which became dominant after several more days. An
interesting feature of this new species was the triple set of
signals, which offered the first hint of an assembly with
three tetrads or a 3T-SGQ (Figures 2b, 3). The question
remained, however, as to what type of 3T-SGQ was it.
The peak pattern of the NMR of the new species
(Figures 2b and S19),11 and the observed mass of m/z
2613.6326 (Figures 2 and S20), led us to hypothesize that
the isomerization toward cis-1 induced the formation of a
dodecamer. This possibility was quickly dismissed due to
1
the near absence of cis-1 in the H NMR in DMSO-d6
(a solvent that disrupts the assembly), in addition to the
observed 1:1 ratio of 1 with a new species (2) that showed
a double set of signals (Scheme 1, Figure 3b). Further
HPLC-MS analysis of 2 demonstrated it to have twice the
mass of 1 (Figures S6, S14) supporting its assignment as a
covalent dimer (Scheme 1). Further NMR experiments
with2 confirmed the presence of a cyclobutyl moiety, likely
formed via a [2 þ 2] photocycloaddition between the
double bonds in the chalconyl moiety.
The reason why irradiation of 116 enables a [2 þ 2]
cycloaddition reaction is the increased proximity of four
pairs of chalconyl double bonds in the supramolecule
(Figure 3a).12,13 The formation of a single diastereomer
of 2 and the negligible formation of cis-1 are indicative of
the kinetic stability, well-defined spatial orientation of the
double bonds, and the relatively crowded environment
within 116. Molecular modeling studies suggest the stereo-
chemistry around the cyclobutyl moiety in 2 is the so-called
δ-truxinate14 due to the relative orientation of the olefins
prior to the dimerization (Figure S26). Minimization of the
dimer by itself, followed by its superposition over the
structure of 1424, shows minimal deviations in the overall
geometry of the subunits (Figure S26). Those small devia-
tions, however, are likely responsible for the reluctance of
2 to form a homomeric octamer due to its reduced flex-
ibility relative to 1. The association of two different tetrads
(14 þ 24) provides enough flexibility to accommodate the
increased rigidity of the subunits of 2.
Figure 2. Phototriggered transformation of 116 into 1424 as a
function of time. (a) Partial 1H NMR (500 MHz, CD3CN,
0.5 equiv KSCN, 298 K) showing the region corresponding to
the N1H peaks. (b) Cartoon depiction of the corresponding
SGQs [116•3K]3þ and [1424•3K]3þ; the measured and calculated
isotope patterns obtained from ESI-MS are included above and
below (respectively) the corresponding cartoon depictions.
After irradiation, many of the potential supramolecular
states, including, for example, the heteromeric dodecamer
1824 or the homomeric octamer 28, are not favored.15 This
islikely acombination ofthe detrimental strainimposed by
the cyclobutyl moiety combined with the putative entropic
advantage of a heteromeric assembly over a homomeric
one. More specifically, isolated dimer 2 (with 0.5 equiv of
KSCN in CD3CN) self-assembles with very poor fidelity
leading to a complex mixture of assemblies (Figure S24),
yet, adding 1 equiv of 1 leads to the relatively slow
formation of 1424 (Figure S25). The many transient peaks
observed right after the irradiation of 116 (Figure 2;
Figures S18, S19), and up until the new equilibrium is
established, points to a rough energy landscape (thus, the
relatively long equilibration process) containing multiple
metastable assemblies (i.e., kinetic traps) where 1424 re-
presents a significantly deeper well to shift the equilibrium
with high fidelity.16 The proportion of subunits (1:1) is
precisely determined by not only the information encoded
ꢀ
´
(11) (a) Rivera-Sanchez, M. d. C.; Andujar-de-Sanctis, I.; Garcıa-
Arriaga, M.; Gubala, V.; Hobley, G.; Rivera, J. M. J. Am. Chem. Soc.
2009, 131, 10403–10405. (b) Gonzalez-Rodriguez, D.; van Dongen,
J. L. J.; Lutz, M.; Spek, A. L.; Schenning, A. P. H. J.; Meijer, E. W.
Nat. Chem. 2009, 1, 151–155.
(12) The [2 þ 2] cycloaddition in solution is not usually the favoured
dissipative process after photoexcitation (heat dissipation, transꢀcis
isomerization, and emission being usually favoured). And while still
relatively rare, relative to examples [2 þ 2] cycloaddition reactions
can occur in crystalline environments: (a) Bhogala, B.; Captain, B.;
Parthasarathy, A.; Ramamurthy, V. J. Am. Chem. Soc. 2010, 132,
ꢀ
13434–13442. (b) MacGillivray, L. R.; Papaefstathiou, G. S.; Friscic,
T.; Hamilton, T. D.; Bucar, D.-K.; Chu, Q.; Varshney, D. B.; Georgiev,
I. G. Acc. Chem. Res. 2008, 41, 280–291. (c) Yang, S.-Y.; Naumov,
P. e.; Fukuzumi, S. J. Am. Chem. Soc. 2009, 131, 7247–7249.
(13) (a) Ushakov, E. N.; Vedernikov, A. I.; Alfimov, M. V.; Gromov,
€
S. P. Photochem. Photobiol. Sci. 2011, 10, 15–18. (b) Svoboda, J.; Konig,
B. Chem. Rev. 2006, 106, 5413–5430. (c) Mizoguchi, J.-i.; Kawanami, Y.;
Wada, T.; Kodama, K.; Anzai, K.; Yanagi, T.; Inoue, Y. Org. Lett. 2006,
8, 6051–6054. (d) Iwaura, R.; Shimizu, T. Angew. Chem., Int. Ed. 2006,
45, 4601–4604.
(14) Although the double bonds in 1 could be positioned in a relative
geometry that would lead to the R-truxilate, analysis of the resulting
mixed octamer of 1424 made with it revealed many detrimental interac-
tions that would prevent its formation. McClenaghan, N. D.; Bassani,
D. M. Int. J. Photoenergy 2004, 6, 185–192.
(15) The peak at m/z = 3471 could correspond to [116•3K]3þ or
[1824•3K]3þ (Figure S20), which leaves open the possibility that the latter
(a mixed dodecamer) could form at least in the gas phase as well as in
small amounts in solution.
(16) Furlan, R. L. E.; Otto, S.; Sanders, J. K. M. Proc. Natl. Acad.
Sci. U.S.A. 2002, 99, 4801–4804.
2352
Org. Lett., Vol. 15, No. 10, 2013