Siegen, Germany. §Present address: WestCHEM, University of Glasgow, Joseph-Black-Building,
University Avenue, Glasgow G12 8QQ, United Kingdom.
goebu@chem.gla.ac.uk; schmittel@chemie.uni-siegen.de
Received February 26, 2009
The mechanism of the photochemical C2-C6 cyclization of enyne-allenes has been studied through
radical clock, intramolecular kinetic isotope effect, and laser flash photolysis (LFP) experiments as
well as density functional theory (DFT) and ab initio computations. While the photochemical
cyclization of enyne-allenes 1 and 2 furnished ene and Diels-Alder products without any
cyclopropyl ring opening, that of 3 carrying the ultrafast diphenylcyclopropylcarbinyl radical clock
afforded products derived from cyclopropyl ring opening. Laser flash photolysis (LFP) studies on
enyne-allene 3 point to an allene triplet excited state (transient A) as a primarily formed short-lived
(τ = 430 ns) intermediate. In addition, we have obtained evidence for the formation of a
diphenylmethyl-type diradical (transient C, τ=1.0 μs) resulting from ring opening of a diphenyl-
cyclopropane ring. C subsequently undergoes a surprisingly slow (τ = 1.0 μs) 1,6-hydrogen shift
leading to the stable 1,3-diene 6.
Introduction
triggered diradical cyclizations, i.e., the Bergman1 and
Myers-Saito2 cycloaromatizations, to ultimately damage
DNA by hydrogen atom abstraction. Not surprisingly, the
photochemical analogues of the Bergman3 and recently of
the Myers-Saito4 cyclization have become equally a major
concern of research due to their high applicability in photo-
dynamic therapy (PDT).5 Along with the photochemical
Myers-Saito (C2-C7) cycloaromatization, the photochemi-
cal C2-C6 cyclization of enyne-allenes was observed as
well.4 Apparently, spin control allows management between
the two photochemical pathways, as the Myers-Saito (C2-
C7) cycloaromatization was observed from the singlet mani-
Natural enediyne antitumor antibiotics, such as neocar-
zinostatin, dynemicin, etc., avail themselves of two thermally
(1) Bergman, R. G. Acc. Chem. Res. 1973, 6, 25–31.
(2) (a) Myers, A. G.; Kuo, E. Y.; Finney, N. S. J. Am. Chem. Soc. 1989,
111, 8057–8059. (b) Nagata, R.; Yamanaka, H.; Okazaki, E.; Saito, I.
Tetrahedron Lett. 1989, 30, 4995–4998. (c) Saito, K.; Watanabe, T.; Takaha-
shi, K. Chem. Lett. 1989, 2099–2102. (d) Myers, A. G.; Dragovich, P. S.;
Kuo, E. Y. J. Am. Chem. Soc. 1992, 114, 9369–9386.
(3) (a) Turro, N. J.; Evenzahav, A.; Nicolaou, K. C. Tetrahedron Lett.
1994, 35, 8089–8092. (b) Funk, R. L.; Young, E. R. R.; Williams, R. M.;
Flanagan, M. F.; Cecil, T. L. J. Am. Chem. Soc. 1996, 118, 3291–3292.
(c) Evenzahav, A.; Turro, N. J. Am. Chem. Soc. 1998, 120, 1835–1841.
(d) Kaneko, T.; Takahashi, M.; Hirama, M. Angew. Chem., Int. Ed. 1999, 38,
1267–1268. (e) Choy, N.; Blanco, B.; Wen, J. H.; Krishan, A.; Russell, K. C.
Org. Lett. 2000, 2, 3761–3764. (f) Clark, A. E.; Davidson, E. R.; Zaleski, J. M.
J. Am. Chem. Soc. 2001, 123, 2650–2657. (g) Chandra, T.; Kraft, B. J.;
Huffman, J. C.; Zaleski, J. M. Inorg. Chem. 2003, 42, 5158–5172. (h) Benites,
P. J.; Holmberg, R. C.; Rawat, D. S.; Kraft, B. J.; Klein, L. J.; Peters, D. G.;
Thorp, H. H.; Zaleski, J. M. J. Am. Chem. Soc. 2003, 125, 6434–6446.
(i) Alabugin, I. V.; Manoharan, M. J. Phys. Chem. A 2003, 107, 3363–3371.
(4) Schmittel, M.; Mahajan, A. A.; Bucher, G. J. Am. Chem. Soc. 2005,
127, 5324–5325.
(5) (a) Boyle, R. W.; Dolphin, D. Photochem. Photobiol. 1996, 64, 469–
485. (b) Ali, H.; van Lier, J. E. Chem. Rev. 1999, 99, 2379–2450. (c) Ackroyd,
R.; Kelty, C.; Brown, N.; Reed, M. Photochem. Photobiol. 2001, 74, 656–669.
(d) Bonnett, R.; Martinez, G. Tetrahedron 2001, 57, 9513–9547.
5850 J. Org. Chem. 2009, 74, 5850–5860
Published on Web 07/14/2009
DOI: 10.1021/jo900439x
r
2009 American Chemical Society

Bucher et al.
JOCArticle
fold, whereas the photochemical C2-C6 cyclization could be
ignited after triplet excitation (Scheme 1). On the basis of a
combined LFP and DFT study, it was recently suggested that
the photochemical C2-C6 cyclization evolves from a triplet
allene intermediate (τ = 30 ns) that initiates benzofulvene
formation.6 In contrast, the exact nature of the long-lived
intermediate (τ = 30 μs) was not identified beyond doubt,
although experimental and computational evidence pointed
to a fulvene triplet diradical. As the thermal C2-C6 cycliza-
tion of enyne-allenes has become important over the past
decade due to its striking potential for the synthesis of
complex carbocycles7 and for DNA cleavage,8 a mechanistic
comparison of the thermal and photochemical diradical
cyclization pathways should open further venues for the
synthetic use of enyne-allenes.
SCHEME 1. Thermal and Photochemical Myers-Saito (C2-
C7) and C2-C6 Cyclizations of Enyne-Allenes
used as a kinetic clock13 and as a mechanistic probe14 for
radical and diradical15,16 intermediates. Formation of ring-
opened products via the cyclopropylcarbinyl substrate is
generally accepted as evidence for the formation of a radical
intermediate. Very recently, we have disclosed experimental
evidence for the stepwise mechanism in the thermal C2-C6
cyclization of enyne-allenes using an ultrafast radical
clock.17 Encouraged by this finding, we decided to interro-
gate also the photochemical route using the radical clock as a
mechanistic probe and to complement these studies with
kinetic isotope and laser flash photolysis investigations to
further understand the nature of the long-lived intermediate.
The thermal C2-C6 cyclization of enyne-allenes has been
extensively studied over the past few years, both mechan-
istically9 and computationally.10 Some years back, our
group as well as Lipton et al. investigated the C2-C6
cyclization using radical clock experiments, but all early
attempts to locate cyclopropyl ring-opened products in the
thermal C2-C6 cyclization of enyne-allenes were met with
failure.9a,11 Cyclopropyl12 substrates have been successfully
Results
As part of our study to comprehend the mechanism of the
photochemical C2-C6 cyclization, we have chosen enyne-
allenes 1-3 (Chart 1) as model compounds due to their
special features: (1) the triisopropylsilanyl group (TIPS) at
the alkyne terminus raises the cyclization barrier thus pre-
cluding any thermal reaction during photolysis; (2) the bis-
(4-bromophenyl)amine unit was used as an internal triplet
sensitizer to initiate the photochemical reaction; and (3) the
“radical clock” reporter group was attached to the allene
terminus to trap the radical intermediate.
(6) Bucher, G.; Mahajan, A. A.; Schmittel, M. J. Org. Chem. 2008, 73,
8815–8828.
(7) (a) Zhang, H. R.; Wang, K. K. J. Org. Chem. 1999, 64, 7996–7999.
(b) Wang, K. K.; Zhang, H. R.; Petersen, J. L. J. Org. Chem. 1999, 64, 1650–
1656. (c) Yang, Y. H.; Petersen, J. L.; Wang, K. K. J. Org. Chem. 2003, 68,
5832–5837. (d) Yang, Y. H.; Petersen, J. L.; Wang, K. K. J. Org. Chem. 2003,
68, 8545–8549.
(8) (a) Schmittel, M.; Kiau, S.; Siebert, T.; Strittmatter, M. Tetrahedron
Lett. 1996, 37, 7691–7694. (b) Schmittel, M.; Maywald, M.; Strittmatter, M.
Synlett 1997, 165–166.
Synthesis. Preparation of enyne-allenes 1-3 is described
elsewhere.17 4 was obtained in a three-step synthesis from
benzaldehyde after addition of BrMgCtC;R, (prepared
from the reaction of EtMgBr with R = 1-ethynyl-2,2-
diphenylcyclopropane18), acetylation of the propargyl alco-
hol with acetic anhydride/DMAP, and concluding with
(9) (a) Schmittel, M.; Strittmatter, M.; Kiau, S. Angew. Chem., Int. Ed.
Engl. 1996, 35, 1843–1845. (b) Schmittel, M.; Strittmatter, M.; Vollmann, K.;
Kiau, S. Tetrahedron Lett. 1996, 37, 999–1002. (c) Schmittel, M.; Steffen, J.
P.; Auer, D.; Maywald, M. Tetrahedron Lett. 1997, 38, 6177–6180.
(d) Schmittel, M.; Keller, M.; Kiau, S.; Strittmatter, M. Chem.;Eur. J.
1997, 3, 807–816. (e) Engels, B.; Lennartz, C.; Hanrath, M.; Schmittel, M.;
Strittmatter, M. Angew. Chem., Int. Ed. 1998, 37, 1960–1963. (f) Schmittel,
M.; Strittmatter, M. Tetrahedron 1998, 54, 13751–13760. (g) Li, H. B.;
Zhang, H. R.; Petersen, J. L.; Wang, K. K. J. Org. Chem. 2001, 66, 6662–
6668. (h) Schmittel, M.; Steffen, J. P.; Maywald, M.; Engels, B.; Helten, H.;
Musch, P. J. Chem. Soc., Perkin Trans. 2 2001, 1331–1339. (i) Yang, Y. H.;
Petersen, J. L.; Wang, K. K. J. Org. Chem. 2003, 68, 5832–5837. (j) Yang, Y.
H.; Petersen, J. L.; Wang, K. K. J. Org. Chem. 2003, 68, 8545–8549.
(10) (a) Engels, B.; Hanrath, M. J. Am. Chem. Soc. 1998, 120, 6356–6361.
(b) Schreiner, P. R.; Prall, M. J. Am. Chem. Soc. 1999, 121, 8615–8627.
(c) Cramer, C. J.; Kormos, B. L.; Seierstad, M.; Sherer, E. C.; Winget, P. Org.
Lett. 2001, 3, 1881–1884. (d) de Visser, S. P.; Filatov, M.; Shaik, S. Phys.
Chem. Chem. Phys. 2001, 3, 1242–1245. (e) Stahl, F.; Moran, D.; Schleyer, P.
V.; Prall, M.; Schreiner, P. R. J. Org. Chem. 2002, 67, 1453–1461.
(f) Schreiner, P. R.; Navarro-Vazquez, A.; Prall, M. Acc. Chem. Res. 2005,
38, 29–37.
(14) (a) Alonso, M. E.; Hernandez, M. I.; Gomez, M.; Jano, P.; Pekerar,
S. Tetrahedron 1985, 41, 2347–2354. (b) Baldwin, J. E.; Adlington, R. M.;
Domayne-Hayman, B. P.; Knight, G.; Ting, H.-H. J. Chem. Soc., Chem.
Commun. 1987, 1661–1663. (c) Suckling, C. J. Angew. Chem., Int. Ed. Engl.
1988, 27, 537–552. (d) Slama, J. T.; Satsangi, R. K.; Simmons, A.; Lynch, V.;
Bolger, R. E.; Suttie, J. J. Med. Chem. 1990, 33, 824–832. (e) Arasasingham,
R. D.; He, G.-X.; Bruice, T. C. J. Am. Chem. Soc. 1993, 115, 7985–7991.
(f) Dixon, C. E.; Hughes, D. W.; Baines, K. M. J. Am. Chem. Soc. 1998, 120,
11049–11053. (g) Hartshorn, R. M.; Telfer, S. G. J. Chem. Soc., Dalton
Trans. 1999, 3565–3571. (h) Newcomb, M.; Toy, P. H. Acc. Chem. Res. 2000,
33, 449–455. (i) Kretzschmar, I.; Levinson, J. A.; Friend, C. M. J. Am. Chem.
€
(11) Brunette, S. R.; Lipton, M. A. J. Org. Chem. 2000, 65, 5114–5119.
(12) (a) Shimizu, N.; Nishida, S. J. Chem. Soc., Chem. Commun. 1972,
389–390. (b) Rudolph, A.; Weedon, A. C. Can. J. Chem. 1990, 68, 1590–1596.
(c) Newcomb, M.; Johnson, C. C.; Manek, M. B.; Varick, T. R. J. Am. Chem.
Soc. 1992, 114, 10915–10921. (d) Adam, W.; Curci, R.; D’Accolti, L.; Dinoi,
A.; Fusco, C.; Gasparrini, F.; Kluge, R.; Paredes, R.; Schulz, M.; Smerz, A.
Soc. 2000, 122, 12395–12396. (j) Adam, W.; Roschmann, K. J.; Saha-Moller,
C. R.; Seebach, D. J. Am. Chem. Soc. 2002, 124, 5068–5073. (k) Tokuyasu,
T.; Kunikawa, S.; Masuyama, A.; Nojima, M. Org. Lett. 2002, 4, 3595–3598.
(l) Samuel, M. S.; Baines, K. M. J. Am. Chem. Soc. 2003, 125, 12702–12703.
(m) Samuel, M. S.; Jenkins, H. A.; Hughes, D. W.; Baines, K. M. Organo-
metallics 2003, 22, 1603–1611. (n) Flemmig, B.; Kretzschmar, I.; Friend, C.
M.; Hoffmann, R. J. Phys. Chem. A 2004, 108, 2972–2981. (o) Tong, X.;
DiLabio, G. A.; Clarkin, O. J.; Wolkow, R. A. Nano Lett. 2004, 4, 357–360.
(15) (a) Wagner, P. J.; Liu, K.-C.; Noguchi, Y. J. Am. Chem. Soc. 1981,
103, 3837–3841. (b) Cheng, K.-L.; Wagner, P. J. J. Am. Chem. Soc. 1994, 116,
7945–7946. (c) Gan, C. Y.; Lambert, J. N. J. Chem. Soc., Perkin Trans. 1
1998, 2363–2372. (d) Dopico, P. G.; Finn, M. G. Tetrahedron 1999, 55, 29–
62. (e) Broyles, D. A.; Carpenter, B. K. Org. Biomol. Chem. 2005, 3, 1757–
1767.
€
(16) For a theoretical study on radical clock openings see: Jager, C. M;
Hennemann, M.; Mieszala, A.; Clark, T. J. Org. Chem. 2008, 73, 1536–1545.
(17) Schmittel, M.; Mahajan, A. A.; Bucher, G.; Bats, J. W. J. Org. Chem.
2007, 72, 2166–2173.
(18) Bettinetti, G. F.; Desimoni, G.; Gruenanger, P. Gazz. Chim. Ital.
1964, 94, 91–108.
.
K.; Veloza, L. A.; Weinkotz, S.; Winde, R. Chem.;Eur. J. 1997, 3, 105–109.
(13) (a) Griller, D.; Ingold, K. U. Acc. Chem. Res. 1980, 13, 317–323.
(b) Engel, P. S.; Keys, D. E. J. Am. Chem. Soc. 1982, 104, 6860–6861.
(c) Campredon, M.; Kanabus-Kaminska, J. M.; Griller, D. J. Org. Chem.
1988, 53, 5393–5396. (d) Castellino, A. J.; Bruice, T. C. J. Am. Chem. Soc.
1988, 110, 1313–1315. (e) Newcomb, M.; Manek, M. B.; Glenn, A. G. J. Am.
Chem. Soc. 1991, 113, 949–958. (f) Adam, W.; Heil, M. J. Am. Chem. Soc.
1991, 113, 1730–1736. (g) Adam, W.; Finzel, R. J. Am. Chem. Soc. 1992, 114,
4563–4568. (h) Newcomb, M. Tetrahedron 1993, 49, 1151–1176. (i) Engel, P.
S.; Wu, A. Y. J. Org. Chem. 1994, 59, 3969–3974. (j) Caldwell, R. A.; Zhou, L.
W. J. Am. Chem. Soc. 1994, 116, 2271–2275. (k) Engel, P. S.; Lowe, K. L.
Tetrahedron Lett. 1994, 35, 2267–2270. (l) Tanko, J. M.; Blackert, J. F. J.
Chem. Soc., Perkin Trans. 2 1996, 1775–1779. (m) Cooksy, A. L.; King, H. F.;
Richardson, W. H. J. Org. Chem. 2003, 68, 9441–9452.
e
J. Org. Chem. Vol. 74, No. 16, 2009 5851

Bucher et al.
JOCArticle
CHART 1. Model Compounds 1-5 Used in the Photochemical Investigation
SCHEME 2. Photochemical Cyclization of Enyne-Allenes 1-3
allene formation19 by introducing the triarylamine unit as
R0ZnCl in the presence of Pd(PPh3)4. The monodeuterated
enyne-allene 5 was prepared by using our established pro-
tocol20,21 by adding the deuterated butyl group as
CH3CH2CH2CHD-MgBr in the presence of CuI/LiBr22 to
the propargyl alcohol derivative in the last step (see the
Supporting Information).
Photolysis. Photolysis of a 1.43 mM solution of 1 in
toluene in the presence of an excess of 1,4-cyclohexadiene
(1,4-CHD) at 300 nm in the Rayonet reactor afforded ene
and formal Diels-Alder products 7 and 9, each in 8% yield
(Scheme 2; Table 1). The 1H and 13C NMR spectra of 7 and 9
are in full agreement with those of formal ene and Diels-
Alder products isolated from the thermal cyclization of
enyne-allene 1.17 Under analogous photolysis conditions,
enyne-allene 2 provided the formal ene product 823 along
with Diels-Alder product 10 in 8% and 12% yield, respec-
tively. In contrast, photolysis of 3 in toluene (1.25 mM
solution) afforded the cyclopropyl ring-opened product 6
in 40% yield (two isomers),24a along with the formal Diels-
Alder product 11, isolated in 4% yield. The structure of
compound 6 was assigned on the basis of H, 13C, H,1H-
COSY, HSQC, and HRMS NMR data. The 1H NMR
showed a characteristic doublet with a coupling constant
∼11.6 Hz for the two olefinic protons Ha and Hb. Moreover,
1H,1H-COSY clearly showed the coupling between Ha and
Hb. A control experiment was carried out with 3 in toluene-
d8. The lack of deuterium incorporation in 6 suggested that
1
1
(24) (a) The second isomer of 6 (isomer about the TIPS-CdC bond) arises
from a follow-up irradiation of the photolabile product 6 as demonstrated
by a photolysis of the thermally generated isomerically pure 6. (b) Transient C
shows up as negative absorption in the difference spectrum shown in Figure 2, but
it can also be seen in Figure 1. At λ=332 nm, the transient decay of C is seen on top
of a strong negative absorption due to bleaching of the precursor. The transient
spectrum recorded 450 ns after LFP (solid circles) shows a less negative
absorbance at λ=332 nm than the transient spectra recorded 1.7 or 7.5 μs after
LFP (open circles and solid triangles). If the 450 ns spectrum is subtracted from the
7.5 μs spectrum (as in Figure 2), a less negative number is subtracted from a more
negative number, resulting in a net negative absorption at λ=332 nm in Figure 2. By
displaying the data in this format, the characteristic sharp band signature of the
diphenylmethyl radical moiety is visualized, which is not easily seen from Figure 1
alone.
(19) Elsevier, C. J.; Stehouwer, P. M.; Westmijze, H.; Vermeer, P. J. Org.
Chem. 1983, 48, 1103–1105.
(20) Schmittel, M.; Vavilala, C. J. Org. Chem. 2005, 70, 4865–4868.
(21) Schmittel, M.; Vavilala, C.; Jaquet, R. Angew. Chem., Int. Ed. 2007,
46, 6911–6914.
(22) Macdonald, T. L.; Reagan, D. R.; Brinkmeyer, R. S. J. Org. Chem.
1980, 45, 4740–4747.
(23) The purification of ene product was difficult due to the unidentified
impurities with the same Rf value.
5852 J. Org. Chem. Vol. 74, No. 16, 2009

Bucher et al.
JOCArticle
TABLE 1. Yields of C2-C6 Products after Irradiation of 1-3 at 300 nm
compd (conc)
λmax/nm (in hexane)
solvent (irradiation time)
ene product (yield)a
DA productb (yield)a
1 (1.43 mM)
2 (1.40 mM)
3 (1.25 mM)
206, 244, 267, 320
212, 247, 262, 317
212, 272, 322
toluene + CHD c (6 h)
toluene + CHD c (7 h)
toluene + CHD c (5 h)
7 (8%)
8 (8%)
6 (40%)d
9 (8%)
10 (12%)
11 (4%)
aYield after isolation, estimated relative error 10%. bDA: formal Diels-Alder. cIn the presence of 1,4-cyclohexadiene (200-fold of excess). d[1,6]-H
shift product.
SCHEME 3. Thermal and Photochemical Cyclization of Enyne-Allene 5
TABLE 2. Intramolecular Kinetic Isotope Effect in the C2-C6 Photo-
chemical and Thermal Cyclization of 5
absorption maxima at λmax at ca. 458 nm (A: kA=(2.3 ( 0.5) ꢀ
106 s-1; B: kB=(3.4 ( 0.8) ꢀ 105 s-1). Transient C showed a
yield, %
λ
max=332 and ∼500 nm (kC=(1.0 ( 0.1) ꢀ 106 s-1, transientC
is most easily seen as negative absorption in Figure 2).24b In
addition, we observed the growing transient D at λmax=380
nm (k = (1.0 ( 0.1) ꢀ 106 s-1). A very weak transient
absorption (kdecay ≈ 3.8 ꢀ 105 s-1) extending to λ=600 nm
also belongs to B. The fast component of the 458 nm transient,
i.e., A (τ = 435 ns), was quenched by 3O2 at a diffusion
DAb
intramolecular ene products product
reaction solvent
conditions (reaction time)
KIE (d1)
1.0016c
1.0088c
12H, 12Da
13a
thermal
photo-
chemical
toluene (48 h)
toluene (7 h)
63
10
15
controlled rate, kA,O =(4.7 ( 0.7) ꢀ 109 M-1s-1. In contrast,
aYield of isolated material, estimated relative error 10%. bDA: formal
2
Diels-Alder. cEstimated error:21 (0.002.
the slow component of the 458 nm transient, i.e., B, showed no
observable reactivity toward 3O2. Transient C (λmax=332 nm)
decays with a lifetime τ=1.0 μs. It was quenched by 3O2 at a
the hydrogen transfer proceeded in an intramolecular fash-
ion, e.g., as sketched for C in Scheme 4.
near diffusion-controlled rate, kC,O =(1.2 (0.2) ꢀ 109 M-1 s-1
.
Concomitant with the decay of C we observed a growing
2
Intramolecular Kinetic Isotope Effect. Recent investiga-
tions on thermal enyne-allene cyclizations have shown
interesting signatures with regard to kinetic isotope effects,21
which we wanted to apply as a test for possibly identical
intermediates in the thermal and photochemical cyclization.
To evaluate the intramolecular kinetic isotope effect, the
thermal cyclization of the monodeuterated enyne-allene 5
was investigated. It afforded a 1:1 mixture of the cyclized ene
products 12H,12D in reasonable yield (63%), resulting in an
intramolecular kinetic isotope effect of 1.0016 (at 110 °C).
Upon photolysis of enyne-allene 5, the formal ene 12H,12D
(1:1) and Diels-Alder products 13 were isolated in 10% and
15% yield (Scheme 3; Table 2). Notably, the two ene
products 12H,12D were formed with a comparable intra-
molecular kinetic isotope effect in both the thermal and
photochemical reaction channel.
transient at λmax =380 nm. This species D was stable on the
longest time scale (ca. 50 ms) accessible to our experimental
setup. Apart from D, no transient growth was detectable in our
experiments even on the shortest time scale accessible to our
instrument (ca. 20 ns). Figure 1 shows the transient spectra
observed during different time intervals after LFP of 3, and
Figure 2 shows a transient difference spectrum. Transient traces
are given in Figures S1 and S2 in the SI.
For comparison, we also investigated an allene compound
devoid of the alkyne moiety in the ortho position. LFP (λexc
=
355 nm, cyclohexane solution) of 4 furnished a transient spec-
trum with λmax=430 and 470 nm, and a transient lifetime τ=1 μs
(Figure 3). This transient was quenched by O2 with a rate
3
constant kO =(2.5 ( 0.5) ꢀ 109 M-1 s-1 (Figure S6, see the SI).
2
Laser Flash Photolysis. In the LFP experiment, excitation
(λexc = 355 nm) of a solution of 3 in cyclohexane purged
with argon led to the observation of three decaying transients
A-C. Transients A and B were found to exhibit very similar
Discussion
Identity of the Transients As Derived from LFP. The LFP
results of linear allene 4 show a short-lived transient (τ=1 μs)
J. Org. Chem. Vol. 74, No. 16, 2009 5853

Bucher et al.
JOCArticle
with λmax = 430 and 470 nm. A similar transient (λmax
400 nm, τ=980 ns) has recently been observed in the LFP and
pulse radiolysis of 1,3-diphenylallene,25 which was assigned
to the first triplet excited state of 1,3-diphenylallene. Triplet
tetraphenylallene (λmax = 385 nm)26 has been reported to
have a shorter lifetime (τ=90 ns). On the basis of these results
and the characteristic reactivity toward oxygen, we assign the
transient species A (of 3, Scheme 4) and the transient of 4 to a
triplet excited state of the allene. Cyclopropyl ring opening at
this stage is slow, as demonstrated also by preparative
photochemical results: when allene 4 was subjected to photo-
lysis at 300 nm in dry degassed toluene, no ring opening
occurred and only starting material 4 was recovered.
=
Transient B could possibly be assigned to a fulvene-diyl
diradical, either in its triplet or singlet spin multiplicity. Due
to steric shielding of the reactive vinylic radical center by the
bulky TIPS group, even a triplet fulvene-diyl may react
rather slowly with oxygen6 and cannot be discarded as
intermediate. As transient C shows a characteristic sharp
absorption maximum at λmax =332 nm and additionally a
high reactivity toward oxygen, we assign it to an intermediate
containing a diphenylmethyl radical moiety (Scheme 4).27
Notably, we do not observe such typical sharp absorption for
the diphenylmethyl radical moiety (λmax = 332 nm) in the
LFP of linear allene 4. Hence, we can reliably exclude
formation of C directly from the allene triplet via intermedi-
ate E, suggesting that the cyclopropane ring opens at the
diradical stage, either singlet or triplet (path 1). Transient C
finally leads to the formation of transient D, which we assign
FIGURE 1. Transient spectra, observed during different time inter-
vals after LFP (355 nm) of 3 in cyclohexane (1 atm of Ar): (solid
circles) time window centered 450 ns after LFP; (open circles) time
window centered 1.7 μs after LFP; and (solid triangles) time window
centered 7.5 μs after LFP.
to 6 as suggested by identical absorption maxima, i.e., λmax
=
380 nm, for both D and 6.
At this point two important questions still remain unre-
solved. (i) What is the multiplicity of the fulvene-diyl
diradical that we assign to transient B? (ii) Does the cyclo-
propylcarbinyl rearrangement take place at the singlet or
triplet diradical stage? Experimentally, these questions are
addressed in the following by DFT calculations as well as by
radical clock and intramolecular kinetic isotope effects.
Calculations. We have performed a series of calculations
on potential reaction intermediates formed upon photolysis
of model compound 30. For this, we have employed a
standard DFT method (B3LYP) in combination with a 6-
31G(d) basis set. To facilitate the calculations on this sizable
system to some degree, the remote bromine atom substitu-
ents in the experimentally studied system 3 were omitted in
the calculations. Calculations on 30 and the transients de-
rived thereof are complicated by the fact that the molecular
framework of 3/30 in itself is chiral. In combination with the
chiral diphenylcyclopropyl unit, two possible diastereomers
(l and u) were received that in our hands could not be
separated experimentally. Unfortunately, this complication
also applies to many of the reaction intermediates to be
discussed here, as most of them are also inherently chiral.
Precursor 30 and its triplet excited states: The equilibrium
geometries of two rotamers (syn, anti) of both diastereomers
FIGURE 2. Transient difference spectrum derived from two of the
transient spectra shown in Figure 1. The spectrum given here
displays the difference between the 7.5 μs and 450 ns time windows.
(25) Yamashita, T.; Nishiguchi, T.; Kato, J.; Muroya, Y.; Katsumura, Y.
J. Photopolym. Sci. Technol. 2005, 18, 95–102.
(26) Brennan, C. M.; Caldwell, R. A.; Elbert, J. E.; Unett, D. J. J. Am.
Chem. Soc. 1994, 116, 3460–3464.
(27) (a) Hadel, L. M.; Platz, M. S.; Scaiano, J. C. J. Am. Chem. Soc. 1984,
106, 283–287. (b) Klett, M. W.; Johnson, R. P. J. Am. Chem. Soc. 1985, 107,
3963–3971.
FIGURE 3. Transient spectra, observed during two different time
intervals after LFP (355 nm, cyclohexane) of 4: (solid circles) 320 ns
after excitation and (open circles) 1.7 μs after excitation.
5854 J. Org. Chem. Vol. 74, No. 16, 2009

Bucher et al.
JOCArticle
SCHEME 4. Initial Hypothesis Assigning the Observed Transients to Possible Intermediates/Productsa
aThe observed transients B and C are kinetically not linked to each other.
SCHEME 5. Singlet Ground State Enthalpies of Isomers of 30,
Relative to anti-u-30 = 0.0 kcal mol-1 a
of 30 (u and l) were localized. In the case of the u-diastereo-
mer, the anti-rotamer is predicted to be the lowest energy
conformer, being 7.4 kcal mol-1 more stable than the syn-
rotamer. The situation is different with the l-diastereomer.
To our surprise, the anti-rotamer is predicted to be only 0.1
kcal mol-1 more stable than the syn-rotamer. The calcula-
tions thus indicate that the anti-rotamer should be exclu-
sively present in u-30, while the syn- and anti-rotamers of l-30
should have roughly equal equilibrium concentrations.
Scheme 5 shows the molecules and their energies (relative
to anti-u-30=0.0 kcal mol-1). On the basis of the geometries
optimized at the B3LYP/6-31G(d) level of theory, we addi-
tionally calculated single-point energies at the RIMP2/
TZVP level of theory, which should better describe the
impact of weak intramolecular dispersive interactions.28a
The results of these ab initio calculations (also given in
Scheme 5) indicate that the stability of the two syn-rotamers
of 30 is underestimated by DFT by 5-6 kcal mol-1. Inter-
estingly, for l-30 the syn-conformation is more stable than the
anti-conformation by 5.5 kcal mol-1
!
According to our calculations, the first triplet excited states
of l-30 and u-30 are typical allene-type triplets with an allyl-
vinyl type diradical character. The triplet states may contain up
to four stereogenic elements, i.e., a stereogenic axis, twice Z/E-
isomerism, and a stereogenic center in the diphenylcyclopropyl
aRegular font: B3LYP/6-31G(d), including ZPE correction. In italics: RIMP2/
TZVP//B3LYP/6-31G(d), without ZPE correction.
group. Helicity arises in the syn-triplet states of 30 along the
allyl-vinyl-type diradical site due to an interaction of the bulky
substituent (diphenylcyclopropyl or diphenylaminophenyl)
forcing the phenylacetylene TIPS unit slightly out of the
common plane. The energetically lower lying anti-conformers
do not suffer this steric repulsion and are nearly planar.28b
Therefore, a total of 12 stereoisomeric triplet states consisting
(28) (a) Unlike MP2 theory, B3LYP will tend to underestimate weak
dispersive interactions (cf.: Zhao, Y.; Tishchecko, O.; Truhlar, D. G. J. Phys.
Chem. B 2005, 109, 19046–19051). This will result in enthalpies of the syn-
rotamers that are too high, as in the case of the precursor ground states, see
above. (b) We have not been able to extend the RIMP2 single-point calculations
to the triplet hypersurface.
J. Org. Chem. Vol. 74, No. 16, 2009 5855

Bucher et al.
JOCArticle
a
SCHEME 6. Reaction Pathways Leading from anti-Conformers of Triplet Excited States of 30 to Triplet States of Fulvene-Diyl Diradicals 140
aIn brackets: calculated enthalpies (in kcal/mol), relative to anti-u-30 = 0.0 kcal mol-1.
SCHEME 7. Possible Interconversion of a Triplet Allene into an
of six pairs of enantiomers (E-u-syn, E-l-syn, Z-u-syn, Z-l-syn,
E-Fulvene Diyl Diradical
E-anti, and Z-anti) are conceivable. It is noted that excitation of
the two rotamers of both diastereomeric precursors (l, u) can
potentially lead to the formation of one (and only one)
enantiomer of each rotameric triplet state diastereomer.29
Scheme 6 shows reaction pathways leading from the anti-
rotamers of triplet 30 to the triplet states of the fulvene-diyl-
type diradicals 140.
Our calculations indicate that the activation enthalpies for
the syn-anti conversion of the triplet excited states of the two
diastereomers of 30 are somewhat smaller than the activation
(fulvene) bond (Scheme 7). However, steric repulsion bet-
ween the TIPS group and the diphenylcyclopropyl group
(= R) or, in a different rotamer, the triphenylamine moiety,
can be expected to exert a strongly destabilizing effect.
Vinyl radicals have been reported to show relatively small
barriers for Z/E-isomerization30 that are further reduced
upon R-silylation.31 Even at the primitive UHF/STO-3G
level of theory, which in the study by Lalitha and Chandra-
sekhar resulted in the highest barrier for isomerization of an
R-silylvinyl radical,31 our attempts to optimize a derivative
of 140 with E-configuration at the vinyl radical moiety
(E-vinyl isomer) were fruitless, resulting in the optimization
of the Z-silylvinyl isomer instead.32 For that reason we have
to assume that (E-silyl)fulvene-diyl-type diradicals are no
minima on the potential energy hypersurface.
enthalpies of a smaller system studied by us earlier (R=Me
instead of diphenylcyclopropyl).6 With ΔH q of the order of
10 kcal mol-1 and ΔS q of the order of -5 cal mol-1 K-1
,
anti-syn isomerization rate constants of the order of 2 ꢀ 104 s-1
can be expected. The activation enthalpies for cyclization of the
syn-triplet allenes are calculated to be much smaller, of the
order of ΔHq=3.8-7.6 kcal mol-1. For that reason, the syn-
triplet allenes, if formed from the anti-triplet allenes, will be
present in quasistationary concentration only, inaccessible to
direct observation. The results obtained by RIMP2/TZVP
indicate, however, that a significant portion of triplet allenes
can be formed directly as syn-rotamer. This population of syn-
triplet-allenes may be detected by LFP.
It is a priori conceivable that cyclization of the syn-allene
triplets may initially result in fulvene-diyl-type diradicals
that, unlike 140, show an E configuration at the CdC
(30) Galli, C.; Guarnieri, A.; Koch, H.; Mencarelli, P.; Rappoport, Z. J.
Org. Chem. 1997, 62, 4072–4077.
(29) Upon excitation, one of the two stereogenic elements present in the
molecule (the stereogenic center in the cyclopropane unit) is retained,
whereas the stereogenic axis of the linear allene system is converted into
other stereogenic elements (helicity, Z/E-isomerism).
(31) Lalitha, S.; Chandrasekhar, J. Proc. Indian Acad. Sci. (Chem. Sci.)
1994, 106, 259–266.
(32) Attempted optimization at the UB3LYP/6-31G(d) level of theory
gave the same result.
5856 J. Org. Chem. Vol. 74, No. 16, 2009

Bucher et al.
JOCArticle
SCHEME 8. Reaction Pathways Leading from Isomers of Fulvene-Diyl-Type Diradicals 140 to Ring-Opened Diradicals 150, Fulvenes
a
160 ,and Benzofluorenes 170
aIn brackets: enthalpies (in kcal mol-1), relative to anti-u-30 = 0.0 kcal mol-1. In italics: calculated absolute entropies (in cal mol-1 K-1).
Diradical Reactions. Follow-up reactions of the triplet
fulvene-diyl diradicals 140 include ISC to the singlet state,
followed by either hydrogen transfer and fulvene formation
(E-isomers) or addition to the proximal benzene ring of the
triphenylamine moiety (Z-isomers). Alternatively, the cyclo-
propane ring may undergo ring opening, either from the
triplet or from the singlet state of 140. Scheme 8 shows the
results of our calculations on these competing pathways.
On the basis of the calculated activation enthalpies, the E-
rotamers of 140 should preferentially undergo a facile 1,5-
hydrogen shift, yielding fulvene 160. The calculated activa-
tion enthalpies for this strongly exothermic hydrogen shift
are very small (of the order of 1-5 kcal mol-1). For that
reason, the two diastereomers E-140 (l, u) are unlikely to have
lifetimes exceeding 1 ns and therefore are not expected to be
observed in our experiments. Both diastereomers of Z-140,
on the other hand, should preferentially undergo intramole-
cular addition to a benzene ring of the triphenylamine
moiety, yielding isomers of 170. In contrast, cyclopropyl-
methyl ring opening of both E- and Z-140 is impeded by a
somewhat higher activation enthalpy. Taking into account
the calculated entropies, however, the cyclopropylmethyl
ring opening may become competitive, as it is much more
favorable in terms of entropy than the hydrogen shift
or addition reactions. A difference in activation entropy
ΔΔSq = 10 cal mol-1 K-1 between hydrogen transfer or
benzofluorene formation on one hand and ring opening on the
other hand translates into a difference of free energy of
activation ΔΔGq=3 kcal mol-1, which may easily outbalance
the relatively small preference for hydrogen transfer or benzo-
fluorene formation in terms of ΔHq. This at least applies to
E-l-140 and Z-u-140, where the calculated activation enthalpies
for hydrogen transfer (E-l-140) or addition (Z-u-140) are of the
order of 5-6 kcal mol-1. Finally, interconversion of the Z- and
E-isomers of 140 is also conceivable, with activation enthalpies
around 6-8 kcal mol-1 33
.
The spin density at the cyclopropylmethyl-carbon atom
can be expected to be a crucial factor determining the ease of
ring opening. To gain an understanding for the surprisingly
high barriers for cyclopropylmethyl ring opening in 140,
we additionally calculated the activation enthalpies for
ring opening of three simple diphenylcyclopropylmethyl
radicals, namely the parent diphenylcyclopropylmethyl
(33) We have been unable to localize the transition state linking Z-u-140
and E-u-140.
J. Org. Chem. Vol. 74, No. 16, 2009 5857

Bucher et al.
JOCArticle
CHART 2. Various Cyclopropylmethyl Radicals
(18), (E-diphenylcyclopropyl)phenylmethyl (19), and )di-
phenylcyclopropyl)diphenylmethyl (20). As the triplet al-
lenes of importance investigated in this study also bear spin
density at the cyclopropylmethyl-carbon atom (and may
thus in principle also undergo ring opening), Z-u-triplet 30
syn and triplet 40 were also included in the comparison
(Chart 2).
Figure 4 shows a plot of activation enthalpy vs. spin
density (for the data, please see Table S1 in the SI). It clearly
demonstrates the correlation between the two parameters.
The data also indicate that due to a relatively low spin density
at the cyclopropylmethyl carbon atom, the cyclopropyl-
methyl ring-opening reactions in the derivatives investigated
by us (14, triplet-4, and triplet-3) are expected to be slower
than originally anticipated, and that the lifetimes for these
processes may reach well into the nanosecond or even
microsecond realm.
FIGURE 4. Plot of the calculated activation enthalpy for the
cyclopropylmethyl ring-opening reaction of a series of 1,1-diphe-
nylcyclopropanes vs. the calculated Mulliken spin density at the
cyclopropylmethyl carbon atom: (A) Z-u-triplet 30 syn; (B) triplet 40;
(C) E-l-140; and (D) E-u-140.
SCHEME 9. Possible Reaction Pathway Leading to 6’ from
a
s-cis-Z-150
Fate of the Ring-Opened Diradicals. Hydrogen transfer in
150, yielding butadiene 60, requires Z-configuration at the
newly formed carbon-carbon double bond. Among the
ring-opened diradicals 150 formed in the cyclopropylmethyl
ring-opening reactions of isomers of 140, only s-cis-Z-150
matches this requirement. We have been able to locate the
transition state for formation of 60 on the singlet diradical
hypersurface (Scheme 9). The calculations indicate that the
hydrogen transfer requires an activation enthalpy of the
order of 10 kcal mol-1, which should bring this process into
the low microsecond time domain.
aIn brackets: calculated enthalpies (in kcal mol-1, relative to anti-u-30 =
0.0 kcal mol-1); in italics: calculated absolute entropies (in cal mol-1 K-1).
observe none of them. Moreover, the reactions competing
with their formation (formation of ene or Diels-Alder
products) are predicted to occur via very small barriers.
Transients A and B are more difficult to assign unambigu-
ously. A(λmax=458 nm, kdecay=(2.3 (0.5) ꢀ 106 s-1, quenched
by oxygen) likely corresponds to a precursor triplet excited
state, possibly the triplet conformer E-3-triplet anti. Due to a
lack of reactivity toward oxygen, transient B (λmax=458 nm,
kdecay =(3.4 ( 0.8) ꢀ 105 s-1) is unlikely to correspond to a
triplet excited state. Our calculations indicate that the fulvene-
diyl-type diradicals 14 may actually have singlet ground states
by a very small margin. In previous LFP studies of related
allene precursors, transients with characteristics similar to B
were assigned to triplet fulvene-diyl-type diradicals.4,6 Never-
theless, we hesitate to give an assignment of B to any stereo-
isomer of 14, as these (singlet) diradicals are predicted to
undergo follow-up reactions far too easily to have a lifetime
in excess of 1 μs. Moreover, as mentioned above, Z-u-14 is
expected to be the most stable isomer of 14. This species
probably is the (invisible) precursor to C. As B is not a
precursor to C, an assignment of B to singlet Z-u-14 can thus
be ruled out. What possible assignments remain for B? Ground
state triplet fulvene-diyl diradicals can be expected to be longer
lived than their counterparts on the singlet spin manifold, as
the triplet-singlet energy gap would additionally have to be
It is noteworthy that s-cis-Z-150 is formed by ring opening
of Z-u-140, which is the energetically lowest lying fulvene-
diyl-type diradical studied, and which also is the diradical
least likely to undergo an extremely facile hydrogen shift or
addition reaction. For that reason, the hydrogen abstraction
mechanism depicted in Scheme 9 appears plausible to us.
Assignment of the Transients Observed. The spectra and
behavior of the transients, the results of the product studies,
and the calculations allow us to make a reasonable assign-
ment of the transient species observed in the LFP experi-
ments (cf. Scheme 10). Transient D (stable on the ms time
scale, formed with a rate constant k=(1.0 ( 0.1) ꢀ 106 s-1
)
has already been assigned to butadiene 6 based on its UV/vis
spectrum. The discussion above strongly suggests that tran-
sient C, which is the direct precursor to D (= 6), should be
assigned to Z-15, as none of the other isomers of 15 is suitable
for hydrogen transfer. The sharp absorption band observed
at λmax=332 nm is strongly suggestive of a diphenylmethyl
radical moiety.27 As mentioned above, a calculated activa-
tion enthalpy of 10.2 kcal mol-1 is consistent with a lifetime
of the order of 1 μs. E-Isomers of 15 are predicted to be less
stable, which would offer an explanation for the fact that we
5858 J. Org. Chem. Vol. 74, No. 16, 2009

Bucher et al.
JOCArticle
SCHEME 10. Possible Reaction Pathway Leading to 60 from syn-l-30
overcome. We cannot rule out the possibility of our calcula-
tions giving an inaccurate picture of the potential energy
hypersurface and wrongly predicting the stereoisomers of 14
to be ground state singlet species. In this case, an assignment of
B to a stereoisomer of triplet-14 would represent a straightfor-
ward solution to the problem.34 The lack of reactivity of B
toward oxygen need not be of concern here, as a related,
sterically less congested triplet fulvene-diyl diradical has pre-
viously been shown not to react with oxygen, either.6
An additional experiment to probe the course of reaction is
to compare the intramolecular kinetic isotope effect (KIE) in
the thermal and photochemical cyclization of 5. Recently, we
learned that the intramolecular KIE of C2-C6 cyclization of
enyne-allenes has a meaningful signature with regard to the
mechanism.21 Hence, it is interesting to see that in both the
thermal and photochemical cyclization, the intramolecular
KIE is around 1.00. Using our understanding of intramole-
cular KIE,21 this is strong evidence for (i) having basically the
same intermediate involved in the hydrogen transfer step for
both reaction protocols and (ii) for having a stepwise reaction
mechanism with dynamic effect about the diradical.35 As the
intermediacy of a singlet diradical is reasonably well estab-
lished for the thermal pathway by a multitude of mechanistic
and computational studies,36 these results suggest that a singlet
diradical is involved equally in the final H-transfer in the
photochemical process and most likely also in the ring opening
of the cyclopropyl radical clock.
enyne-allenes. The reaction sequence is started with excita-
tion of the allene moiety to its triplet state thus igniting C2-
C6 cyclization to furnish a fulvene-diyl diradical. The inter-
mediacy of the diradical can be derived indirectly from the
time-resolved detection of a radical clock derived ring-
opened product showing a diphenylmethyl radical function-
ality (characteristic sharp signal at λmax=332 nm). Compu-
tational results indicate that radical clock opening in the
fulvene-diyl diradical is much slower than that in mono-
radicals. The investigation, however, failed to provide an
experimental kinetic analysis of the ring opening of the
singlet fulvene-diyl diradical, as the visible transient B was
not a direct precursor of C.
Experimental Section
General Procedure for the Photochemical Cyclization. The
reaction mixture of enyne-allene and 1,4-cyclohexadiene
(CHD) (200-fold excess) in dry degassed toluene was irradiated
at 300 nm in an inert atmosphere. The reaction progress was
monitored by TLC. The product mixture was concentrated and
purified with preparative TLC (aluminum sheet, silica gel
60F254, n-pentane). For characterization of compounds 6, 7, 9,
and 10, see ref 17.
1
Physical Measurements. For all H and 13C NMR spectra
chemical shifts refer to the residual protio solvent signals
(CDCl3 7.26 and 77.0 ppm, C6D6 7.15). Laser flash photolysis:
The system employed has already been described.37 Enyne-
allene 3 was excited with use of the third harmonic (λ=355 nm)
of a Nd:YAG laser (130 mJ/pulse, 7 ns pulse duration). The
concentration of 3 was ca. 3.5 ꢀ 10-5 M in cyclohexane. A flow
cell was used to avoid buildup of photoproducts. Prior to
experiments, the solutions were purged with argon for ca. 20
min. In oxygen quenching studies, variable mixtures of O2 with
Ar were obtained with use of mass flow controllers (MKS).
Calculations. Geometry optimizations were performed at the
(U)B3LYP level of theory,38 using a 6-31G(d) basis.39 The DFT cal-
culations were performed employing the Gaussian03 suite of pro-
grams.40 RIMP2 single point energy calculations41 with a TZVP42
Conclusions
The present investigation reveals interesting new details of
the mechanism of the photochemical C2-C6 cyclization of
(34) We note, however, that a series of calculations on model systems has
indicated the B3LYP method to overestimate rather than to underestimate
the stability of the triplet state of fulvene-diyls,6 and therefore the identity of
B cannot be regarded as fully settled. An alternative scenario would involve
the formation of a phenylvinylcarbene by aryl shift of the singlet excited
allene. While such reactions do occur in arylallene photochemistry, they
generally have a very low quantum yield.27b As we have not isolated any
products indicative of such chemistry, we do not want to embark on any
speculations in this direction at this point.
(37) Bucher, G. Eur. J. Org. Chem. 2001, 2463–2475.
(38) Becke, A. D. J. Chem. Phys. 1993, 98, 5648–5652.
(39) Ditchfield, R.; Hehre, W. J.; Pople, J. A. J. Chem. Phys. 1971, 54,
724–728.
(35) Bekele, T.; Christian, C. F.; Lipton, M. A.; Singleton, D. A. J. Am.
Chem. Soc. 2005, 127, 9216–9223.
(36) (a) Wenthold, P. G.; Lipton, M. A. J. Am. Chem. Soc. 2000, 112,
9265. (b) Musch, P. W.; Engels, B. J. Am. Chem. Soc. 2001, 123, 5557. (c)
Stahl, F.; Moran, D.; Schleyer, P. R.; Prall, M.; Schreiner, P. R. J. Org.
ꢀ
Chem. 2002, 67, 1453. (d) Schreiner, P. R.; Navarro-Vazquez, A.; Prall, M.
Acc. Chem. Res. 2005, 38, 29–37.
(40) Frisch, M. J. et al., Gaussian 03, Revision C02; Gaussian, Inc.,
Wallingford, CT, 2004.
€
(41) Weigend, F.; Haser, M. Theor. Chem. Acc. 1997, 97, 331–340.
(42) Weigend, F.; Ahlrichs, R. Phys. Chem. Chem. Phys. 2005, 7, 3297–
3305.
J. Org. Chem. Vol. 74, No. 16, 2009 5859

Bucher et al.
JOCArticle
basis set were performed with Turbomole software.43 The DFT
enthalpies given are corrected for zero-point vibrational energy.
4-Bromo-N-(4-bromophenyl)-N-{4-((1E)-1-[(1E)-1-((triisopro-
pylsilyl)methylene)-1H-inden-2-yl]-4,4-diphenylbuta-1,3-dienyl)-
phenyl}benzenamine (6):17 IR (film) 3051 (m), 2942 (s, C-H),
2865 (s), 1582 (m), 1486 (s), 1312 (s), 1265 (s), 1178 (w), 1072 (m),
1008 (m), 882 (m), 823 (s), 740 (s), 702 (s) cm-1; HRMS calcd for
C33H2781Br79BrN 889.215, found 889.215.
J = 9.1 Hz, 1H), 6.30 (s, 1H), 6.81 (s, 1H), 6.84 (d, J = 8.8 Hz,
4H), 6.99 (d, J = 8.8 Hz, 2H), 7.04-7.06 (m, 2H), 7.12-7.18 (m,
6H), 7.28 (d, J = 9.0 Hz, 4H), 7.36 (d, J = 8.8 Hz, 2H), 7.37 (d,
J = 6.9 Hz, 2 H), 7.69 (d, J = 7.5 Hz, 1H); MS-EI (70 eV) m/z
813.2 (Mþ); HRMS calcd for C33H2781Br79BrN 813.183, found
813.183.
Bis(4-bromophenyl)[10-(2,2-diphenylcyclopropyl)-11,11a-di-
hydro-4aH-benzo[b]fluoren-7-yl]amine (11): 1H NMR (400
MHz, CDCl3) δ 2.06 (dd, J = 9.4, 5.2 Hz, 1H), 2.51 (dd, J =
7.0, 5.3 Hz, 1H), 3.37 (dd, J = 8.2, 8.2 Hz, 1H), 3.99 (s, 2H),
6.78-6.88 (m, 4H), 6.96 (d, J = 8.8 Hz, 4H), 7.11 (dd, J = 2.3,
9.2 Hz, 1H), 7.27-7.33 (m, 4H), 7.36 (d, J = 8.8 Hz, 4H), 7.41
(d, J = 7.3 Hz, 3H), 7.48 (d, J = 7.5 Hz, 2H), 7.51 (d, J = 7.7
Hz, 1H), 7.76 (d, J = 6.8 Hz, 1H), 7.81 (s, 1H), 8.20 (d, J = 9.2
Hz, 1H); 13C NMR (100 MHz, CDCl3) δ 18.7, 29.7, 29.9, 36.8,
115.5, 116.5 120.3, 121.5, 123.2, 124.9, 125.5, 125.7, 126.3,
126.8, 126.9, 127.3, 127.5, 128.2, 128.6, 128.8, 129.7, 130.3,
132.3, 134.3, 140.6, 140.9, 141.2, 143.6, 143.7, 145.9, 146.5
ppm; MS-EI (70 eV) m/z 733.1 (Mþ); HRMS calcd for
1st isomer: yield 7.7 mg; 1H NMR (400 MHz, C6D6) δ 1.03 (d,
J = 7.4 Hz, 18H), 1.32 (septet, J = 7.4 Hz, 3H), 6.52 (d, J = 8.8
Hz, 4H), 6.60 (d, J = 8.7 Hz, 2H), 6.69 (s, 1H), 6.83 (s, 1H),
6.90-6.93 (m, 3H), 7.10 (d, J = 8.8 Hz, 4H), 7.12-7.16 (m, 3H),
7.16-7.24 (m, 8H), 7.30 (d, J = 11.5 Hz, 1H), 7.37 (d, J = 6.9
Hz, 2H), 7.94 (dd, J = 5.4, 2.7 Hz, 1H); 13C NMR (100 MHz,
CDCl3) δ 12.9, 19.1, 115.6, 120.9, 122.2, 123.7, 125.0, 125.5,
126.4, 127.3, 127.4, 127.6, 127.7, 127.9, 128.1, 128.2, 130.7,
132.3, 132.7, 133.5, 136.2, 137.2, 137.7, 139.9, 142.3, 142.7,
143.3, 143.9, 145.9, 146.2, 155.4 ppm.
2nd isomer:24 yield 7.7 mg; 1H NMR (400 MHz, CDCl3)
δ 0.98 (d, J = 7.4 Hz, 18H), 1.29 (m, 3H), 6.28 (s, 1 H), 6.53 (d,
J = 11.6 Hz, 1H), 6.68 (s, 1H), 6.97 (d, J = 8.8 Hz, 4H), 7.00 (d,
J = 8.6 Hz, 2H), 7.01 (d, J = 11.7 Hz, 1H), 7.08-7.12 (m, 1H),
7.21-7.30 (m, 14H), 7.35 (d, J = 8.8 Hz, 4H), 7.59 (d, J = 7.6
Hz, 1H); 13C NMR (100 MHz, CDCl3) δ 12.9, 19.1, 115.8, 120.6,
122.1, 123.0, 124.9, 125.5, 125.8, 127.4, 127.5, 127.8, 128.2,
128.8, 130.6, 131.0, 131.2, 132.4, 133.4, 135.6, 136.8, 137.9,
139.7, 142.7, 143.3 144.1, 146.0, 146.2, 146.7, 154.8 ppm.
4-Bromo-N-(4-bromophenyl)-N-{4-((1E)-1-[(1E)-1-((triisopro-
pylsilyl)methylene)-1H-inden-2-yl]-4,4-diphenylbuta-1,3-dienyl)-
phenyl}benzenamine.
C
44H3181Br79BrN 733.081, found 733.080.
Acknowledgment. We are very much indebted to the
Deutsche Forschungsgemeinschaft and the Fonds der Che-
mischen Industrie for support of our research. Work per-
formed in Glasgow was supported by the Glasgow Centre for
Physical Organic Chemistry as funded by the EPSRC. G.B.
is grateful for this support. This publication is dedicated
€
to Prof. Dr. C. Ruchardt on the occasion of his 80th
birthday.
Bis(4-bromophenyl){4-[(1E)-[(1E)-1-[(triisopropylsilanyl)-
methylene]-1H-inden-2-yl][(2-phenylcyclopropylidene)methyl]-
phenyl}benzenamine (8): H NMR (400 MHz, CDCl3) δ 1.01-
1.03 (m, 18H), 1.15-1.21 (m, 2H), 1.31-1.39 (m, 3H), 1.86 (t,
Supporting Information Available: Experimental proce-
dures for compounds 4, 5, 12H, 12D, and 13, 1H, 13C, and
HRMS spectra for all compounds, LFP results; and details of
the computation. This material is available free of charge via the
Internet at http://pubs.acs.org.
1
(43) TURBOMOLE V5-9-1, University of Karlsruhe, 2007.
5860 J. Org. Chem. Vol. 74, No. 16, 2009