Templation of the excited-state chemistry of α-(n-alkyl) dibenzyl ketones: how guest packing within a nanoscale supramolecular capsule influences photochemistry

Article abstract of DOI:10.1021/ja7107917

Excited-state behavior of eight α-alkyl dibenzyl ketones (alkyl = CH₃ through n-C₈H₁₇) that are capable of undergoing type II and/or type I photoreactions has been explored in isotropic solution and within a water-soluble capsule. The study consisted of two parts: photochemistry that explored the excited-state chemistry and an NMR analysis that revealed the packing of each guest within the capsule. The NMR data (COSY, NOESY, and TOCSY experiments) revealed that ternary complexes between α-alkyl dibenzyl ketones and the capsule formed by two cavitands are kinetically stable, and the guests fall into three packing motifs modulated by the length of the α-alkyl chain. In essence, the host is acting as an external template to promote the formation of distinct guest conformers. The major products from all eight guests upon irradiation either in hexane or in buffer solution resulted from the well-known Norrish type I reaction. However, within the capsule the excited-state chemistry of the eight ketones was dependent on the alkyl chain length. The first group consisted of α-hexyl, α-heptyl, and α-octyl dibenzyl ketones that yielded large amounts of Norrish type II products within the host, while in solution the major products were from Norrish type I reaction. The second group consists of α-butyl and α-pentyl dibenzyl ketones that yield equimolar amounts of two rearranged starting ketones within the capsule (combined yield of ca 60%), while in solution no such products were formed. The third group consisted of α-methyl, α-ethyl, and α-propyl dibenzyl ketones that within the capsule yielded only one (not two) rearranged starting ketone in larger amounts (21-35%) while in solution no rearrangement product was obtained. Variation in the photochemistry of the guest within the capsule, with respect to the α-alkyl chain length of the guest, highlights the importance of how a small variation in supramolecular structure can influence the selectivity within a confined nanoscale reactor.

Full text of DOI:10.1021/ja7107917

Published on Web 03/06/2008
Templation of the Excited-State Chemistry of r-(n-Alkyl)
Dibenzyl Ketones: How Guest Packing within a Nanoscale
Supramolecular Capsule Influences Photochemistry
Corinne L. D. Gibb,‡ Arun Kumar Sundaresan,† V. Ramamurthy,*,† and
Bruce C. Gibb*,‡
Department of Chemistry, UniVersity of Miami, Coral Gables, Florida 33124, and Department
of Chemistry, UniVersity of New Orleans, New Orleans, Louisiana 70148
Received December 3, 2007; E-mail: murthy1@miami.edu
Abstract: Excited-state behavior of eight R-alkyl dibenzyl ketones (alkyl ) CH3 through n-C8H17) that are
capable of undergoing type II and/or type I photoreactions has been explored in isotropic solution and
within a water-soluble capsule. The study consisted of two parts: photochemistry that explored the excited-
state chemistry and an NMR analysis that revealed the packing of each guest within the capsule. The
NMR data (COSY, NOESY, and TOCSY experiments) revealed that ternary complexes between R-alkyl
dibenzyl ketones and the capsule formed by two cavitands are kinetically stable, and the guests fall into
three packing motifs modulated by the length of the R-alkyl chain. In essence, the host is acting as an
external template to promote the formation of distinct guest conformers. The major products from all eight
guests upon irradiation either in hexane or in buffer solution resulted from the well-known Norrish type I
reaction. However, within the capsule the excited-state chemistry of the eight ketones was dependent on
the alkyl chain length. The first group consisted of R-hexyl, R-heptyl, and R-octyl dibenzyl ketones that
yielded large amounts of Norrish type II products within the host, while in solution the major products were
from Norrish type I reaction. The second group consists of R-butyl and R-pentyl dibenzyl ketones that yield
equimolar amounts of two rearranged starting ketones within the capsule (combined yield of ca 60%),
while in solution no such products were formed. The third group consisted of R-methyl, R-ethyl, and R-propyl
dibenzyl ketones that within the capsule yielded only one (not two) rearranged starting ketone in larger
amounts (21-35%) while in solution no rearrangement product was obtained. Variation in the photochemistry
of the guest within the capsule, with respect to the R-alkyl chain length of the guest, highlights the importance
of how a small variation in supramolecular structure can influence the selectivity within a confined nanoscale
reactor.
soluble, deep cavity cavitand is used as a reaction vessel.5 In
Introduction
the presence of a suitable guest (or guests), two molecules of 1
The remark that Hammett made almost seven decades ago,
that “a major part of the job of the chemist is the prediction
and control of the course of reactions”, still remains true
although the techniques employed to “predict and control”
reactions have changed.1 In recent times, the use of confined
media to control photoreactions has gained importance,2-4 and
we present results of one such investigation where a water-
assemble in aqueous solution to form supramolecular nanocap-
sules.6 These capsular complexes, in which the templating guest
or guests reside within an essentially dry interior, are held
together by weak interactions such as van der Waals and CH-π
interactions and the hydrophobic effect. Capsular assembly and
disassembly are slow on the NMR time scales, and therefore in
the time scale of excited-state reactions they remain essentially
intact.7
† University of Miami.
‡ University of New Orleans.
(1) (a) Hammett, L. P. Physical Organic Chemistry, Reaction Rates, Equilibria
and Mechanisms; McGraw-Hill: New York, 1940. (b) Hammett, L. P.
Physical Organic Chemistry, Reaction Rates, Equilibria, and Mechanisms,
2nd ed.; McGraw-Hill: New York, 1970.
(2) (a) Photochemistry in Organized and Constrained Media; Ramamurthy,
V., Ed.; VCH: New York, 1991. (b) Ramamurthy, V.; Eaton, D. F. Acc.
Chem. Res. 1988, 21, 300-306. (c) Weiss, R. G.; Ramamurthy, V.;
Hammond, G. S. Acc. Chem. Res. 1993, 26, 530-536. (d) Ramamurthy,
V.; Eaton, D. F.; Caspar, J. V. Acc. Chem. Res. 1992, 25, 299-307. (e)
Sivaguru, J.; Natarajan, A.; Kaanumalle, L. S.; Shailaja, J.; Uppili, S.; Joy,
A.; Ramamurthy, V. Acc. Chem. Res. 2003, 36, 509-521.
(3) (a) Turro, N. J. Acc. Chem. Res. 2000, 33, 637-646. (b) Turro, N. J. Chem.
Commun. 2002, 2279-2292. (c) Turro, N. J. Proc. Natl. Acad. Sci. U.S.A.
2005, 102, 10766-10770. (d) Turro, N. J. Proc. Natl. Acad. Sci. U.S.A.
2002, 99, 4805-4809. (e) Turro, N. J.; Gra¨tzel, M.; Braun. A. M. Angew.
Chem., Int. Ed. Engl. 1980, 19, 675-696.
(4) (a) Tung, C. H.; Wu, L. Z.; Zhang, P. P.; Chen, B. Acc. Chem. Res. 2003,
36, 39-47. (b) Gamlin, J. N.; Jones, R.; Leibovitch, M.; Patrick, B.;
Scheffer, J. R.; Trotter, J. Acc. Chem. Res. 1996, 29, 203-209. (c) Toda,
F. Acc. Chem. Res. 1995, 28, 480-486. (d) Zimmerman, H. E.; Nesterov,
E. E. Acc. Chem. Res. 2002, 35, 77-85. (e) Garcia-Garibay, M. A. Acc.
Chem. Res. 2003, 36, 491-498. (f) Whitten, D. G.; Russell, J. C.; Schmehl,
R. H. Tetrahedron 1982, 38, 2455-2487. (g) Whitten, D. G. Acc. Chem.
Res. 1993, 26, 502-509. (h) Weiss, R. G. Tetrahedron 1988, 44, 3413-
3475.
(5) Gibb, C. L. D.; Gibb, B. C. J. Am. Chem. Soc. 2004, 126, 11408-11409.
(6) (a) Gibb, C. L. D.; Gibb, B. C. Chem. Commun. 2007, 1635-1637. (b)
Gibb, C. L. D.; Gibb, B. C. J. Am. Chem. Soc. 2006, 128, 16498-16499.
(7) Natarajan, A.; Kaanumalle, L. S.; Jockusch, S.; Gibb, C. L. D.; Gibb, B.
C.; Turro, N. J.; Ramamurthy, V. J. Am. Chem. Soc. 2007, 129, 4132-
4133.
9
10.1021/ja7107917 CCC: $40.75 © 2008 American Chemical Society
J. AM. CHEM. SOC. 2008, 130, 4069-4080
4069

A R T I C L E S
Gibb et al.
section that explores the excited-state chemistry of these guests
2a-h, and (b) an NMR analysis revealing the packing of each
guest within the capsule.
Experimental Section
Materials. Host 1 was synthesized according to the literature.5 The
different R-(alkyl) DBK guests were synthesized by the general
procedure outlined below (Scheme 2). DBK, sodium hydride, and alkyl
bromides were obtained from Aldrich and used as such. THF was
freshly distilled over sodium.
Sodium hydride (60% dispersion in mineral oil, 0.6 equiv; washed
with dry pentane) was stirred with THF and cooled to 0 °C. DBK (1
equiv, dissolved in THF) was added with stirring. Alkyl bromide (0.6
equiv) was added over 10-15 min, and the reaction was continued at
0-10 °C for 1-1.5 h (40-50% alkylation). The reaction was quenched
by slow addition of 10% aqueous NH4Cl solution. THF was removed
under reduced pressure, and the product was extracted with three
portions of ether. The ether layer was washed with brine, dried over
Na2SO4, and concentrated to obtain the crude product as a viscous liquid
containing about 5-10% of dialkylated DBK. Purification by flash
chromatography (silica gel, hexanes/dichloromethane) afforded the
We have shown recently that products of oxidation of olefins,7
geometric isomerization of stilbenes,8 dimerization of acenaph-
thylene and styrenes,8,9 R-cleavage reaction of dibenzyl ketones
(DBK),10 â-cleavage reaction of naphthyl esters,11 and excimer
formation of anthracene12 could be controlled by enclosing
photoreactive molecules within the confined space of the
capsule. These studies have established the importance of free
space and weak interactions between the host interior and the
guest molecules in altering the excited-state behavior of the
latter. A detailed photochemical investigation of DBK10 and
1-(4-alkylphenyl)-3-phenyl propan-2-ones (para-alkyl DBK)
revealed13 that the extent of free space at the tapered ends of
the capsule plays an important role in the extent of rearrange-
ment product formed (Scheme 1, yield of product 9 (R ) H)
from dibenzyl ketone: 49% and corresponding product from
1-(4-pentylphenyl)-3-phenyl propan-2-ones: 0%).
1
a-alkyl DBK as a viscous oil. H NMR and 13C NMR data for 2a-h
are presented in the Supporting Information.
Protocols for NMR Binding Studies. NMR experiments were
performed on an Inova 500 MHz spectrometer (Varian Inc.). All binding
studies involved titration of the R-alkyl DBK 2 to the host solution to
confirm the stoichiometry of the complexes. The general protocol was
as follows: 0.6 mL of 1 mM host 1 and 10 mM sodium tetraborate in
D2O were titrated with aliquots of 2 µL of a 30 mM solution of 2 in
DMSO-d6. Full complexation occurred at 0.5 equiv of 2 (i.e., 10 µL).
Complexation was rapid on the experimental time scale, but the NMR
spectra were rerecorded after 24 h to ensure no further changes had
occurred.
Formation of 9 from DBK within a capsule is envisioned to
proceed as follows: The encapsulated DBK 2 (R ) H) is
orientated in such a manner that each aromatic ring occupies
one cavitand or hemisphere. Upon irradiation, one of the C-C
bonds R to the carbonyl group homolytically cleaves, but the
resulting radicals do not optimally pack the cavity and thus the
benzyl radical rotates 180° relative to the acyl radical. Coupling
of the radicals at the para position of the benzyl ring followed
by 1,5 H-shift results in 9.
With these results in hand, and with the idea of expanding
on the repertoire of this capsule as a water-based nanoreactor,
we sought to investigate the photochemistry of encapsulated
R-(n-alkyl) DBK derivatives 2a-h, which undergo both Norrish
type II and/or Norrish type I reactions. Our interest was to
examine how guest packing within the capsule (as dictated by
the variable R-n-alkyl chain) influenced the partitioning of R-(n-
alkyl) DBK derivatives 2 by the two reaction pathways. The
presence of an R-(n-alkyl) group desymmetrizes the guest, and
in relation to this study, the two benzyl carbon-carbonyl carbon
bonds and the two phenyl rings. In this work, we differentiate
between the latter by referring to them as being proximal or
distal to the alkyl group (red and blue, respectively, Scheme
1). The study consisted of two parts: (a) a photochemistry
To perform COSY, TOCSY (mixing time 80 ms), and NOESY
(mixing time 0.5 s) experiments, 5 mM solutions of the complexes
were prepared: 5 mM host 1, 2.5 mM R-alkyl DBK 2 (R ) various),
and 50 mM sodium tetraborate in D2O. Variable-temperature 1H NMR
experiments ranged from 5 to 60 °C.
Protocols for NMR Diffusion Measurements. Diffusion measure-
ments were performed on an Inova 500 MHz (Varian Inc.) instrument
equipped with a Performa II pulsed field gradient (PFG) module capable
of producing pulses up to 52 G/cm. The experiments were carried out
on a 5-mm PFG indirect detection probe. The stimulated echo (STE)
diffusion experiment using the Varian pulse sequence “pge” (stimulated
option on) was performed with pulse gradients of 2 ms in duration
separated by 155 ms. Calibration utilized D2O samples with a diffusion
constant of 1.88 10-5 cm2 s-1. The experiments were performed at
25 °C, at a host concentration of 1 mM (in 10 mM sodium tetraborate).
The given diffusion constants were an average of three measurements.
Data were analyzed using the optional Varian diffusion software and
gave, for example, a value of D ) 1.82 10-6 cm2 s-1 for the free host
1 and a value of D ) 1.30 10-6 cm2 s-1 for the 2 a@1 complex.
Assuming the entity is spherical, a hydrodynamic radius (Rh) and
corresponding volume can be calculated from D according to the
following equation:
kT
Rh )
6πηD
(8) Parthasarathy, A.; Kaanumalle, L. S.; Ramamurthy, V. Org. Lett. 2007, 9,
5059-5062.
where Rh is the hydrodynamic radius of the sphere in meters, k is the
Boltzmann constant, T is the temperature in Kelvin, η is the solvent
(9) Kaanumalle, L. S.; Ramamurthy, V. Chem. Commun. 2007, 1062-1064.
(10) Kaanumalle, L. S.; Gibb, C. L. D.; Gibb, B. C.; Ramamurthy, V. J. Am.
Chem. Soc. 2004, 126, 14366-14367.
viscosity, and D is the diffusion constant in m2 s-1
.
Protocols for Photochemistry Experiments. Irradiations were
carried out using a medium-pressure Hg lamp. All samples were
bubbled with N2 for 30 min before irradiations, and bubbling continued
during photolysis. Solutions of the ketones 2a-h (10-3 M) in hexanes
(11) Kaanumalle, L. S.; Gibb, C. L. D.; Gibb, B. C.; Ramamurthy, V. Org.
Biomol. Chem. 2007, 5, 236-238.
(12) Kaanumalle, L. S.; Gibb, C. L. D.; Gibb, B. C.; Ramamurthy, V. J. Am.
Chem. Soc. 2005, 127, 3674-3675.
(13) Sundaresan, A. K.; Ramamurthy, V. Org. Lett. 2007, 9, 3575-3578.
9
4070 J. AM. CHEM. SOC. VOL. 130, NO. 12, 2008

Excited-State Chemistry of
R
-(n-Alkyl) DBK
A R T I C L E S
Scheme 1. Overall Reaction Manifold for the Photochemistry of Encapsulated DBK and R-Alkyl DBKsa
a The three routes away from starting material are: (1) Norrish type I photochemistry that leads to (a) a range of decarbonylated products 3-8 and (b)
rearranged products 9 and 10, and (2) Norrish type II photochemistry to yield products 11-13. In structures 11 and 13, R′ represents alkyl chain that is less
by CH2CH2 from R.
Scheme 2. Synthesis of the Different R-(Alkyl) DBK Guests
or aqueous sodium tetraborate (10-2 M) taken in a Pyrex test tube were
irradiated for 30 min. Reaction conversion of 30-50% was achieved
during this period. The products (extracted from buffer solution using
CHCl3) were analyzed by GC (HP 5890 series, SE-30 column) and
GC-MS (HP-6890N with 5975B MSD; HP-5 column).
A stock solution of cavitand 1 (10-3 M) was prepared in sodium
tetraborate (10-2 M) buffered H2O or D2O. Stock solutions of the guests
were prepared in DMSO-d6. To 0.6 mL of a solution of 1 taken in an
NMR tube, an aliquot of guest solution was added, maintaining a host-
guest ratio of 4:1, and the solution was sonicated for 30 min. 1H NMR
showed formation of complex. The complex was irradiated for 30 min,
maintaining nitrogen purge during irradiation. The products were
extracted using chloroform and analyzed by GC and GC-MS.
Photoproducts were identified by comparison with those obtained from
irradiation of homogeneous solutions. Structures of rearrangement
products (8 and 9) were established by comparison with authentic
samples prepared independently.
concentrations of free guest were apparent when the titration
went beyond stoichiometric, demonstrating that guest exchange
between the free and encapsulated environments was slow on
the (500 MHz) NMR timescale. Confirmation of the capsular
nature of these complexes was also obtained using pulse gradient
spin-echo NMR experiments,14 which yielded the diffusion
rates and corresponding particle sizes (Table 1).
Photochemical Results. In general, R-(n-alkyl) DBKs un-
dergo two types of photoreactions: Norrish type I and type II
(Scheme 1). The second reaction occurs only with guests that
possess γ-hydrogen atoms (2c-h). Our primary interest was to
explore the photochemical behavior of these molecules when
included in the capsule formed by 1. For comparative purposes,
their excited-state behaviors in hexane and borate buffer were
also investigated. Results of photolysis are summarized in Table
2. As expected, irradiation of ketones 2a and 2b in hexane and
Results
Stoichiometry of the Complexes. Eight guests, R-methyl-
through R-(n-octyl) DBK (Scheme 1, 2a-h), were examined.
1H NMR titration studies involving 1 and 2 revealed that all
formed capsular 2:1 host-guest complexes. The significant
upfield shift (vide infra) of the guest signals revealed that all
were held within the essentially dry interior of the capsule. Low
(14) Cohen, Y.; Avram, L.; Frish, L. Angew. Chem., Int. Ed. 2005, 44, 520-
554.
9
J. AM. CHEM. SOC. VOL. 130, NO. 12, 2008 4071

A R T I C L E S
Gibb et al.
Table 1. Diffusion Rates, Complex Radius, and Volume for the
Complexes Formed between 2a-h and 1 as Derived from Pulse
Gradient Spin-Echo NMR Spectrometry
encapsulated R-octyl DBK that gave 90% of Norrish type II
products, 11 and 12, while in hexane solution the predominant
products (∼85%) were from Norrish type I reaction. (c) The
second group consists of R-butyl and R-pentyl DBKs (2d and
2e) that yield the two rearranged starting ketones 9 and 10 in
major amounts (>60%) almost in equal ratio. Products of
Norrish type II reaction were obtained in less than 10% yield.
(d) The third group consists of R-methyl, R-ethyl, and R-propyl
DBKs (2a-c) that within capsule yield only one (not two)
rearranged starting ketone 10 in larger amounts. The second
one (9) is formed in less than 5%. In this set, as expected, only
R-propyl DBK gave Norrish type II products.
1H NMR Structural Characterization of Host-Guest
Complexes. This section is subdivided into three parts pertaining
to signals from the alkyl chain of the guest, the guest aromatic
region, and the host aromatic region. Characterization relied on
signal shift data in the 1D spectrum (the general rule is that
atoms residing at the tapered top and bottom regions of the
capsule are shifted more from the free state than those residing
at the equatorial region), as well as COSY, TOCSY (short- and
long-range coupling within the guest), and NOESY (through-
space host-host, guest-guest, and host-guest interactions)
experiments. With the exception of Figures 2-4, all these data
are presented in the Supporting Information.
guest
(2, R
diffusion rate
hydrodynamic
radiusa (nm)
hydrodynamic
volumeb (nm3)
1
)
)
(
×
10-10 m2 s-
)
CH3
C2H5
1.30
1.33
1.28
1.32
1.32
1.46
1.27
1.31
1.7
1.6
1.7
1.7
1.7
1.5
1.7
1.7
19.8
18.5
20.8
18.9
18.9
14.0
21.3
19.4
n-C3H7
n-C4H9
n-C5H11
n-C6H13
n-C7H15
n-C8H17
a Derived from Rh (hydrodynamic radius) ) kT/(6πηD). b Derived from
V ) 4/3πr3 (R ) Rh).
Table 2. Product Distribution upon Photolysis of R-(n-Alkyl)
Dibenzyl Ketones in Hexane and Borate Buffer and as Complexes
with the Capsule in Borate Buffera,b
guest 2
type 1
type 2
rearranged
DBK
medium
AA (7) ABc (3) BB (8)
(9)
(10)
DBK (12) cyclobutanol (11)
2a hexane
buffer
capsule
2b hexane
buffer
capsule
2c hexane
buffer
capsule
2d hexane
buffer
capsule
2e hexane
buffer
capsule
2f hexane
buffer
capsule
2g hexane
buffer
capsule
2h hexane
buffer
24
21
48
53
75
48
50
51
40
43
50
45
52
29
41
46
28
46
51
73
49
49
60
46
42
10
28
26
0
0
4
0
0
5
0
0
4
0
0
0
0
21
0
0
29
0
0
35
0
23
23
7
16
16
29
27
8
24
21
Alkyl Chain NMR Data. Figure 1 reveals that the R-alkyl
11
11
4
9
9
7
chain signals (numbered such that the terminal methyl groups
1
are labeled H1) appear in the upfield region of the H NMR.
The shift in the methyl signal of 2a upon complexation (∆δ )
-0.98 ppm) is relatively small. However, the corresponding
shifts increased with lengthening alkyl chain: 2b, -1.82 ppm;
2c, -2.79 ppm; 2d, -3.72 ppm; 2e, -3.79 ppm; 2f, -3.66 ppm;
2g, -3.97 ppm; and 2h, -4.07 ppm. Within the confines of
the capsule, most of the methylene hydrogens were observably
diastereotopic, with two signals for almost all methylene groups.
However, some signal overlaps were apparent in longer homo-
logues (2f-h). In almost all cases, the methylene hydrogens â
to the carbonyl chromophore demonstrated the largest anisot-
ropy. For guests 2c-h, NOESY NMR revealed through-space
interactions between the host Hbenzal atoms (see 1) and the
terminal methyl. For the longer homologues, 2e-h, an interac-
tion between the Hbenzal atoms and the penultimate methylene
was also observed. This latter point suggests that packing is
such that the methyl group is anchored deeply in the tapering
cavity of the cavitand.
17
17
0
16
17
0
15
14
0
13
15
0
20
18
0
23
26
0
32
27
0
26
26
0
7
7
11
6
0
28 36
0
0
2
5
0
0
4
4
16
7
34 32
3
3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
4
3
6
2
25
4
2
8
4
6
32
5
20
83
8
10
5
16
13
0
23
20
0
capsule
7
a See Scheme 1 for structures of products. b The GC results are average
of at least six independent runs, and reported yields are based on the
integrated area of the GC signals. c Includes AB 3 and its isomers 4, 5, and
6.
borate buffer resulted in 1,3-diphenyl ethanes 7 (AA), 3 (AB),
and 8 (BB) in 1:2:1 ratios. Furthermore, excitation of 2c-h
possessing γ-hydrogen atoms gave products of both Norrish type
I and type II reactions. On the other hand, irradiation of the 2:1
complexes formed between host 1 and guests 2a-h gave a
distribution of products strikingly different from that in solution.
Trends observed were: (a) Irradiation of seven R-alkyl DBKs
(2a, 2c-h) did not give the cage escape products 7 and 8 (AA
and BB). The only decarbonylated products obtained from these
were AB and its rearrangement isomers (3-6). R-Ethyl DBK
(2b) gave small amounts of AA and BB along with major
amounts of AB. (b) On the basis of their photobehavior, we
can divide the eight guests into three groups. The first consists
of R-hexyl, R-heptyl, and R-octyl DBKs (2f-h) that do not give
any rearranged starting ketones either in isotropic solution or
within the capsule. These ketones yield large amounts of Norrish
type II products within the host. This was especially true with
Aromatic Signals of the Guest. For convenience, we label
the two aromatic rings of the guest as proximal and distal to
the R group (red and blue, respectively, in Scheme 1). In the
case of 2a, all six signals for the two rings were visible as pairs
at δ ≈ 5.50 ppm (2H each), 4.70 ppm (2H each), and 3.50 ppm
(1H each). Confirmation of their identity required both COSY
and TOCSY NMR. The COSY spectrum identified several key
interactions within the guest, including the methyl-methine
coupling, both ortho-para couplings, and one of the meta-
para couplings. In regards to the latter, both meta-para
couplings were apparent with TOCSY NMR. Taken together,
these data identified the pairs of signals at δ ≈ 5.50, 4.70, and
3.50 ppm as belonging to the ortho-, meta-, and para-hydrogens,
respectively, of the two phenyl rings. The NOESY NMR
spectrum of the complex revealed many interactions, including
those between the para-hydrogens of the guest and the Hbenzal
9
4072 J. AM. CHEM. SOC. VOL. 130, NO. 12, 2008

Excited-State Chemistry of
R
-(n-Alkyl) DBK
A R T I C L E S
1
Figure 1. Guest region of the H NMR spectra of the complexes formed between cavitand 1 and R-alkyl DBK guests 2a-h. Guests are: (a) 2a, (b) 2b,
(c) 2c, (d) 2d, (e) 2e, (f) 2f, (g) 2g, and (h) 2h. Numbering is such that the methyl group is position one. Numbering for 2h is shown as an example. *
Denotes impurity.
atoms of the host. A similar situation was observed with the
complex of 2b. The R-propyl DBK (2c) complex behaved
slightly differently. In this complex, one set of aromatic signals
was considerably more shifted that the other, confirming that
one ring was not located deep in a hemisphere but rather in the
equatorial region of the capsule. The presence of an NOE
interaction between the ortho-hydrogens of this ring and the
methine hydrogen identified it as the proximal (to the R group)
phenyl. Only five of the six aromatic ring signals for encapsu-
lated 2d were observed. The ring with only two signals
accounted for was the least shielded, and as was the case for
the complex with 2c, this was assigned as the proximal ring. In
contrast to the smaller homologues, signals from only one
aromatic ring were observable in the complex with 2e. The
observable ring was shielded to a slightly greater extent than
the most shielded ring in the 2d complex and showed no NOE
between its ortho-hydrogen atoms and the methine C-H. Again,
therefore, this deeper phenyl was assigned as the distal ring.
For encapsulated 2f, only two aromatic signals from the guest
were apparent. COSY and TOCSY NMR revealed no coupling
between these, suggesting that they are not in the same ring
(or, less likely, that the dynamic nature of the encapsulated guest
masked coupling). In the 2g complex, four aromatic signals of
the guest were visible, but COSY NMR did not yield coupling
information. However, TOCSY NMR did identify that the
signals at 2.80 and 5.10 ppm belong to the same ring.
Additionally, integration confirmed that the former corresponded
to a para-hydrogen, while an NOE interaction between the latter
and the methine hydrogen of the guest confirmed that this signal
originated form the ortho-hydrogens of the proximal phenyl.
Finally, in the 2h complex the two phenyls of the guest were
again observed: one bound deeply, the other shielded only to
a small amount. The ortho-hydrogens of the deeply bound ring
showed an NOE interaction with the methine H of the guest,
again demonstrating that it was the proximal ring.
Aromatic Signals of the Host. As the guests 2a-h are chiral,
the enantiotopic host atoms Ha and Hb (1) are diastereotopic
(nonequivalent) in the complex. However, with the complexes
of 2a and 2b, only the signals for the Ha atoms were observed
to be magnetically nonequivalent; a singlet was observed for
the Hb atoms (Figure 2). It is not clear why in this case the Ha
atoms on the exterior of the host are better reporters for the
encapsulation of chiral guests than the Hb atoms that make up
part of the wall of the cavity. In the complex with 2c, the
anisotropy of the diastereotopic Ha atoms was diminished
(Figure 2c) and the formally diastereotopic Hb atoms continued
to be indistinguishable. In contrast, the NMR spectra of the
complexes with 2d and 2e (Figures 2d and 3a) revealed the
diastereotopicity of not only the Ha atoms, but also the Hb atoms.
However, in the 2f complex the nonequivalence of the Ha atoms
9
J. AM. CHEM. SOC. VOL. 130, NO. 12, 2008 4073

A R T I C L E S
Gibb et al.
Figure 2. Aromatic region of 1H NMR spectra of the complexes formed between cavitand 1 and guests 2a-d. R guests are: (a) 2a, (b) 2b, (c) 2c, and (d)
2d. * Denotes residual water signal.
1
Figure 3. Aromatic region of H NMR spectra of the complexes formed between cavitand 1 and guests 2e-h. Guests are: (a) 2e, (b) 2f, (c) 2g, and (d)
2h. * Denotes residual water signal.
is diminished, and only a broad signal for the Hb atoms is
apparent. The NMR signals due to the host in the complex with
2g are interesting; the host signals, which undergo relatively
small shifts upon complex formation, were broadened in this
complex (Figure 3c), presumably because of the reduced
mobility of this larger guest. Nevertheless, it was still possible
to identify two Ha signals, and for the first time two Hbenzal
signals. Increased temperature leads to increased mobility, and
the NMR spectrum of this complex at 60 °C was more akin to
those of the smaller homologues (Figure 4a). Correspondingly,
9
4074 J. AM. CHEM. SOC. VOL. 130, NO. 12, 2008

Excited-State Chemistry of
R
-(n-Alkyl) DBK
A R T I C L E S
(long-range coupling), and NOESY (through-space host-host,
guest-guest, and host-guest interactions) data.
As we shall see, NMR revealed three guest-packing motifs
modulated by the length of the R-alkyl chain of the guest. To
define these different packing motifs, we define two distinct
regions within the capsule and three moieties of the guest. Thus,
the capsule has two identical polar regions and an equatorial
region, while the guests are viewed in terms of the variable R
group and the phenyl rings proximal (red in Figures 1 and 5)
and distal (blue) to the R group. We observed three packing
motifs: the two aromatic rings occupy the polar region of the
capsule and the R group the equatorial region (e.g., Figure 5a);
the distal phenyl and R group occupy the polar regions (e.g.,
Figure 5d); and the proximal phenyl and the R group occupy
the polar regions (e.g., Figure 5h).
(1) Guests that Pack the Two Hemispheres with Two
Aromatic Rings. Packing of guests 2a and 2b is dominated by
their aromatic rings. Multiple lines of evidence from NMR
confirm that for 2a and 2b each phenyl ring occupies a
hemisphere (Figure 5a,b). Thus, the methyl groups of 2a and
2b are shifted -0.98 and -1.82 ppm, respectively, from their
free states (Figure 1). These are relatively small shifts compared
to the ca. 4.00 ppm shifts of the larger homologues, suggesting
that their R groups are located in the spacious equatorial region
of the capsule. However, no NOE interactions between these
methyl groups and the suitably located Hendo atoms of the host
were observed. The shift in the methyl signal of 2c is larger
(-2.97 ppm) and in contrast shows an NOE interaction with
the inward point Hbenzal atoms of the host. In combination, these
data suggest that the side chain of the propyl guest is beginning
to compete with one of the aromatic rings for occupation of a
hemisphere (Figure 5c). Remaining with the alkyl groups of
these guests, it is worth noting that the CH2 groups are
observably diastereotopic. The anisotropy within a CH2 group
is small and consistent, while the anisotropy between the CH2
and the terminal methyl is large.
1
Figure 4. Variable-temperature H NMR spectra of the complex formed
between cavitand 1 and guest 2g. Temperatures are: (a) 60, (b) 25, and (c)
5 °C.
at a relatively low 5 °C tumbling of the guest whereby the
contents of each hemisphere exchange places was slow on the
NMR time scale, with the result that the host signals underwent
splitting (Figure 4c). The NMR host signals of the complex
with 2h are also generally broad at room temperature (Figure
3d). However, in contrast to the 2g complex, the Hbenzal signal
is not split. Variable-temperature studies gave results analogous
to the 2g complex.
Discussion
In aqueous solution, host 1 dimerizes around a broad range
of molecules.5-12 In the study here, a series of eight R-(n-alkyl)
DBK derivatives 2a-h yielded 2:1 capsular complexes (2a-
h@12). The first part of this section is devoted to determining
the structures of these complexes using NMR. The second part
discusses the photochemistry of these different complexes. It
is satisfying to note that the NMR-derived structures of the
guests within the capsule remarkably predict the observed
photochemical behavior of the guest ketones.
Structural Characterization of the Host-Guest Complexes
through 1H NMR. Characterization by NMR relied on 1- and
2D experiments. In the former, a comparison between the spectra
of the free host and guest and the corresponding complex is
useful because the binding pocket of each cavitand is akin to a
truncated cone. Hence, atoms that reside deeper in the cavity
are more shifted (shielded) by the aromatic wall of the capsule
than those that are located at the equator. This phenomenon
is par for the course for resorcinarene-based hosts and cap-
sules.5-12,15-21 Additionally, for full characterization it was
necessary to collect COSY (short-range coupling), TOCSY
The proposed structures for the three complexes are also
supported by the aromatic signals of each guest. For 2a and
2b, both phenyls undergo significant shifts relative to the free
state. All six signals are apparent in both complexes, and a
combination of COSY and TOCSY experiments confirmed that
the ortho-, meta-, and para-hydrogens of both rings occur in
pairs at ca. 5.5, 4.7, and 3.5 ppm. It is the para-hydrogens that
are the most deeply bound, a conclusion further supported by
NOE interactions between the para-hydrogens of each guest
and the Hbenzal atoms of the host. In contrast, in the complex
with 2c only one aromatic ring is markedly shielded. The other
ring is much less so, suggesting that its average position is closer
to the equatorial region. Indeed, CPK models confirm that there
is little room for both the propyl side chain and an aromatic
ring to occupy a hemisphere. Which aromatic ring is displaced
toward the equator? The ortho-hydrogens of the aromatic ring
experiencing less shielding show strong NOEs with the methine
hydrogen of the guest and the adjacent methylene group on the
side chain, confirming that it is the proximal aromatic rings
that are being displaced from the hemisphere by the propyl
group (Figure 5c). The propyl group is less voluminous than
the phenyl ring (69 versus 98 Å3), but it presumably packs the
narrowest part of the cavity more efficiently. These points stated,
comparing the NMR data and photochemical products from this
(15) Biros, S. M.; Rebek, J., Jr. Chem. Soc. ReV. 2007, 36, 93-104.
(16) Rebek, J., Jr. Angew. Chem., Int. Ed. 2005, 44, 2068-2078.
(17) Purse, B. W.; Rebek, J. J. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 10777-
10782.
(18) Rebek, J., Jr. J. Org. Chem. 2004, 69, 2651-2660.
(19) Hof, F.; Rebek, J., Jr. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 4775-4777.
(20) Rebek, J., Jr. Chem. Commun. 2000, 637-643.
(21) Rebek, J., Jr. Acc. Chem. Res. 1999, 32, 278-286.
9
J. AM. CHEM. SOC. VOL. 130, NO. 12, 2008 4075

A R T I C L E S
Gibb et al.
Figure 5. Representations of the complexes formed by: (a) 2a, (b) 2b, (c) 2c, (d) 2d, (e) 2e, (f) 2f, (g) 2h, and (h) 2h.
complex to those formed by 2d and 2e (vide infra) suggests
that the equilibrium between the two states depicted in Figure
5c is in favor of the phenyl ring occupying the northern
hemisphere.
with 2d and 2e (Figure 1). Apparently, one of these diaste-
reotopic pairs is considerably more shielded than the other.
Presumably, it lies close to, and close to the plane of, the
proximal phenyl ring.
Guest packing changes are also evident in the NMR signals
of the host. Such changes can be illustrated using the atoms
(see structure 1) Ha, Hb, Hendo, Hexo, and Hbenzal (Figure 2). In
regards to changes in absorption, the inward pointing Hbenzal
atoms are good reporters of guest binding. In the case at hand,
there is an upfield shift in this signal as the alkyl group changes
from methyl to n-propyl, a trend that continues through to the
complex 2d. Host signal splitting is also informative. Formally,
atoms Ha and Hb are enantiotopic in the free host and
diastereotopic in the complexes. Interestingly, however, the
atoms more remote from the chiral guest (Ha) are better reporters
for encapsulation than the Hb atoms that make up part of the
wall of the cavity. Although slightly more removed from the
binding site, changes to their surrounding environment upon
guest binding are apparently more significant.
For 2d, only five of the six aromatic signals are apparent in
the complex. The ring with all three observable signals was the
most shielded and, following on from the complex with 2c, was
assigned as the distal phenyl. Another difference between the
aromatic signals of 2d and encapsulated 2c was the degree of
shielding of this distal ring. It was not possible to fully assign
all the signals of free guest in aqueous solution and therefore
not possible to accurately determine the corresponding ∆δ
values for both complexes. Nevertheless, the -0.60 ppm shift
for the para-hydrogen in encapsulated 2d versus the 2c complex
is much larger than any difference between the free guests.
Hence, the increased size of the alkyl group forces the distal
phenyl deeper into the opposing hemisphere.
The complex with 2e proved to be less informative; no signals
for the less shielded, equatorial, proximal ring were apparent.
Furthermore, variable-temperature studies (5-55 °C) failed to
yield any information about this “invisible” ring. As it “reap-
pears” in the larger complexes (vide infra), a dynamic confor-
mational process on or near the NMR time scale is assumed to
be behind this phenomenon. Figure 5e denotes, by way of
example, a slowed rotation around the C-Ph bond.
Changes in packing are also indirectly observable in the NMR
spectrum of the host. For example, following a decreasing trend
in anisotropy of the Ha atoms in the complexes with 2a to 2c,
there is a sudden increase (Figure 2) in their diastereotopicity
in the complex with 2d. Furthermore, it is possible for the first
time to observe magnetic nonequivalence for the Hb atoms,
atoms that, although formally diastereotopic, appear equivalent
in the complexes of 2a-c. Interestingly, the complex with 2e
again reveals the diastereotopicity of these atoms (Figure 3),
but the anisotropy in both is diminished relative to that of 2d.
One interesting feature of the host region of the complex with
2e is broadening of the Hendo signal (Figure 3). These eight atoms
in the complex point inward at the equatorial region of the
capsule and are good reporters of events at the region. It will
become apparent (vide infra) that gross flipping of the guest
within the capsule, whereby the contents of one hemisphere trade
places with the contents of the other, slows down to the NMR
In summary, guests 2a-c pack the capsule such that the two
aromatic rings occupy the two hemispheres. However, with
guest 2c the propyl group competes with the proximal phenyl
ring for packing a hemisphere.
(2) Guests that Pack the Hemispheres with an Alkyl
Group and the Distal Ring. Whereas the alkyl group in the
2c complex struggles for occupancy of the northern hemisphere,
NMR evidence demonstrates that those of the 2d and 2e
complexes are large enough (87 and 105 Å3, respectively) to
fully displace the proximal phenyl ring (98 Å3). Thus, the methyl
shift in these two complexes is near maximal (-3.72 and -3.79
ppm, respectively). These greater shifts can be attributed to two
phenomena: a longer residency time in the hemisphere and
occupation of deeper/narrower regions of the cavity. The latter
is particularly evident in the NOESY NMR spectrum of the
complex with 2e, which shows NOEs between the Hbenzal atoms
of the host and both the terminal methyl and the penultimate
methylene group. The chain is now long enough that the methyl
group can move beyond the crown of Hbenzal atoms.
For both guests, the alkyl group methylene signals are
distributed over a range of ca. 4 ppm. There is therefore no
signal coincidence and it is possible to observe the large
anisotropy of the respective H4 and H5 atoms in the complex
9
4076 J. AM. CHEM. SOC. VOL. 130, NO. 12, 2008

Excited-State Chemistry of
R
-(n-Alkyl) DBK
A R T I C L E S
time scale as the guest increases in size. However, such
broadening occurs for all host signals. For the 2e complex, only
the Hendo signal broadens, and with the bigger guests this peak
is sharper. In other words, this Hendo broadening is out of step
both with other signals from this complex and with Hendo signals
in different complexes. We attribute this manifestation to the
same (on the NMR time scale) movement responsible for the
“missing” aromatic ring. Again, variable-temperature studies do
not reveal more information about this process, but we view it
as quite distinct from the global movements of the guest such
as gross flipping.
between them. This point notwithstanding, photolysis of this
complex and NMR data from the complexes of 2g and 2h
(vide infra) suggest that the 2f complex represents the transition
between different packing motifs, a packing motif in which
the aromatic rings exchange positions with each other (Figure
5f). Thus, now that the alkyl group is sufficiently long, it is
energetically feasible to locate the proximal ring in the
opposite hemisphere and the distal ring at the equator. The net
effect of this change is to make the guest a bit shorter (by
-CH2C(O)-), but slightly more rotund (by the same amount).
This dynamic nature likely destabilizes the capsule leading to
fast diffusion data (Table 1).
In summary, guests 2d and 2e pack the capsule such that the
hemispheres are filled with the alkyl group and the distal ring.
In this packing arrangement, the proximal ring can adopt one
of at least two isoenergetic positions.
For encapsulated 2g, four of the six aromatic signals are
apparent. A COSY experiment did not yield coupling informa-
tion, but a TOCSY experiment did identify the signals at 5.1
and 2.8 ppm as originating from the same ring. From the studies
of the smaller guests, these signals can be attributed to the ortho
and para atoms of a hemispherically bound (rather than
equatorially bound) aromatic ring. Importantly, the signal at 5.1
ppm showed a strong NOE with the methine hydrogen of the
guest, indicating that the most deeply bound phenyl is the
proximal ring. This type of interaction was observed with the
complex of 2c, but in that instance the proximal ring was located
at the equatorial region; the proximal and distal rings have traded
places.
(3) Guests that Pack the Hemispheres with an Alkyl
Group and the Proximal Ring. NMR data reveal that guests
2f-h pack the cavity of the capsule so that the alkyl group and
the proximal ring fill the two hemispheres, while the distal ring
resides in the equatorial region of the host. NMR also reveals
that, like the guest 2c, guest 2f lies on the border between two
modes of packing.
The methyl group of the encapsulated 2f is slightly less shifted
(-3.66 ppm) than those of guests 2d and 2e. Indeed, this shift
is slightly less (Figure 1) than larger guests, suggesting that 2f
is unusual. As with the other guests, all the signals of the n-alkyl
chains of encapsulated 2f-h can be accounted for. However,
packing is tight and there is signal overlap for the H2-H4 atoms.
This is also apparent with CPK models, which suggest that these
guests can fit only if their alkyl chains adopt highly compact
conformations. The fact that there is overlap of the alkyl signals,
and that accompanying COSY data are incomplete, means that
the H2-H4 assignments are based on the trends observed for
the smaller guests. Nearer the capsule equator, the readily iden-
tifiable H6 and H7 hydrogens of the n-hexyl and n-heptyl chains
again demonstrate an anisotropy much larger than those other
groups in their chain. However, guest 2h bucks this trend.
This greater amount of data relative to 2f suggests that the
local, dynamic guest movement of 2g is slower, and/or that the
equilibrium suggested for 2f (Figure 5f) is pushed more toward
the isomer in which the proximal phenyl is embedded in the
“southern” hemisphere (Figure 5g). The NMR data for guest
2h continue this trend. Thus, all six aromatic signals of the guest
are observed, and it is apparent that one ring is bound deeply
in the southern hemisphere while the other is shielded only to
a small amount. As with the 2g complex, an NOE is observed
between the methine H and the ortho-hydrogen atom of the
deeply embedded proximal ring.
In addition to the changes to the local dynamics identified
through the aromatic signals of the guests 2f-h are changes at
the next level of scale. These are best illustrated by the aromatic
signals of the host. For all the complexes, the host exerts more
of an influence on the NMR spectrum of the guest than vice
versa. In other words, the host signals undergo relatively small
shifts upon complex formation. As a result, slowing of guest
tumbling is observed with the broadening of signals from atoms
such as Ha, Hb Hendo, and Hexo, while the guest signals remain
well defined (kcoal ) πx2∆ν). The broadening of the signal
from Hb as the guest increases in size is typical (Figure 3a-d).
The effects of temperature upon guest flipping are as expected
and are illustrated by the complex with 2g (Figure 4). At higher
temperature, the signals sharpen up somewhat to give a spectrum
reminiscent of the smaller guests, whereas at lower temperatures
there is considerable splitting of the signals to give a spectrum
similar to that observed when rigid guests such as steroids are
encapsulated.3 This “freezing” of guest mobility means that with
the exception of diastereotopic atoms Ha and Hb, all the host
signals split into sharp doubles; the northern and southern
hemispheres are magnetically nonequivalent, and each set of
atoms in one hemisphere is heterotopic with respect to its
counterpart in the other. This north-south divide also applies
to the Ha and Hb atoms. In addition however, because the guests
Although it is not possible to fully assign the H2-H4 atoms
of these three guests, the NOE spectrum does provide some
intriguing information concerning the general packing of the
side chains; for each guest, NOEs are observed in two distinct
packets. For example, in the case of the encapsulated 2g, there
are NOEs between atoms H5-H7 (0.5 to -0.5 ppm range), a
second packet of NOEs between atoms H1-H4 (-1.0 to -2.0
ppm), but no interactions between these two groups. It is difficult
to envision a conformation in which there is no possibility of
NOEs between, for example, H4 and H5 and H6. We therefore
attribute this observation to dynamics; conformational changes
within each of the two sections of chain allow observation of
NOE signals, but movement around the “hinge” section between
them is such that intersection NOEs are weak. Returning to the
bunching of the signals from H2-H4, it would appear that these
atoms are situated at a similar depth in the cavity. Thus, a
possible packing motif for this chain would be an anchored
methyl, three CH2 groups filling the next level of the cavity,
and a “kink” to accommodate the rest of the chain lying at the
equator.
As was the case with the 2e complex, the 2f complex provides
little information from its aromatic rings. Only two signals were
observed, and neither COSY nor TOCSY revealed coupling
9
J. AM. CHEM. SOC. VOL. 130, NO. 12, 2008 4077

A R T I C L E S
Gibb et al.
are chiral, each set of eight Ha/Hb atoms in a hemisphere is
split into two magnetically dissimilar (diastereotopic) sets of
four atoms. The resulting splitting of splitting is illustrated with
Ha; four signals are apparent at 5 °C.
respect to 7 and 8 (Scheme 1).25 Uniquely, in some of these
confined media rearrangements, product 9 was obtained in less
than 10% yield.26 Thus, the reaction medium that so greatly
influences the coupling of the RP-2 and leads to a large cage
effect with essentially only AB products generally has little
effect on the reactions of the RP-1 (only a small % of rearranged
starting ketone is normally formed). We have recently reported
that irradiation of para-methyl dibenzyl ketone included within
the capsule formed by 1 gave AB (and its rearranged isomers)
as the exclusive product of decarbonylation.10 Most importantly,
the rearrangement product 9 from the primary radical pair
(Scheme 1) was obtained in significant yield (>50%). Such a
remarkable change in product distribution that has not been
noted earlier in any other media suggests that this capsule is
unique in its ability to influence the behavior of both primary
and secondary radical pairs.
Whereas the aromatic signals of the host follow a systematic
broadening as the capacity of the capsule is approached, the
Hbenzal signal remains relatively sharp except in the case of guest
2g (Figure 3). This unique case of broadening cannot be
attributed to the small shift that it undergoes relative to the free
host; the shift of the (broad) Hbenzal of the complex with 2g is
more than the shift of the corresponding (sharp) signal in the
2h complex. Rather, we suggest that the Hbenzal of the former is
broad because tumbling is slowing down to the NMR time scale,
but that the tumbling of the latter is faster because packing is
beyond optimal and the two hemispheres of the capsule cannot
close properly. Tumbling is still, however, slow enough to cause
broadening of the aromatic signals.
To further explore the effectiveness of this capsule as a
reaction medium, we have currently examined the photochem-
istry of R-(n-alkyl) dibenzyl ketones (2a-h) that have multiple
chemical channels in the triplet excited-state surface. Ketones
2a-h can cleave on either side of the carbonyl chromophore
leading to two rearrangement products (9 and 10) and can
abstract γ-hydrogen leading to Norrish-Yang products (11-
13). The major products from all eight guests upon irradiation,
either in hexane or in buffer solution, resulted from the well-
known Norrish type I reaction (Table 2). Independent of the
solvent and the alkyl chain, the only products obtained from
decarbonylation via Norrish type I process were 3, 7, and 8.
No rearrangement products of 3, namely 4, 5, and 6, and that
of the starting ketone, namely 9 and 10, were formed in any
solvent (Scheme 1 and Table 2). The guests that possess
γ-hydrogen atoms in the R-alkyl chain gave, via a Norrish type
II pathway, products 11, 12, and 13 in less than 20% yields.
Clearly, in hexane and buffer solutions the Norrish type II
process is a minor pathway compared to the type I process.
Independent of the fact that both R-cleavage and γ-hydrogen
processes were expected to proceed with similar rates, the lower
yield of products from the Norrish type II versus Norrish type
I could be attributed to the conformational flexibility of the alkyl
chain resulting in a lower efficiency of hydrogen abstraction.27
Perusal of Table 2 reveals that, relative to either hexane or
aqueous solution, the photobehavior of every ketone in the series
To recap, these larger guests adopt a well-defined packing
arrangement in which the alkyl group and the proximal phenyl
ring fill the hemispheres; it is the distal ring that fills the
equatorial regions of the capsule. In contrast, guest 2f lies at
the cusp between the two packing modes demonstrated by 2e
and 2g.
Effect of Supramolecular Conformational Control on the
Excited-State Behavior of r-(n-Alkyl) Dibenzyl Ketones. The
NMR data discussed above reveal that (a) all ternary complexes
between R-(n-alkyl) DBKs 2a-h and the capsule formed by 1
are kinetically stable on the NMR time scale, and (b) the guests
fall into three packing motifs modulated by the length of the
R-alkyl chain. In essence, the host is acting as an external
template to promote the formation of distinct guest conformers.
We utilize the structures suggested by NMR to understand the
variations in photochemical behavior between the eight guests
within the capsule.
Following the initial discovery by Turro and co-workers that
the product distribution resulting from photolysis of dibenzyl
ketone could be controlled by micelles,22,23 this reaction has
become the benchmark to assess the efficacy of a medium as a
“cage”.24 The sequence of reactions upon excitation of dibenzyl
ketone can be understood on the basis of Scheme 1 (R ) H).
In solution, as well as in most confined media, the primary
radical pair (RP-1) resulting from R-cleavage decarbonylates
to yield a secondary aryl methyl radical pair (RP-2).24 Detailed
studies have revealed that the changes in product distribution
in organized assemblies are attributable to the restriction on the
translational mobility (diffusion) of the primary and secondary
radical pairs. Photochemistry of dibenzyl ketones has been
investigated in organized assemblies such as micelles, cyclo-
dextrins, and zeolites and in the crystalline state, and a common
thread in these processes is the enhanced formation of 3 with
(25) (a) Quinkert, G.; Tabata, T.; Hickmann, E. A. J.; Dobart, W. Angew. Chem.,
Int. Ed. Engl. 1971, 10, 199-200. (b) Veerman, M.; Resendiz, M. J. E.;
Garcia-Garibay, M. A. Org. Lett. 2006, 8, 2615-2617. (c) Resendiz, M. J.
E.; Garcia-Garibay, M. A. Org. Lett. 2005, 7, 371-374. (d) Garcia-Garibay,
M. A.; Zhang, Z.; Turro, N. J. J. Am. Chem. Soc. 1991, 113, 6212-6218.
(e) Turro, N. J.; Cheng, C. C.; Abrams, L.; Corbin, D. R. J. Am. Chem.
Soc. 1987, 109, 2449-2456. (f) Gould, I. R.; Zimmt, M. B.; Turro, N. J.;
Baretz, B. H.; Lehr, G. F. J. Am. Chem. Soc. 1985, 107, 4607-4612. (g)
Turro, N. J.; Weed, G. C. J. Am. Chem. Soc. 1983, 105, 1861-1868. (h)
Turro, N. J.; Jockusch, S.; Lei, X.-G. J. Org. Chem. 2002, 67, 5779-
5782. (i) Lei, X.-G.; Jockusch, S.; Ottaviani, M. F.; Turro, N. J. Photochem.
Photobiol. Sci. 2003, 2, 1095-1100. (j) Ruzicka, R.; Bara´kova´, L.; Kla´n,
P. J. Phys. Chem. B 2005, 109, 9346-9353. (k) Petrova, S. S.; Kruppa, A.
I.; Leshina, T. V. Chem. Phys. Lett. 2004, 385, 40-44. (l) Roberts, C. B.;
Zhang, J.; Brennecke, J. F.; Chateauneuf, J. E. J. Phys. Chem. 1993, 97,
5618-5623. (m) Frederick, B.; Johnston, L. J.; de Mayo, P.; Wong, S. K.
Can. J. Chem. 1984, 62, 403-410. (n) Johnston, L. J.; Wong, S. K. Can.
J. Chem. 1984, 62, 1999-2005. (o) Hrovat, D. A.; Liu, J. H.; Turro, N. J.;
Weiss, R. G. J. Am. Chem. Soc. 1984, 106, 5291-5295.
(26) (a) Johnston, L. J.; de Mayo, P.; Wong, S. K. J. Am. Chem. Soc. 1982,
104, 307-309. (b) Kraeutler, B.; Turro, N. J. Chem. Phys. Lett. 1980, 70,
266-269. (c) Kraeutler, B.; Turro, N. J. Chem. Phys. Lett. 1980, 70, 270-
275.
(27) (a) Lewis, F. D.; Johnson, R. W.; Johnson, D. E. J. Am. Chem. Soc. 1974,
96, 6090-6099. (b) Lewis, F. D.; Johnson, R. W.; Kory, D. R. J. Am.
Chem. Soc. 1974, 96, 6100-6107. (c) Wagner, P. J. Acc. Chem. Res. 1983,
16, 461-467.
(22) (a) Robbins, W. K.; Eastman, R. H. J. Am. Chem. Soc. 1970, 92, 6076-
6079. (b) Engel, P. S. J. Am. Chem. Soc. 1970, 92, 6074-6075.
(23) (a) Turro, N. J.; Cherry, W. R. J. Am. Chem. Soc. 1978, 100, 7431-7432.
(b) Turro, N. J.; Kraeutler, B. J. Am. Chem. Soc. 1978, 100, 7432-7434.
(24) (a) Turro, N. J.; Kraeutler, B. Acc. Chem. Res. 1980, 13, 369-377. (b)
Turro, N. J. Proc. Natl. Acad. Sci. U.S.A. 1983, 80, 609-621. (c)
Bhattacharjee, U.; Chesta, C. A.; Weiss, R. G. Photochem. Photobiol. Sci.
2004, 3, 287-295. (d) Cozens, F. L.; Scaiano, J. C. J. Am. Chem. Soc.
1993, 115, 5204-5211. (e) Chesta, C. A.; Mohanty, J.; Nau, W. M.;
Bhattacharjee, U.; Weiss, R. G. J. Am. Chem. Soc. 2007, 129, 5012-5022.
(f) Kleinman, M. H.; Shevchenko, T.; Bohne, C. Photochem. Photobiol.
1998, 67, 198-205. (g) Kleinman, M. H.; Shevchenko, T.; Bohne, C.
Photochem. Photobiol. 1998, 68, 710-718.
9
4078 J. AM. CHEM. SOC. VOL. 130, NO. 12, 2008

Excited-State Chemistry of
R
-(n-Alkyl) DBK
A R T I C L E S
2a-h has been dramatically altered within the capsule. The first
important fact is that the chemical behavior of the secondary
radical pair (A, B radical pair, RP-1, Scheme 1) resulting from
all eight ketones is uniformly influenced by the capsule. The
two radicals A and B resulting via decarbonylation are trapped
and forced to combine with or without rearrangement. In all
cases, except R-(ethyl) dibenzyl ketone 2b, no AA or BB was
formed within the capsule. Such a large cage effect on the
secondary radical pair although noteworthy is not unprecedented
in an organized assembly. Similar large cage effect has been
noted earlier in micelles, zeolites, and crystals.24,25 Lack of
formation of AA and BB suggests that the capsular assembly
is stable in the time scale of the lifetime of the secondary radical
pair.5 The structures of the host-guest complexes shown in
Figure 5 account for the presence of cage effect and formation
of AB and rearranged AB products.
efficiencies of product formation are. In the two structures
shown in Figure 5d,e, the “northern” hemisphere is filled with
the alkyl chain and accommodation of an aromatic ring
following the rearrangement to 10 is expected to be difficult.
We believe that templation reduces the efficiency of rotation
of the R-alkyl-substituted benzyl radical resulting from 2d and
2e forces it to recombine with its partner and regenerate 2d
and 2e. Such a situation allows the less favored R-cleavage
process (unsubstituted R-C-C bond) to compete and yield 9;
rotation of the benzyl radical in the southern hemisphere is
readily accomplished, and the packing of the incipient methyl
group in the tapering portion of the cavity ensures a driving
force toward 9. Note that in 2d and 2e the product 9 is formed
at the expense of AB product (Table 2). For example, while in
2c the ratio of {AB + rearranged AB} to the rearranged ketone
is 39:50; in 2d it is 29:64. However, the data at hand preclude
firmly ascertaining the origin of this difference.
One of the most interesting features within the capsule is the
formation of the rearrangement product(s) of the starting ketone.
Such products are formed in less than 10% yield in micelles
and zeolites and are not formed in crystals.25,26 Even more
fascinating is that the methyl-, ethyl-, and propyl-substituted
dibenzyl ketones yield mostly one rearrangement product (10),
whereas butyl- and pentyl-substituted dibenzyl ketones give the
two rearrangement products 9 and 10, and hexyl-, heptyl-, and
octyl-substituted dibenzyl ketones do not photorearrange. For-
mation of 10 as the major product from methyl-, ethyl-, and
propyl-substituted dibenzyl ketones is consistent with the
expectation that the alkyl-substituted R-C-C bond would cleave
more readily than the unsubstituted one.28 Looking at the
structures of the complexes presented in Figure 5, it is clear
that once the alkyl-substituted R-C-C bond cleaves, the
R-substituted benzyl radicals would rotate to optimally fill the
cavity and lead to the rearrangement product 10. Although, in
principle, the unsubstituted R-C-C bond could cleave, rotate,
and yield 9 as the final product, this probably does not occur
because the relative rates of R-cleavage of the substituted and
unsubstituted R-C-C bonds favor the former.28
As we progress to the butyl- and pentyl-substituted DBKs
2d and 2e, the structure of the host-guest complex changes
(Figure 5a-e) and, correspondingly, so does the photochemistry.
Guests 2a-c pack the capsule such that the two aromatic rings
occupy the two hemispheres. Guests 2d and 2e pack the capsule
such that the hemispheres are filled with the alkyl group and
the distal ring. In this packing arrangement, the proximal ring
resides at the equatorial region of the capsule. These butyl- and
pentyl-substituted dibenzyl ketones yield two rearrangement
products 9 and 10 almost in equal amounts. This distinct
behavior falls in between shorter chain DBKs (2a-c) that
mainly give one rearrangement product and longer chain DBKs
(2f-h) that give no rearrangement products. Now the question
is how the normally slow cleaving unsubstituted R-C-C bond
is able to compete with the faster cleaving substituted R-C-C
bond. Once again, it is important to recognize that this change
in chemistry is specific to the capsule and does not occur in
hexane and buffer. We believe that the rates of the R-C-C bond
cleavages are not influenced by the capsule, but that the
Ketones 2f-h do not yield rearrangement products 9 and 10
within the capsule. Structures suggested by 1H NMR provide a
clue to this unexpected behavior. In these three cases, the capsule
is congested with the guest molecule and there is very little
room left for any readjustment of any formed radicals. NMR
data reveal that the guests 2f-h pack the cavity of the capsule
so that the alkyl group and now the proximal ring fill the two
hemispheres, with the distal ring forced to reside in the
equatorial region of the host. In this arrangement, the R-sub-
stituted benzyl radical would not be able to rotate to yield 10
and there is no driving force for the benzyl radical to rotate. It
is interesting to note that the secondary radical pairs stay within
the capsule and yield only AB products. Again, this is suggested
by the structures shown in Figure 5 where ketones 2f-h are
fully accommodated within the capsule and no part of them
protrude outside into the external aqueous medium from the
slowly assembling and disassembling capsule.
The final point we wish to touch upon relates to the products
of Norrish type II reaction. Clearly, in 2h this becomes the major
product (90%). Yields of the Norrish-Yang products are
significantly higher in 2f and 2g compared to that in 2c-e.
Predominance of Norrish-Yang products in 2f-h is consistent
with the structures of the host-guest complexes shown in Figure
5. In these structures, the γ-hydrogen is closer to the carbonyl
chromophore, a feature that most likely enhances the efficiency
of hydrogen abstraction.
Conclusions
The capsule formed by the dimerization of host 1 readily
accommodates the R-(n-alkyl) DBK guests 2a-h. For the series,
however, three different packing motifs are observed. NMR
spectroscopy shows that when the R group is small it is the
two aromatic rings of the guest that occupy the two hemispheres
of the capsule. As the R group increases in length, it displaces
the proximal ring from a hemisphere forcing it to reside in the
equatorial region of the complex. Ultimately, however, as the
R group increases in length further, the proximal ring displaces
the distal ring from its hemisphere, forcing the latter to take up
residence at the equator. These three packing motifs each lead
to distinct photochemical outcomes.
(28) (a) Turro, N. J.; Gould, I. R.; Baretz, B. H. J. Phys. Chem. 1983, 87, 531-
532. (b) Lunazzi, L.; Ingold, K. U.; Scaiano, J. C. J. Phys. Chem. 1983,
87, 529-530. (c) Tsentalovich, Y. P.; Fischer, H. J. Chem. Soc., Perkin
Trans. 2 1994, 729-733. (d) Zhang, X.; Nau, W. M. J. Phys. Org. Chem.
2000, 13, 634-639.
Variation in the photochemistry of the guest within the
capsule, with respect to the R-alkyl chain length of the said
guest, highlights the importance of how a small variation in
9
J. AM. CHEM. SOC. VOL. 130, NO. 12, 2008 4079

A R T I C L E S
Gibb et al.
supramolecular structure can influence the selectivity within a
confined nanoscale reactor. It is quite fascinating to be able to
control the excited behavior of several R-(alkyl) dibenzyl
ketones, all of which behave in an identical manner in solution.
It is common knowledge that enzyme active sites can impart
specificity on the bound substrate. The results presented here
establish that carefully designed synthetic hosts can do likewise,
in terms of both their ability to interact with the guest and how
templation can influence the reaction course of the guests. The
observed selective influence of the capsule formed by cavitand
1 on a variety of guests elucidates specific host designing needs
in achieving a specific goal.
Acknowledgment. C.L.D.G. and B.C.G. gratefully acknowl-
edge the National Institutes of Health for financial support
(GM074031). V.R. and A.K.S. thank the National Science
Foundation for financial support (CHE-0213042 and CHE-
0531802).
Supporting Information Available: 1H NMR, 13C NMR,
COSY, TOCSY, and NOESY spectra. This material is available
free of charge via the Internet at http://pubs.acs.org.
JA7107917
9
4080 J. AM. CHEM. SOC. VOL. 130, NO. 12, 2008