Intramolecular photosensitization of the pinene-ocimene rearrangement

Article abstract of DOI:10.1139/v03-077

Bonding of nopol to the para position of acetophenone produces 5,5-dimethyl-2-(2-(p-acetylphenoxy)ethyl)bicyclo[3.1.1]hept-2-ene 1, which contains two chromophores: a para-alkoxyacetophenone and an α-pinene, connected by a single methylene group. UV irrad

Full text of DOI:10.1139/v03-077

669
Intramolecular photosensitization of the pinene–
ocimene rearrangement
Kevin McMahon and Peter J. Wagner
Abstract: Bonding of nopol to the para position of acetophenone produces 5,5-dimethyl-2-(2-(p-acetylphenoxy)ethyl)bi-
cyclo[3.1.1]hept-2-ene 1, which contains two chromophores: a para-alkoxyacetophenone and an α-pinene, connected by
a single methylene group. UV irradiation of 1 in both benzene and methanol produces none of the intramolecular
[2 + 2] cycloaddition that most para-(3-buten-1-oxy)acetophenones undergo. Instead, the pinene unit rearranges to a
triene skeleton identical to that of ocimene, a known photoproduct of pinene. At modest conversion the diene portion
of the triene is cis but gradually is converted to a 52:48 trans:cis ratio. It is concluded that intramolecular triplet energy
transfer from the excited ketone chromophore forms the 1,2-biradical triplet state of the pinene moiety, which then un-
dergoes cyclobutylcarbinyl ring opening to a 1,4-biradical that cleaves to the 1,3,6-triene structure of ocimene. This
mechanism is suggested to be responsible for the earlier reported intermolecularly sensitized rearrangement of α-pinene
to the ocimene isomers.
Key words: intramolecular energy transfer, triplet pinene, cyclobutylcarbinyl ring opening, photosensitization.
Résumé : La fixation du nopol en position para de l’acétophénone produit le 5,5-diméthyl-2-(2-(p-acétylphé-
noxy)éthyl)bicyclo[3.1.1]hept-2-ène 1 qui contient deux chromophores : la para-alkoxyacétophénone et l’α-pinène, re-
liés par un seul groupe méthylène. L’irradiation UV du composé 1 dans le benzène et dans le méthanol ne donne
aucun des produits de cycloaddition intramoléculaire [2 + 2] que les para-(3-butén-1-oxy)acétophénones donnent géné-
ralement. L’unité pinène se réarrange de préférence en un triène dont le squelette est identique à celui de l’ocimène, un
photoproduit connu du pinène. A un taux modéré de transformation la portion diène du triène est cis, mais elle se
transforme graduellement pour donner un rapport trans : cis de 52 : 48. On en conclut que le transfert d’énergie intra-
moléculaire du triplet à partir du chromophore cétone excitée forme l’état triplet 1,2-biradical de l’unité pinène qui, par
ouverture subséquente du cycle du cyclobutycarbinyle, donne un 1,4-biradical qui s’ouvre à son tour pour donner la
structure 1,3,6-triène de l’ocimène. On suggère que ce mécanisme est responsable du réarrangement intermoléculaire
sensibilisé de l’α-pinène en isomères de l’ocimène rapporté antérieurement.
Mots clés : transfert d’énergie intramoléculaire, pinène triplet, ouverture du cycle du cyclobutylcarbinyle, photosensibi-
lisation.
[Traduit par la Rédaction] McMahon and Wagner 672
Introduction
wise, as in norbornene, the biradical triplet would have ab-
stracted hydrogen atoms from the solvent. Both authors
noted that the initially formed cis-ocimene undergoes sensi-
tized cis → trans isomerization, gradually forming a nearly
1:1 equilibrium.
This paper describes our discovery of an intramolecular
version of the photosensitized pinene rearrangement and a
simple mechanism for the rearrangement based on the 1,2-
biradical nature of triplet alkenes (3). To test whether intra-
molecular [2 + 2] cycloaddition of remote double bonds to
triplet benzenes (4) works with cyclic alkenes, nopol was at-
tached para to acetophenone to form 1. Irradiation of 1
caused rearrangement of the dimethylbicyclo[3.1.1]heptene
ring to a mixture of the cis- and trans-isomers of 3-(2-(p-
Many years ago two papers were published reporting the
acetophenone-photosensitized isomerization of α-pinene to
cis- and trans-ocimene (Scheme 1) (1, 2). Both authors sug-
gested triplet energy transfer from the ketone as the initiator
of pinene’s rearrangement; but given the incomplete under-
standing of excited-state reactivity in those years, neither
could provide a compelling mechanism describing how trip-
let pinene rearranges. Frank suggested the possibility of
thermal rearrangement following intersystem crossing of the
triplet pinene to a vibrationally excited ground state, a then
relatively popular notion already subject to suspicion. Kropp
merely noted that rearrangement must be facile since other-
Received 20 March 2003. Published on the NRC Research Press Web site at http://canjchem.nrc.ca on 17 June 2003.
This paper is dedicated to Don Arnold for his many important contributions to chemistry in North America, and just as important,
for his long-time friendship.
K. McMahon1 and P.J. Wagner.2 Chemistry Department, Michigan State University, East Lansing, MI 48824, U.S.A.
1Present address: Department of Chemistry and Biochemistry, Carroll College, Waukesha, WI 53186, U.S.A.
2Corresponding author (e-mail: wagnerp@msu.edu).
Can. J. Chem. 81: 669–672 (2003)
doi: 10.1139/V03-077
© 2003 NRC Canada

670
Can. J. Chem. Vol. 81, 2003
Scheme 2. Photoinduced rearrangement of compound 1.
Scheme 1. Photosensitized rearrangement of α-pinene.
acetylphenoxy)ethyl)-7-methylocta-1,3,6-triene, the same
transformation that α-pinene itself undergoes (Scheme 2).
Results
ference between the cis and trans isomers, most notably the
0.5 ppm difference between the H2 peaks (see Experimen-
tal.)
Compound 1 was prepared via SN2 reaction of the tosylate
of nopol with potassium p-acetylphenoxide in DMF as de-
scribed in the Experimental section.
Discussion
This experiment has revealed that triplet energy transfer
from the excited acylbenzene chromophore to a tethered
alkene can occur at the expense of possible cycloaddition.
That such energy transfer could occur was not unexpected,
given the early bimolecular examples, although no such total
quenching of cycloaddition occurs for acetophenone with a
para 3-methyl-3-penten-1-oxy tether (5), which also has a
trialkyl-substituted double bond. Caldwell and co-workers’
(6) measurement of the triplet energies of various olefins in-
dicate an ET of 77 kcal/mol for 2-methyl-2-butene, but only
75 kcal/mol for 1-methylcyclohexene, a better model for α-
pinene. Both ETs are higher than that of p-acetylanisole
(~71 kcal/mol (7)), so any energy transfer is quite endother-
mic. The apparent lack of much if any energy transfer to the
trisubstituted acyclic double bond undoubtedly reflects its
higher endothermicity. We should note that the lowest triplet
of p-alkoxy alkanophenones is π,π*, whereas acetophenone
itself has a n,π* lowest triplet. However, there appear to be
only slight differences between n,π* and π,π* ketones in rate
constants for intramolecular energy transfer (8).
It must be noted that we cannot distinguish quantitatively
between intermolecular and intramolecular energy transfer
in the case of 1, although its low concentration should have
minimized any intermolecular interactions. The 4 kcal
endothermicity of energy transfer to pinene would be ex-
pected to slow down a bimolecular process to a rate constant
~1 × 107 M–1 s–1 (9). Intramolecular reactions usually enjoy
rate enhancement, but it has been shown that compounds
with a three-atom chain between triplet donor and acceptor
have an effective molarity near 1, with very similar first- and
second-order rate constants (10). That being the case, energy
transfer to the pinene moiety should be only about 10% as
fast as the ~1 × 108 s–1 rate constant for [2 + 2] cyclo-
addition by acylbenzenes with simple acyclic alkenoxy teth-
ers (4). The lack of cycloaddition in 1 then probably
involves some form of steric hindrance to proximity between
the pinene unit and the benzene ring, either at the charge
transfer stage or at the C—C bonding stage.
1 was irradiated in the same way that other p-alkenoxy-
acetophenones had been (4). Preliminary experiments were
run in deaerated NMR tubes containing benzene-d6 or meth-
anol-d4 solutions 0.02 M in both 1 and methyl benzoate (as
an internal standard). Similar results were obtained in both
solvents. NMR spectra at various conversions revealed no
significant changes in the alkoxyacetophenone portion of the
molecule, and the UV spectrum of the product mixture was
dominated by an intense peak at 270 nm characteristic of p-
alkoxyphenyl ketone chromophores. However, changes in
the NMR spectrum for the pinene portion of the molecule
were consistent with the formation of an ocimene-like triene.
Only tiny traces of peaks ascribable to other unidentified
products were noted, but it was obvious that no significant
amount of cycloaddition of the pinene double bond to the
benzene ring occurred. As with the early studies on pinene
itself, the cis-ocimene unit was the only product at low con-
versions but it gradually equilibrated to a 48:52 cis–trans ra-
tio in 73% combined yield at 84% conversion of starting 1.
The products of larger scale reactions were purified by pre-
parative TLC. That the pinene portion of 1 was no longer
present was indicated by the lack of significant optical activ-
ity in the product solutions at high conversion.
The conversion of the pinene moiety to a triene was dem-
onstrated by the addition of four vinyl proton peaks (6.8 or
6.3 and 5.30–5.06 ppm) to the single vinyl proton peak at
5.34 ppm as well as the increase in 13C vinyl peaks (110–
122 ppm) from two to six. The chemical shifts and coupling
constants of the nonaromatic protons correspond very
closely to those reported for ocimene (2), including the dif-
Ab initio computations at the 3-21G and 6-31G*/UHF
levels affirm the 1,2-biradical character of triplet α-pinene,
with the two p-like SOMOs at a roughly 45° angle and the
one on the carbon next to the bridgehead nearly parallel to
© 2003 NRC Canada

McMahon and Wagner
671
the bridgehead—Me2C bond. This alignment should provide
for efficient cleavage of the four-membered ring to yield a
1,4-biradical that can then cleave to two double bonds.
Scheme 3 depicts this chain of events for 1. We suspect that
this same biradical double cleavage occurs in pinene itself.
The ring-opening of cyclobutylcarbinyl radicals to 4-penten-
1-yl radicals was not widely recognized and studied until the
early 1980s, although Scheme 4 shows an early example of
this same ring opening in the benzoyl-peroxide-initiated ad-
dition of carbon tetrachloride to β-pinene (11). The rate con-
stant for cyclobutylmethyl ring opening has been measured
as 4 × 103 s–1 in Australia (12) and 1.5 × 104 s–1 in Canada
(13). A comparable rate constant for opening of the four-
membered ring in triplet α-pinene would seem unable to
compete with measured triplet decay rates in the 1 × 107 s–1
range for alkenes (6). However, in this case radical ring-
opening involves loss of both electronic excitation and
bicyclic ring strain, as well as formation of a tertiary radical
site and thus, may be sufficiently exothermic to be quite
rapid. Whatever the exact kinetics, the likelihood that the re-
arrangement begins with a known radical reaction further
confirms the biradical nature of triplet alkenes (3, 6).
Another possibility is that charge transfer interaction be-
tween the bicycloheptene double bond and the triplet
phenone produces enough radical-cation character in the
pinene to allow ring opening with spread of the radical-
cation character between tertiary and allylic sites. Further
bond cleavage could then generate a diene radical-cation that
collapses to an ocimene structure after back electron trans-
fer. This possibility deserves further study.
Scheme 3. Mechanism for triplet pinene rearrangement.
Scheme 4. Radical cyclobutylcarbinyl ring opening in pinene.
6.00, 1.50 Hz), 1.21 (s, 3H, Me), 1.05 (d, 1H, 8.40 Hz), 0.74
(s, 3H, Me). 13C NMR (CDCl3, 75.5 MHz) δ: 144.63,
142.61, 133.28, 129.77, 127.85, 119.67, 68.57, 45.53, 40.56,
37.99, 36.08, 31.50, 31.27, 26.16, 21.62, 21.04
5,5-Dimethyl-2-(2-(p-acetylphenoxy)ethyl)bi-
cyclo[3.1.1]hept-2-ene (1)
Nopol tosylate (2.00 g, 0.00675 mol), p-hydroxy-
acetophenone (1.38 g, 0.0101 mol), and K2CO3 (2.80 g,
0.0202 mol) were added to DMF (50 mL) and heated to
60°C under argon for 15 h. The mixture was then cooled to
room temperature and water (50 mL) was added. The solu-
tion was extracted with ether (3 × 30 mL). The combined or-
ganic layer was washed with 2 N NaOH (2 × 30 mL) and aq
NaCl (2 × 30 mL), and then dried over MgSO4. After re-
moval of the ether at reduced pressure, the desired product
was obtained without further purification as a colorless oil
(1.35 g, 0.0047 mol, 71%). [α]25°C (CHCl3, c = 0.0273):
–27.56 (589 nm), –28.66 (578 nm), –32.56 (546 nm), –54.77
(436 nm), –84.82 (365 nm). UV (CH3OH) λmax: 275
(14 789), 216 (11 668), 204 (11 965). MS m/z: 284 (<1%),
241 (18), 137 (25), 121 (15), 105 (100), 93 (21), 91 (29), 79
Experimental
Preparation of reactant
Nopol tosylate
(1R)-(–)-Nopol (Aldrich, 20.00 g, 0.11 mol) in pyridine
(32 mL) was cooled to 0°C. Tosyl chloride (13.65 g,
0.0716 mol) was added portion-wise under argon over 1 h.
The mixture was stirred mechanically for a further 5 h at
low temperature. Then H2SO4 (2 N, 100 mL) was added to
the mixture. The resulting solution was extracted using ether
(3 × 100 mL). The combined organic extracts were washed
with 2 N NaOH, NaHCO3, and NaCl. After drying over
MgSO4, the ether was removed by distillation at reduced
pressure. The crude product was purified by dry column
flash chromatography (silica gel, CH2Cl2), to give a color-
less oil (13.33 g, 0.045 mol, 63%). [α]25°C (CHCl3, c =
0.0273): –23.81 (589 nm), –24.94 (578 nm), –28.46
(546 nm), –48.02 (436 nm), –75.46 (365 nm). 1H NMR
(CDCl3, 300 MHz) δ: 7.76 (d, 2H, 8.40 Hz, Tos), 7.32 (d,
2H, 7.80 Hz, Tos), 5.22 (br s, 1H, =CH), 4.01 (t, 2H,
6.90 Hz, O-CH2), 2.43 (s, 3H, Tos), 2.32–2.24 (m, 3H),
2.17–2.15 (m, 2H), 2.04–2.01 (m, 1H), 1.91 (td, 1H, 6.00,
1
(23), 77 (22), 43 (44). H NMR (CDCl3, 300 MHz) δ: 7.90
(d, 2H, 9.00 Hz, H11), 6.88 (d, 2H, 9.00 Hz, H10), 5.34 (br s,
1H, H3), 4.01 (t, 2H, 7.05 Hz, H9), 2.53 (s, 3H, COMe), 2.44
(tdd, 2H, 7.05, 1.2, 1.2 Hz, H8), 2.36 (ddd, 1H, 8.70, 5.62,
5.62 Hz, H5), 2.23 (br d, 2H, 8.40 Hz, H7), 2.08 (br d, 2H,
5.40 Hz, H4), 1.26 (s, 3H, 6-Me), 1.16 (d, 1H, 8.40 Hz, H1),
0.80 (s, 3H, 6-Me). 13C NMR (CDCl3, 75.5 MHz) δ: 196.74,
162.90, 144.17, 130.56, 130.18, 118.90, 114.15, 66.56,
© 2003 NRC Canada

672
Can. J. Chem. Vol. 81, 2003
45.89, 40.73, 38.08, 36.33, 31.64, 31.37, 26.28 21.17. HR-
MS calcd. for C19H24O2: 284.1770; found: 284.1776.
to those reported for ocimene (2). Italicized data again repre-
sent NMR peaks for the aryl ketone portion of the mole-
cules, which remain nearly identical to their analogs in the
reactant.
Irradiation procedures
Samples of 1 were irradiated in both benzene and metha-
nol solutions. Solutions of 1 (0.0047 g, 1.65 × 10–5 mol) and
methyl benzoate (0.0020 g, 1.47 × 10–5 mol) in deuterated
solvent (0.75 mL) were placed in an NMR tube. These were
irradiated at room temperature in a Rayonet reactor with
300 nm lights. After 1 h and 84% conversion of starting
ketone, NMR analysis indicated that the pinene portion of 1
had been converted to two ocimene structures in a 48:52
(cis–trans) ratio in 73% combined yield. This ratio was
based on the relative intensities of the 6.77 and 6.34 ppm
proton peaks. In the early stages of the reaction, only the
cis-isomer was observed. Further irradiation led to equilibra-
tion of the two isomers. No optical rotation was observed for
the product mixture. Larger scale reactions were purified by
PTLC (silica gel, CHCl3).
3-(2-(p-Acetylphenoxy)ethyl)-7-methyl-cis-octa-1,3,6-triene
1H NMR (CDCl3, 300 MHz, COSY) δ: 7.94 (d, 2H,
9.00 Hz), 6.95 (d, 2H, 9.00 Hz), 6.77 (dd, 1H, 17.70,
11.40 Hz, H2), 5.47 (t, 1H, 7.80 Hz, H4), 5.32 (d, 1H,
17.70 Hz, H1), 5.15–5.06 (m, 1H, H6), 5.15 (d, 1H, 9.60 Hz,
H1), 4.13 (t, 2H, 6.90 Hz, O–CH2), 2.88 (dd, 2H, 7.20 Hz,
2H5), 2.69 (t, 2H, 6.90 Hz, 2H7), 2.54 (s, COMe), 1.68 (br s,
Me), 1.63 (br s, Me).
3-(2-(p-Acetylphenoxy)ethyl)-7-methyl-trans-octa-1,3,6-
triene
1H NMR (CDCl3, 300 MHz, COSY) δ: 7.94 (d, 2H,
9.00 Hz), 6.95 (d, 2H, 9.00 Hz), 6.34 (dd, 1H, 17.40,
10.5 Hz, H2), 5.60 (t, 1H, 7.65 Hz, H4), 5.20 (d, 1H,
17.70 Hz, H1), 5.15–5.06 (m, 1H, H6), 4.97 (d, 1H,
10.80 Hz, H1), 4.09 (t, 2H, 7.20 Hz, O–CH2), 2.90 (t, 2H,
7.20 Hz, 2H5), 2.80 (t, 2H, 6.90 Hz, 2H7), 2.54 (s, COMe),
1.68 (br s, Me), 1.63 (br s, Me).
Acknowledgment
This work was supported by National Institute of Health
(U.S.A.) grant GM-39821.
The isomer mixture obtained at full conversion was ana-
lyzed by 13C NMR, UV, and MS. Italicized data represent
the NMR peaks for the aryl ketone portion of the molecules,
which remain nearly identical to their analogs in the reac-
tant. The mass spectrum indicated that the products are iso-
mers of the reactant 1, while an intense 270 nm UV peak
indicated that the alkoxyacetophenone portion of the mole-
cule survived irradiation unscathed.
References
1. G. Frank. J. Chem. Soc. B, 130 (1968).
2. P.A. Kropp. J. Am. Chem. Soc. 91, 5783 (1969).
3. R.A. Caldwell. Pure Appl. Chem. 56, 1167 (1984); R.A.
Caldwell and L. Zhou. J. Am. Chem. Soc. 116, 2271 (1994).
4. P.J. Wagner. Acc. Chem. Res. 32, 1 (2001), and refs. therein.
5. P.J. Wagner and K.-L. Cheng. Tetrahedron Lett. 34, 907
(1993).
6. T. Ni, R.A. Caldwell, and L.A. Melton. J. Am. Chem. Soc.
111, 457 (1989).
7. P.J. Wagner, A.E. Kemppainen, and H.N. Schott. J. Am. Chem.
Soc. 95, 5604 (1973).
3-(2-(p-Acetylphenoxy)ethyl)-7-methyl-cis-octa-1,3,6-triene
and 3-(2-(p-acetylphenoxy)ethyl)-7-methyl-trans-octa-
1,3,6-triene
UV (CH3OH) λmax: 270. MS m/z: 284 (6), 241 (7), 137
(20), 133 (48), 121 (53), 105 (100), 79 (62), 77 (56), 69
(27), 55 (27), 43 (73). 13C NMR (CDCl3, 75.5 MHz, DEPT)
δ: 196.62, 162.83, 162.77, 139.86 (CH), 134.60 (CH),
133.38, 132.53, 132.40 (CH), 131.77, 131.69 (CH), 130.50
(CH), 130.49 (CH), 130.21, 130.16, 121.97 (CH), 121.71
(CH), 114.16 (CH), 114.07 (CH), 113.64 (CH2), 110.81
(CH2), 67.34 (CH2), 66.61 (CH2), 32.90 (CH2), 27.43 (CH2),
26.57 (CH2), 26.24 (CH3), 26.11 (CH2), 25.61 (CH3), 17.75
(CH3), 17.72 (CH3).
8. P.J. Wagner and P. Klán. J. Am. Chem. Soc. 121, 9625 (1999).
9. P.J. Wagner. J. Am. Chem. Soc. 89, 2820 (1967); P.J. Wagner
and I. Kochevar. J. Am. Chem. Soc. 90, 2232 (1968); P.J. Wag-
ner. Mol. Photochem. 1, 71 (1969).
10. G.L. Closs, P. Piotrowiak, J.M. MacInnis, and G.R. Fleming. J.
Am. Chem. Soc. 110, 2652 (1988).
11. D.M. Oldroyd, G.S. Fisher, and L.A. Goldblatt. J. Am. Chem.
Soc. 72, 2407 (1950).
12. A.L.J. Beckwith and G. Moad. J. Chem. Soc., Perkin Trans. 2,
1083 (1980).
1H NMR spectra of the individually isolated isomers were
measured to compare chemical shifts and coupling constants
13. K.U. Ingold, B. Maillard, and J.C. Walton. J. Chem. Soc.,
Perkin Trans. 2, 970 (1981).
© 2003 NRC Canada