Photochemical studies on ladderane formation from conjugated esters in solution or solid phase

Article abstract of DOI:10.1016/j.tetlet.2006.06.066

This letter describes a combined photochemical and X-ray crystallographic study of routes for the synthesis of cyclobutane or ladderane structures by a [2+2]-cycloaddition pathway and leads to a clearer definition of the 3-D structural requirements for su

Full text of DOI:10.1016/j.tetlet.2006.06.066

Tetrahedron Letters 47 (2006) 5879–5882
Photochemical studies on ladderane formation from
conjugated esters in solution or solid phase
Vincent Mascitti and E. J. Corey*
Department of Chemistry and Chemical Biology, Harvard University, 12 Oxford Street, Cambridge, MA 02138, United States
Received 31 May 2006; revised 6 June 2006; accepted 12 June 2006
Available online 3 July 2006
Abstract—This letter describes a combined photochemical and X-ray crystallographic study of routes for the synthesis of cyclobu-
tane or ladderane structures by a [2+2]-cycloaddition pathway and leads to a clearer definition of the 3-D structural requirements
for such processes in the solid state.
Ó 2006 Elsevier Ltd. All rights reserved.
Recently, we reported the first enantioselective total syn-
thesis1 of the unusual microbial fatty acid pentacyclo-
anammoxic acid 1: a molecule that harbors a unique
carbon framework composed of five cis-fused cyclo-
butane rings (referred to as a [5]-ladderane). In connection
with this work, we conducted some exploratory studies
to test the feasibility of constructing the complete [5]-
ladderane unit of this molecule based on photoirradia-
tion of easily accessible polyenes in solution and/or in
the solid phase. This letter reports some of our prelimin-
ary results in this arena.
cent double bonds are further away from each other
(>4.7 A) and are no longer properly aligned. Moreover,
in solution only trans/cis isomerization occurs.
˚
Following this initial report, the dimerization of cin-
namic acid derivatives has been extensively studied
and the concept of locking two olefinic systems in a spe-
cial arrangement favoring a [2+2]-photocycloaddition
has been extended to the synthesis of ladderane motifs
(i.e., an assembly of cis-fused cyclobutane rings) starting
from polyolefinic substrates. Recent work reported by
Hopf,3 who employed a polyunsaturated substrate rig-
idly held by a [2,2]-paracyclophane scaffold in solution,
and by MacGillivray4 with a polyene in the solid state
are good examples.
O
OH
We started our investigations by studying the UV-irradi-
ation5 of dienoate 2 (Scheme 1), easily synthesized in
quantitative yield via Horner–Wadsworth–Emmons
reaction between m-anisaldehyde and methyl-4-(di-
methylphosphono)crotonate in THF at À50 °C using
1
pentacycloanammoxic acid
(+)-enantiomer
The pioneering study of Schmidt over 50 years ago on
the solid-state photodimerization of trans-cinnamic acid
showed that when crystal packing placed double bonds
of neighboring molecules in proximity, a [2+2]-photo-
cycloaddition occurred upon irradiation.2 Thus, in the
crystalline a- and b-forms of the acid a solid-state
photo[2+2]-cycloaddition occurs, involving face-to-face
MeO
MeO
MeO2C
solid state
O
˚
UV-irradiation
double bonds at a distance of 4.1 A. However, in the
c-form no photodimerization occurs since the two adja-
CO2Me
OMe
MeO
63%
2
3
*
Corresponding author. Tel.: +1 617 495 4033; fax: +1 617 495
0376; e-mail: corey@chemistry.harvard.edu
Scheme 1.
0040-4039/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2006.06.066

5880
V. Mascitti, E. J. Corey / Tetrahedron Letters 47 (2006) 5879–5882
NaHMDS as a base. When a sample of 2 (mp = 62–64 °C,
recrystallized from Et2O–hexanes) was placed between
two layers of PyrexÒ glass and irradiated with ultravio-
let light from a Hanovia medium-pressure Hg lamp, the
reaction proceeded cleanly to produce the monocyclo-
butane head-to-tail derivative 3 (Scheme 1) in 63%
isolated yield (80% conversion after 7 h).6 The formation
of 3 correlates with the X-ray crystal structure of 2 that
is shown in Figure 1. The double bonds that react to
form 3 are close to one another (C–C distances: 3.9
4.26 Å
4.35 Å
4.41 Å
Figure 2. Arrangement of two closest neighbors in the crystal structure
of 4.
˚
and 3.96 A) and are arranged in a head-to-tail manner
with only a slight offset. Encouraged by this result, we
then investigated the solid-state UV-irradiation of tri-
enoate 4 (mp = 114 °C, recrystallized from EtOAc/
heptane7 in the solid state). Unfortunately, the UV-irra-
diation of this adduct (Scheme 2) in the solid state,
either as suspension in water or in dry form (finely
ground crystals placed between two layers of PyrexÒ
glass), led solely to polymeric material as evidenced by
1H NMR.
4.71 Å
4.75 Å
4.75 Å
Although initially unexpected, these results were readily
rationalized once we obtained the X-ray crystal struc-
ture of 4. That structure (see Fig. 2) clearly indicated
that the two closest neighbors in the lattice are not only
˚
too far away from each other (4.3 A) to do a [2+2]-
cycloaddition, but are also not properly oriented. This
result emphasizes the importance of a face-to-face paral-
lel stacking arrangement rather than a criss-cross orien-
tation of double bonds for solid-state photoaddition.
Consequently, we reasoned that the formation of a
dienoate such as 5 (Fig. 3; mp = 102–104 °C, recrystal-
lized from EtOAc–heptane) might favor the orientation
of the two triene units in an arrangement where an intra-
molecular solid-state [2+2]-photocycloaddition could
occur. Although 5 was readily synthesized, irradiation
of the solid between two layers of PyrexÒ glass for
10 h failed to produce any of the desired ladderane skel-
eton and only a chloroform insoluble material (presum-
O
MeO
O
t-Bu
O
MeO
O
5
Figure 3. Arrangement of two triene units in the crystal structure of 5.
ably polymeric) could be recovered. The reason for this
lack of intramolecular photocyclization was clarified
once again by the X-ray crystal structure of 5, which
showed that the two triene units are too far away
˚
(4.7 A) and are also not well aligned (Fig. 3).
In order to potentially reduce the intramolecular
distance between the two triene units, we decided to
investigate the feasibility of an intramolecular [2+2]-
photocycloaddition of the olefinic anhydride 6 (Scheme
3). We were pleased to observe that when a dilute
solution of 6 (this compound is an oil) in MeCN was
irradiated with UV light at 23 °C for 20 h, a [2+2]-cyclo-
addition occurred to produce 7 in over 90% yield.8
Unfortunately, UV-irradiation of the polyolefinic anhy-
dride 8 (an oil) in MeCN led to only isomerization of the
double bonds and formation of polymeric material.8
3.90 Å
3.96 Å
Figure 1. Arrangement of two closest neighbors in the crystal structure
of 2.
Ph
Ph
Another option for aligning the two triene units in a
suitable arrangement for an intramolecular [2+2]-
photocycloaddition was investigated using the 26-mem-
bered macrocyclic structure 9 (Scheme 4). This test
substrate (9) was synthesized in six steps starting from
methyl sorbate as summarized below. All the steps pro-
solid state
Ph
O
CO2Me
CO2Me
UV-irradiation
OMe
4
Scheme 2.

V. Mascitti, E. J. Corey / Tetrahedron Letters 47 (2006) 5879–5882
5881
O
O
O
MeO
MeO
MeO
hν, pyrex filter
O
O
5.17 Å
23 ºC, MeCN
5.26 Å
MeO
90%
O
5.21 Å
6
7
Ph
Ph
Figure 4. Arrangement of two triene units in the crystal structure of 9.
O
hν, pyrex filter
23 ºC, MeCN
O
O
O
O
roform solvate is shown in Figure 4. In this case too,
even though the two p-systems are somewhat parallel
to each other they are unfortunately not well aligned
O
8
˚
Scheme 3.
and are too distant from one another (5.2 A) for intra-
molecular photocyclization to occur without major dis-
ruption of crystal packing. Such disruption is likely
very slow as compared to relaxation of the excited sub-
strate to the ground state (likely to be very fast in the
solid phase).
O3, CH2Cl2–MeOH
O
O
O
–30 ºC
>90%
OMe
OMe
55%
1. Ph3P=C(H)CHO
CH2Cl2, reflux
The studies described herein provide useful guidelines
on future approaches to the synthesis of ladderanes by
a poly-[2+2]-cycloaddition pathway. It is clear that X-
ray crystallographic analysis can serve as a useful screen
for promising substrates in solid-state photoreactions.
What is still required for the efficient application of this
approach is a methodology for crystal engineering,9,10
which in this context, means the prediction of crystal
packing as a function of substrate structure. It is also
apparent that photochemical [2+2]-cycloaddition strate-
gies are only workable when the olefinic units of the
substrate(s) are properly positioned relative to one
another.11
2. (MeO)2P(O)CH2CO2Me
LiCl, DBU, MeCN, 0 ºC
OMe
Cl
O
1. LiOH, THF–H2O
23 ºC to reflux
O
O
O
2. (COCl)2, DMF
THF, reflux
OMe
Cl
77%
HO
HO
t-Bu
Pyridine, THF
reflux
5%
O
O
O
O
t
-Bu
O
O
t
-Bu
O
References and notes
O
9
1. (a) Mascitti, V.; Corey, E. J. J. Am. Chem. Soc. 2006, 128,
3118; (b) Mascitti, V.; Corey, E. J. J. Am. Chem. Soc.
2004, 126, 15664.
Scheme 4. Synthesis of macrocycle 9.
2. (a) Cohen, M. D.; Schmidt, G. M. J. J. Chem. Soc. 1964,
1996, 2000, 2014; (b) Bassani, D. M. In The dimerization of
cinnamic acid derivatives; 2nd ed.; Horspool, W. M., Lenci,
F., Eds., CRC Handbook of Organic Photochemistry and
Photobiology; CRC Press, 2004; pp 1–20, Section 20; (c)
Lahav, M.; Schmidt, G. M. J. Tetrahedron Lett. 1966, 26,
2957.
3. Hopf, H.; Greiving, H.; Jones, P. G.; Bubenitschek, P.
Angew. Chem., Int. Ed. Engl. 1995, 34, 685.
4. Gao, X.; Friscic, T.; MacGillivray, L. R. Angew. Chem.,
Int. Ed. 2004, 43, 232.
5. All the UV irradiations reported in this letter were done
using a Hanovia 450 W medium pressure Hg lamp
equipped with a PyrexÒ cooling jacket.
6. When 2 was suspended in water, the solid stayed in
solution during the irradiation and we observed the clean
formation of 3. However, the reaction was slower (40%
conversion after 7 h). Irradiation of 2 in hexanes led only
to isomerization of the double bonds.
7. Compound 4 was synthesized in one step and 40% yield by
a Horner Wadsworth Emmons reaction starting from
trans-cinnamaldehyde and methyl-4-(dimethylphosphono)-
crotonate.
ceeded in high yield except for the macrocyclization
steps between the bis acid chloride and the diol, which
produced the desired product 9 in 5% isolated yield.
Although the last step was not optimized, the yield
was sufficient to provide enough material to test the
critical UV irradiation step, and 9 provided the bonus
of being a solid. Unfortunately UV irradiation of 9,
either in solution or in the solid state, failed to produce
the desired [5]-ladderane motif. Instead, in addition to
the formation of some polymer material, we observed
1
by H NMR the formation of a complex mixture of
polyolefinic products containing numerous olefinic pro-
ton peaks.
After extensive screening of different solvents, we
found that 9 could be recrystallized from CHCl3–hep-
tane to produce single crystals suitable for X-ray
diffraction analysis (mp not determined since at tem-
peratures as high as 280 °C the solid shrank but did
not melt). The X-ray crystal structure of 9 as the chlo-

5882
V. Mascitti, E. J. Corey / Tetrahedron Letters 47 (2006) 5879–5882
8. As measured by NMR analysis of the total reaction
10. Ichikawa, M.; Takahashi, M.; Aoyagi, S.; Kibayashi, C.
J. Am. Chem. Soc. 2004, 126, 16553.
product.
9. (a) Liu, J.; Boarman, K. J. Chem. Commun. 2005, 340; (b)
Liu, J.; Wendt, N. L.; Boarman, K. J. Org. Lett. 2005, 7,
1007.
11. The X-ray crystal structures of 2, 4, 5, and 9 have
been deposited with the Cambridge Crystallographic
database.