Deep-cavity cavitand octa acid as a hydrogen donor: Photofunctionalization with nitrenes generated from azidoadamantanes

Article abstract of DOI:10.1021/jo301499t

1-azidoadamantane and 2-azidoadamantane form a 1:1 complex with hosts octa acid (OA) and cucurbit[7]uril (CB7) in water. Isothermal titration calorimetric measurements suggest these complexes to be very stable in aqueous solution. The complexes have been characterized by ¹H NMR in solution and by ESI-MS in gas phase. In both phases, the complexes are stable. Irradiation of these complexes (λ > 280 nm) results in nitrenes via the loss of nitrogen from the guest azidoadamantanes. The behavior of nitrenes within OA differs from that in solution. Nitrenes included within octa acid attack one of the four tertiary benzylic hydrogens present at the lower interior part of OA. While in solution intramolecular insertion is preferred, within OA intermolecular C-H insertion seems to be the choice. When azidoadamantanes included in CB7 were irradiated (λ > 280 nm) the same products as in solution resulted but the host held them tightly. Displacement of the product required the use of a higher binding guest. In this case, no intermolecular C-H insertion occurred. Difference in reactivity between OA and CB7 is the result of the location of hydrogens; in OA they are in the interior of the cavity where the nitrene is generated, and in CB7 they are at the exterior. Reactivity of nitrenes within OA is different from that of carbenes that do not react with the host.

Full text of DOI:10.1021/jo301499t

Article
pubs.acs.org/joc
Deep-Cavity Cavitand Octa Acid as a Hydrogen Donor:
Photofunctionalization with Nitrenes Generated from
Azidoadamantanes
Rajib Choudhury,† Shipra Gupta,† Jose P. Da Silva,*,‡ and V. Ramamurthy*,†
́
†Department of Chemistry, University of Miami, Coral Gables, Florida 33124, United States
‡Centro de Investigaca
̧ o em Química do Algarve, FCT, Universidade do Algarve, Campus de Gambelas, 8005-139 Faro, Portugal
̃
S
* Supporting Information
ABSTRACT: 1-azidoadamantane and 2-azidoadamantane
form a 1:1 complex with hosts octa acid (OA) and
cucurbit[7]uril (CB7) in water. Isothermal titration calori-
metric measurements suggest these complexes to be very
stable in aqueous solution. The complexes have been
1
characterized by H NMR in solution and by ESI-MS in gas
phase. In both phases, the complexes are stable. Irradiation of
these complexes (λ > 280 nm) results in nitrenes via the loss of
nitrogen from the guest azidoadamantanes. The behavior of
nitrenes within OA differs from that in solution. Nitrenes
included within octa acid attack one of the four tertiary
benzylic hydrogens present at the lower interior part of OA. While in solution intramolecular insertion is preferred, within OA
intermolecular C−H insertion seems to be the choice. When azidoadamantanes included in CB7 were irradiated (λ > 280 nm)
the same products as in solution resulted but the host held them tightly. Displacement of the product required the use of a higher
binding guest. In this case, no intermolecular C−H insertion occurred. Difference in reactivity between OA and CB7 is the result
of the location of hydrogens; in OA they are in the interior of the cavity where the nitrene is generated, and in CB7 they are at
the exterior. Reactivity of nitrenes within OA is different from that of carbenes that do not react with the host.
intramolecular rearrangement within OA.10a Cucurbiturils with
INTRODUCTION
■
their secondary and tertiary hydrogens (projecting outside the
cavity) are known to be inert toward excited guest molecules
and reactive intermediates derived from them.10b,11−13 This
study explores the reactivity of transient intermediate nitrene
with the host frameworks of OA and cucurbit[7]uril (CB7).
Photolysis of azidoadmantanes is established to generate
reactive nitrenes that undergo intramolecular and intermolec-
ular (solvent) insertion reactions.14−16 We envisioned that an
examination of such systems included within OA and CB7
would provide an opportunity to probe the hydrogen atom
donating ability of these hosts. This approach would also afford
a route to host cavity functionalization. With these in mind we
examined the excited state chemistry of 1-azidoadamantane and
2-azidoadamantane (Figure 1) included within OA and CB7.
The nitrenes generated from these azidoadamantanes reacted
with OA and produced a covalently linked host−guest product
while within CB7 nitrenes yielded products that were
noncovalently bound to the host stronger than the reactants.
Isolation of the product from CB7 required its displacement
The value of deep-cavity cavitands in excited state processes has
gained momentum recently. Octa acid (OA) has recently been
established as an effective reaction vessel in controlling the
excited-state chemistry and physics of included guest
molecules.1−5 OA presents several features that separate it
from other well-known hosts such as cyclodextrins, cucurbitur-
ils (CB), and calixarenes.6 In the presence of a guest, OA
generally forms a capsular assembly that completely incarcer-
ates the guest molecule and provides better confinement and
protection from the aqueous exterior.7 It absorbs in the region
250−300 nm and has an electron-rich molecular framework
and thus upon excitation has a potential to act as an energy and
electron donor to the included guest molecules. Most
importantly, OA with four tertiary benzylic hydrogens (labeled
Hg in Figure 1) in the internal framework of its cavity can be a
hydrogen atom donor to an excited molecule or a reactive
intermediate. This investigation focuses on these abstractable
hydrogens present at the interior of OA. Although products of
reaction between excited guest molecules and cyclodextrin
hosts are known,8,9 no such products have been isolated to date
with OA as host. Neither excited carbonyl compounds nor
radical intermediates derived from them react with OA.
Similarly, highly reactive carbenes generated from adamanta-
nediazirenes that reacted with cyclodextrins only undergo
Special Issue: Howard Zimmerman Memorial Issue
Received: July 16, 2012
Published: August 29, 2012
© 2012 American Chemical Society
1824
dx.doi.org/10.1021/jo301499t | J. Org. Chem. 2013, 78, 1824−1832

The Journal of Organic Chemistry
Article
Figure 1. Structures of host OA, CB7, and azidoadamantanes used in this study.
with a better binding guest molecule. We present the results
with OA first, followed by those with CB7.
RESULT AND DISCUSSION
■
Complexation of Azidoadamantanes with OA. Both 1-
azidoadamantane and 2-azidoadamantane that are sparingly
soluble in water formed a highly soluble 1:1 host−guest
complex with OA in aqueous borate buffer solution. Complex-
ation was evident from the characteristic upfield shift of the
adamantyl hydrogens of these guest molecules (1 and 2) upon
inclusion within OA (Figure 2).17 The host−guest ratio was
1
determined by H NMR titration experiments. As seen in
Figures S1 and S2 (Supporting Information, SI), addition of
more than 1 equiv of the guest to the host resulted in signals
due to uncomplexed guest molecules. The possibility of 1:1
stoichiometry being either a 1:1 cavitandplex or a 2:2
capsuleplex, could be narrowed down by their significantly
different diffusion constants in general of ∼1.5 × 10−6 cm2/s
and ∼1.2 × 10−6 cm2/s, respectively.17 Free OA has a diffusion
constant closer to 1.8 × 10−6 cm2/s. With the help of 2D-
DOSY experiments the diffusion constants for OA complexes
of 1-azidoadamantane and 2-azidoadamantane were determined
to be 1.51 × 10−6 cm2/s and 1.56 × 10−6 cm2/s, respectively,
Figure 2. 1H NMR spectra of (a) OA (1 mM, 10 mM sodium
tetraborate buffer), (b) 1 mM of 1 in 10 mM sodium tetraborate
buffer, (c) 1 in OA (H:G = 1:1, 10 mM sodium tetraborate buffer),
(d) 1 mM of 2 in 10 mM sodium tetraborate buffer, and (e) 2 in OA
(H:G = 1:1, 10 mM sodium tetraborate buffer).
numbers consistent with a 1:1 cavitandplex in solution.
The complexation behavior was further probed by isothermal
titration calorimetric (ITC) measurements.18 Thermodynamic
parameters for complexation of 1 and 2 with OA were
determined by monitoring the heat changes during titration of
the guest into a host solution (for representative ITC data see
Figures S3 and S4 in SI). Since 1 and 2 were not sufficiently
soluble in aqueous solution, all ITC experiments were carried
out in 50% DMSO + 50% water mixture. The association
constant (K), ΔG, ΔH, ΔS, and stoichiometry of the complex
were obtained by fitting the experimental titration curve with
the computed one based on an independent binding model
(Table 1). The binding constants in the range of 105−106 M−1
suggested all complexes to be stable in the above solvent
mixture. Since the guest inclusion within host OA is driven by a
hydrophobic effect, the complexes should be more stable in
100% water in which the photochemical studies were
conducted. We believe that the observed high enthalpies of
binding (∼ca. 8 kcal/mol) result from van der Waals
1825
dx.doi.org/10.1021/jo301499t | J. Org. Chem. 2013, 78, 1824−1832

The Journal of Organic Chemistry
Article
Table 1. Binding Constants (K) and Relevant Thermodynamic Parameters for Complexation of 1- and 2-Azidoadamantanes
with OA and CB7 at 25 °C
a
b
c
d
guest
media
Ka (M−1
)
ΔG (kcal/mol)
ΔH (kcal/mol)
ΔS (kcal/mol)
stoichiometry
1
1
2
2
1@OA
1@CB7
2@OA
2@CB7
1.95( 0.5) × 106
1.2( 0.2) × 106
7.04( 0.2) × 105
1.4( 0.2) × 106
−8.6 0.1
−8.3 0.1
−8.0 0.02
−8.4 0.1
−8.2 0.3
−17.3 0.3
−7.68 0.1
−14.0 0.2
0.4 0.2
−9.0 0.5
0.3 0.1
0.98 0.07
0.90 0.04
0.99 0.03
−5.6 0.3
0.90 0.004
a
b
Mean values measured from at least three ITC experiments at 25 °C in DMSO−water (1:1). Standard deviations are given in parentheses. Gibbs
free energy values calculated from Ka values. Enthalpy values measured by ITC. Entropic contributions to ΔG calculated from Ka and ΔH values. A
solvent mixture of DMSO and sodium borate buffer in water (1:1) was used for all the titrations.
c
d
Figure 3. ESI-MS of an aqueous solution of 1@OA complexes in the presence of NH3 (1 μL/mL): (a) full scan spectrum; (b) fragmentation (MS2)
of m/z 952.6 (1@OA). The arrow indicates the fragmented peak. Assignments: m/z 863.6 [OA − 2H]2−; 952.6 [OA − 2H + 1]2−.
interactions between host and guest. Stoichiometry of the host
to guest complex (1:1) inferred from ITC data (Table 1) is
consistent with the above H NMR titration and 2D-DOSY
it is clear that 1-azidoadamantane and 2-azidoadamantane form
stable 1:1 cavitandplexes with OA.
1
For insight into the orientation of the guest molecules within
OA, 2D-COSY, and 2D-NOESY experiments with 1@OA and
2@OA complexes in D2O borate buffer solutions were
performed. Integration of the peaks and COSY correlations
helped us assign the signals due to various guest protons and
are indicated in the spectra provided in Figures S7 and S8 (SI).
NOESY spectra provided in Figures S9 and S10 (SI) were most
helpful to infer the orientation of azidoadamantanes within OA.
In both cases, we believe that the N3 group is pointed toward
the tapered bottom part of OA, an arrangement that would
place most of the adamantyl protons closer to the opposite,
broader rim. This visualization is consistent with the NOESY
correlations. For example, in the case of 1-azidoadamantane
protons δ1 and δ2 correlate with Hd of the host that is present at
the broader rim of OA. Consistent with this, proton β that is
closer to the N3 group of the guest correlates with Hg and He of
the host that are present at the bottom and middle of the cavity.
Similar analysis in the case of 2-azidoadamantane led us to
conclude that the N3 group is buried at the narrower end of
OA. For example, the proton closer (α and β) to the N3 group
correlates with Hg and He of the host. The two protons (γ)
farther away from the N3 group strongly correlate with Hd and
most importantly show no interaction with Hg, suggesting that
the N3 group is at the bottom and adamantyl part is at the
larger periphery of OA.
results.
Further confirmation for formation of 1:1 complex came
from electrospray ionization mass spectrometry (ESI-MS) data.
The ESI-MS technique provides valuable information regarding
the stoichiometry, reactivity, and binding constants of host−
guest complexes as well as of higher order supramolecular
assemblies in gas phase.19−21 Employing ESI-MS studies, we
recently reported one of the few examples of gas-phase-stable
capsuleplexes (2:2 and 2:1) with octa amine as host.22
However, to our knowledge, there are no known examples of
stable gas phase complexes of OA. Free OA was readily seen in
the gas phase under ESI-MS in both negative and positive
polarities. The spray of an aqueous solution of OA containing
NH3 (1 μL/mL) gave rise to peaks at m/z 863.6 ([OA −
2H]2−) and 1747.4 ([OA + NH3 + H]+) in the negative and
positive modes, respectively (Figure S5, SI). The addition of 1
to OA gave rise to a new doubly charged ion peaking at m/z
952.6 (Figure 3a) under negative ESI-MS. Its fragmentation led
to release of the guest and observation of free OA (Figure 3b).
We therefore assign the peak at m/z 952.6 to 1@OA complex.
A similar behavior was observed for 2@OA (Figure S6, SI).
The azidoadamantane−OA complexes were observed under
gentle ionization conditions, namely low desolvation gas flows
(<2 L/min) and low capillary exit potentials (<150 V). Under
harder ionization conditions (desolvation gas flows >2 L/min
and capillary exit potentials >150 V) only free OA was
observed, indicating that guest is being released. To our
knowledge, this is the first example of a stable gas-phase 1:1
complex of OA host with any guest. As there are no solvent
molecules in the gas phase pure hydrophobic effect is absent.
The observation of gas-phase stable 1:1 cavitandplexes suggests
van der Waals interaction to be likely responsible for the
binding and is in agreement with the observed high enthalpies
of binding. From the above 1H NMR, ITC, and ESI-MS studies
To visualize the structure of the two host−guest complexes
we generated the most likely structures using the molecular
docking program AutoDock Vina.23 Iterated local search global
optimizer with Broyden−Fletcher−Goldfarb−Shanno (BFGS)
method was used for the local optimization.24 A rectangular
region, 10 Å × 9 Å × 9 Å with a grid spacing of 1 Å, was
selected to define the binding site of OA. The atomic partial
charges were calculated by the Kollman method, and other
docking parameters were set as default. The program
determined the total interaction energies between random
1826
dx.doi.org/10.1021/jo301499t | J. Org. Chem. 2013, 78, 1824−1832

The Journal of Organic Chemistry
Article
pairs of ligands and various selected portions of OA to
determine docking poses. For each guest molecule a total of 10
different poses were obtained, and the structures with lowest
energy for 1@OA and 2@OA are provided in Figure 4. It is
readily seen form these cartoon representations that the N3
group in these structures is pointed toward the narrower part of
OA.
is covalently linked to it. To further confirm that we really have
OA-functionalized product, a competitive control experiment
with 1-adamantylamine that is likely to form a stronger complex
was performed. No displacement of photoproduct occurred.
Signals due to 1-adamantylamine in water were observed in the
1H NMR spectrum. In addition to the OA-functionalized
product, a small amount of 6 was also isolated (Table 2).
The photoproduct 5 was further probed by ESI-MS under
positive and negative ionization conditions to obtain further
support for the conclusion that it is the result of reaction
between 1-adamantylnitrene and OA. Figure 6a presents the
mass spectrum of the photoproduct 5 monitored under
conditions similar to those used for the detection of 1@OA
complex (Figure 3a). A new m/z value was observed at 938.2 as
opposed to 952.6 obtained for 1@OA complex. However,
under hard ionization conditions (desolvation gas flows >2 L/
min and capillary exit potentials >150 V), unlike 1@OA
complex (Figure 3a), the ion at m/z 938.2 was stable,
suggesting the formation of a covalently linked system. The
fragmentation of m/z 938.2 (MS2) gave rise to neutral losses of
18 and 44, corresponding to H2O and CO2, respectively
(Figure S15a, SI) a pattern different from that of 1@OA
complex, in which the release of the guest and free OA
formation was observed (Figure 3b). The fragmentation (MS2)
of the doubly charged negative ion of free OA (Figure S15b, SI)
showed the same neutral losses. These observations are
consistent with the photoproduct being a result of a chemical
reaction between 1-adamantanylnitrene and OA. ESI-MS under
positive polarity gave further insight into the structure of the
photoproduct. The spray of an aqueous solution of the
photoproduct in the presence of NH3 (1 μL/mL) in the
positive mode gave rise to a base peak at m/z 1879.6 and to the
already observed signal for free OA (m/z 1747.4) (Figure 6b).
The isotope distributions obtained using time of flight (TOF)
analyzer and the correspondent theoretical patterns near m/z
1879 (Figure S16, SI) are consistent with a product of reaction
between the host OA and 1-adamantanylnitrene or a strong
noncovalent complex between 1-adamantanylnitrene-derived
product and OA. With highly reactive molecules having
precedence26 in being trapped in capsules we fancied that the
above mass peak might correspond to a strong noncovalent
host−guest complex of the highly reactive 4 with OA. The
fragmentations (MS2) of these two ions (1879 and 1747) are
shown in Figure S17 (SI). Unlike the negative ions (Figure S15,
SI), the fragmentation of the single positively charged OA and
photoproduct are different and reflect the presence of the
adamantyl moiety in the photoproduct. While the fragmenta-
tion of OA indicates the loss of NH3 and H2O, the major losses
in the case of the photoproduct are 135 and 150 Da,
corresponding to the adamantyl and amino adamantyl moieties,
respectively (see Figure S17, SI).
Figure 4. Orientations of the azidoadamantanes in (a) 1@OA and (b)
2@OA based on molecular docking study. Color code for atoms: C,
gray; N, blue; O, red; H, white.
Photochemistry of 1-Azidoadamantane@OA and 2-
Azidoadamantane@OA Complexes. Solution photochem-
istry of 1-azidoadamantane and 2-azidoadamantane is known
from literature reports, with 1-azidoadamantane upon photol-
ysis yielding a highly strained reactive imine 4 (characterized
only in argon matrix at 12 K) via the nitrene intermediate 3
(Scheme 1a).14−16,25 In solution, 4 adds to a water molecule to
produce aminoalcohol 6 and/or dimerizes to 7. When
complexes of 1 with α- or β-cyclodextrins were irradiated 6
was obtained as the main product (>90%).25 Unlike cyclo-
dextrins and cucurbiturils, OA absorbs between 250 and 300
nm (Figure S11, SI). This host therefore has the potential to act
as energy and electron donor upon absorption of light and
promote the phototransformation of guest molecules that show
very low extinction coefficients in this spectral region, such as
compounds 1 and 2. In this study, we are not certain whether
1-azidoadamantane and 2-azidoadamantane included within
OA are excited by direct light absorption or via energy transfer
following light absorption by OA.
The cavitandplex 1@OA upon irradiation for 40 min (λ >
1
280 nm) showed complete conversion as evident from H
1
NMR spectrum (Figure S12, SI). The H NMR spectrum of
the isolated major photoproduct (5) contained signals due to
OA as well as the adamantyl protons suggestive of the likely
reaction of the nitrene generated from 1-azidoadamantane with
the host OA (Figure 5). Had a simple noncovalent host−guest
complex been formed between an intramolecular nitrene
product (4) and OA, it would retain the original symmetry,
the signals due to OA should be more simple and similar to that
of host OA. However, the signals due to OA protons of the
isolated product suggested the host to be no longer
symmetrical and the likely insertion of the 1-adamantanylni-
trene to the molecular frame of OA. 2D-NOESY correlations of
the photoproduct (Figure S13, SI) confirmed the interaction
between the adamantyl and aromatic groups (signals in the
region δ 0.2 to −1.2 ppm and δ 5.5 to 8 ppm respectively).
Detailed analysis for the presence of −NH revealed signals at
∼−1.2 ppm and at ∼−0.85 ppm for the photolyzed 1@OA and
2@OA, respectively. Integration of the signals and COSY
interaction analyses indicated the presence of a single −NH
hydrogen in the photoproduct (see Figure S14, SI). This
observation also supports the conclusion that in the major
photoproduct the adamantyl group is within the OA cavity and
Based on the above analysis of mass spectral data, we believe
that the 1-adamantanylnitrene generated from 1-azidoadaman-
tane has inserted itself at the C−Hg site of OA, and the most
likely structure of the product is the one shown in Scheme 1 a.
Formation of this product is consistent with the structure of
1@OA derived from 1H NMR and molecular modeling. In this
structure, the nitrene generated from irradiation of 1 by
extrusion of nitrogen would be very close to the four C−Hg
bonds present at the narrower end of the OA cavity. Nitrenes
and carbenes attacking the aryl units of cavitands and
hemicarcerands and nitrenes inserting into C−H bond of
hemicarcerands are known in the literature.27−29
1827
dx.doi.org/10.1021/jo301499t | J. Org. Chem. 2013, 78, 1824−1832

The Journal of Organic Chemistry
Article
Scheme 1. Photolysis Products of (a) 1 and (b) 2
2-Azidomantane yields an unstrained imine (major) and
adamantanone (minor) via a nitrene intermediate (Scheme 1
b) in solution and when irradiated as a solid or as inclusion
complexes of α- or β-cyclodextrins in water.25 However,
irradiation of 2-azidomantane@OA in aqueous borate buffer
solution (λ > 280 nm) resulted in a nitrene insertion product
and S19, SI) support the conclusion that the photoproduct is
11. Formation of this product is consistent with the structure of
the 2@OA shown in Figure 4. Unstrained imine 9 and
adamantanone 12 were formed as minor products (∼20%)
(Table 2).
Complexation of Azidoadamantanes with CB and
Their Photochemistry. Both 1-azidoadamantane and 2-
azidoadamantane formed a 1:1 complex with CB7 in water.
This conclusion is supported by 1H NMR titration experiments
(Figures S20 and S21, SI) as well as ITC measurements (Table
1
11 similar to 5 with the H NMR spectrum also bearing close
similarity to it (Figure 5). Further, NOESY correlations shown
in Figure S18 in the SI suggest that the adamantyl and OA parts
are linked and the former is tucked inside the later. To probe
this further, the ESI-MS of the photoproduct was recorded
under positive and negative modes. Under the positive mode,
the m/z value (1878.46), the fragmentation behavior, and the
experimental and simulated isotope distributions (Figures S16
1
1 and Figures S22 and S23, SI). The H NMR signals of the
guests 1 and 2 included within CB7, as is often the case with
most guests, were upfield shifted (Figure 7). Mass spectra
recorded by ESI-MS technique gave the mass ion correspond-
1828
dx.doi.org/10.1021/jo301499t | J. Org. Chem. 2013, 78, 1824−1832

The Journal of Organic Chemistry
Article
Figure 7. 1H NMR spectra of (a) CB7 (1 mM, in D2O), (b) 1 in CB7
(H:G = 1:1), (c) 2 in CB7 (H:G = 1:1), * indicates signal for residual
water.
and ΔH and ΔS were more negative than in the case of OA
complexes. These suggested that the guests were more tightly
held within CB7 and less mobile than within OA. Given the
smaller size of CB7 with respect to OA this is not surprising.
Irradiation of 1@CB7 in aqueous solution (λ > 280 nm), as
Figure 5. 1H NMR spectra of (a) 1 in OA (H:G = 1:1, [H] = [G] = 1
mM in 10 mM sodium tetraborate buffer), (b) H NMR spectra of
1
1
per the H NMR spectrum (Figure 8), gave a new product in
product 1@OA in D2O after photolysis followed by extraction with
1
CDCl3, H NMR spectra of (c) 2 in OA (H:G = 1:1, [H] = [G] = 1
1
mM in 10 mM sodium tetraborate buffer), (d) H NMR spectra of
product 2@OA in D2O after photolysis followed by extraction with
CDCl3. * indicates signal for residual water. For signal assignments of
OA, see Figure 1.
Table 2. Relative Percentages of Products upon Photolysis of
Azidoadamantanes in OA and CB7
yield (%)
a
b
c
c
c
guest
media
5
6
7
alkane solution
1@OA
88
1
86
14
>95
1@CB7
yield (%)
c
c
c
9
11
12
Figure 8. 1H NMR spectra (500 MHz in D2O) of (a) 1 in CB7 (H:G
= 1:1, [H] = [G] = 2 mM), (b) photoproduct of 1@CB7 after 1 h
irradiation, and (c) photoproduct of 1@CB7 after irradiation followed
by extraction with CDCl3 (λ > 280 nm). * indicates signals due to
residual water.
alkane solution
2@OA
87
19
5
2
71
10
2@CB7
>95
a
The samples were irradiated using a medium-pressure mercury lamp
b
(Pyrex tube, λ > 280 nm). Products yield were determined by GC
after total conversion with adamantane as internal standard.
c
Concentrations of 1 and 2: OA, CB7 = 1 mM/1 mM.
quantitative yield within 1 h. Further characterizations of the
product from irradiation similar to those described above to
determine complexation of CB7 were carried out (GC and
ing to 1:1 complex (Figures S24, SI). According to ITC
1
measurements, the binding constants were high (∼106 M−1),
GC−MS, ESI-MS, and H NMR). GC and GC−MS indicated
Figure 6. ESI-MS spectra (full scans) of (a) photolyzed 1@OA in the negative polarity, (b) photolyzed 1@OA in the positive polarity. Assignments:
m/z 863.6 [OA − 2H]2−; 938.2 [5 − 2H]2−; 1747.4 [OA + NH3 + H]+; 1879.6 [5 + H]+.
1829
dx.doi.org/10.1021/jo301499t | J. Org. Chem. 2013, 78, 1824−1832

The Journal of Organic Chemistry
Article
Figure 9. ESI-MS spectra (full scans) of (a) photoproduct of 1@CB7 and (b) photoproduct of 1@CB7 in the presence of dimethylviologen2+.
Assignments: m/z 604.4 [CB7 + 2Na]2+; 674.4 [Dimethylviologen@CB7]2+; 1163.4 [CB7 + H]+; 1185.4 [CB7 + Na]+; 1201.4 [CB7 + K]+; 1330.6
[CB7 + 6 + H]+.
the formation of amino alcohol 6 in small amounts (∼10%).
1
The H NMR spectrum of the aqueous extract consisted of
signals due to both CB7 and the adamantyl part of the guest
(Figure 8). The ESI-MS spectrum of the extract of 1@CB7
showed a major peak at m/z 1330.6 (Figure 9a), a value
corresponding to a system containing both 6 and CB7. To get
further insight into the nature of this ion we performed MS/MS
studies. The fragmentation of m/z 1330.6 gave rise to a major
peak at 1312.6, which corresponds to the loss of H2O, together
with free CB7 at m/z 1163.4 and further fragmentation of
former leading to the breaking of the remaining guest and
release of CB7 (Figure S25, SI). The release of guest fragments
upon fragmentation of CB host−guest complexes has been
reported when the binding is strong.30,31 These results could
mean either CB7 was functionalized with 1-adamantylnitrene or
6 was tightly held within CB7. Interestingly when the ESI-MS
of the extract was recorded after the addition of dimethyl
viologen2+ the signal at 1330.6 gave rise to a new signal at 674.4
corresponding to dimethyl viologen2+@CB7 complex (Figure
9b), a result we interpreted as displacement of the photo-
product that was tightly held within CB7. The 1H NMR
spectrum recorded to probe if this displacement could occur in
solution showed noticeable but partial displacement. Addition
of 1-adamantylamine, known to bind to CB7 with much higher
binding constant than dimethyl viologen2+,32,33 to the irradiated
1
Figure 10. H NMR spectra (500 MHz in D2O) of (a) 1@CB7 after
mixture resulted in total displacement of the product 6 to
photolysis followed by extraction with CDCl3, (b) 6 in CB7 (H:G =
1:1, [H] = [G] = 2 mM), (c) addition of 1 equiv of 1-adamantylamine
produce 1-adamantylamine@CB7 exclusively as detected by1H
NMR (Figure 10) and 6 could be quantitatively extracted with
in (a), (d) 1-adamantylamine in CB7 (H:G = 1:1, [H] = [G] = 2
CDCl3 (Figure 10). Thus, irradiation of 1@CB7 in water (λ >
mM), and (e) 6 in D2O. * indicates signals due to residual water.
280 nm) gave quantitatively the amino alcohol 6 that was held
tightly and could only be extracted by displacement. Unlike in
the case of OA there was no reaction of 1-azidoadamantyl
nitrene and CB7.
some cases the product might bind so strongly to the host and
that exchange with a more powerful guest would be the only
way to separation.
Similar to 1@CB7, irradiation of 2@CB7 in aqueous solution
1
gave a new set of H NMR signals that remained even after
SUMMARY
extraction with CDCl3 (Figure S26, SI). Based on ESI-MS data
of the irradiated sample in the absence and presence of
dimethyl viologen2+ and on the fragmentation of the ion at m/z
1312.6 (Figure S27, SI), we concluded that the photoproduct 9
is tightly held within CB7. Thus, irradiation of 2@CB7 gave 9
quantitatively, and there was no reaction between 2-
adamantylnitrene and CB7. The photobehaviors of both 1
and 2 are identical within CB7, and no functionalization of the
host could be carried out with highly reactive nitrenes. This
result is not unexpected if the hydrogens’ location on the
exterior of CB7 is considered. The experience with 1 and 2
within CB7 has demonstrated that one needs to be careful in
isolating products of photoreactions. It is quite likely that in
■
Supramolecular control of photoreactions requires a thorough
understanding of the reaction cavity in which a reaction is being
carried out. This study is focused toward understanding the
reactivity of two hosts, namely OA and CB7. Thus far, both are
known to be inert toward an excited molecule and a reactive
intermediate derived from it. The interior of octa acid contains
four hydrogens that could be susceptible for abstraction by an
excited carbonyl compound, a radical, or a carbene. Thus far,
none of them are reported to react with the host OA. We have
shown in this presentation that nitrenes generated by
photolysis of azidoadamantanes undergo intermolecular C−H
insertion and thereby get attached to OA. As far as we are
1830
dx.doi.org/10.1021/jo301499t | J. Org. Chem. 2013, 78, 1824−1832

The Journal of Organic Chemistry
Article
while stirring at 350 rpm. Electrical calibrations were performed before
each experiment. Nanoanalyze software was used to calculate the
equilibrium constant and standard molar enthalpy of reaction from the
titration curve. Each microcalorimetric titration experiment consisted
of 40−60 successive injections. In each titration, a constant volume (6
μL/injection) of guest solution was injected into the reaction cell (969
μL) charged with host solution. A standard solution of the host and
guest in deionized water−DMSO mixture was diluted to provide a
working solution at 0.1 to 0.15 mM and 1 to 2 mM respectively. The
titrations were performed in 50% DMSO−water solution. The heat of
dilution was calculated as the difference between the data from the last
10 injections (heat generated due to the dilution of the guest into
solvent mixture (DMSO + water)) and those of the host−guest
titration. The accuracy of the calculated thermodynamic quantities for
the 1:1 complexation was checked by performing several independent
titration runs. The reported standard deviation was obtained from the
Nanoanalyze software.
Guest Orientation through Molecular Docking. Molecular
docking (predicting the preferred orientation of one molecule to a
second when bound to each other to form a stable complex) of the
guest molecules with the OA was performed using AutoDock Vina. An
algorithm called Iterated Local Search global optimizer with the
Broyden−Fletcher−Goldfarb−Shanno (BFGS) method was used for
the local optimization according to the interatomic potentials of all
atom types present in the host and guest molecules, including the 12−
6 Lennard-Jones potentials for van der Waals interactions and
coulomb potentials for electrostatic interactions. The program
determined the total interaction energies between random pairs of
ligands and various selected portions of host capsule to determine
docking poses. A grid map of dimension 10 Å × 9 Å × 9 Å with a grid
spacing of 1 Å was placed to cover the OA capsular assembly. The
atomic partial charges were calculated by the Kollman method and
other docking parameters were set as default. For each guest molecule
a total of 10 different poses were obtained and of the various structures
generated, the ones with the lowest energy in each case are reported.
Mass Spectrometry. Sample Preparation. The ESI-MS studies of
OA and OA photoproducts were performed using 100 μM aqueous
solutions prepared with Milli-Q water containing 1 μL/mL of
ammonia. The host−guest complexes were prepared in a 1:1 ratio at
100 μM concentration and using a 50 mM stock solution of each guest
in DMSO and Milli-Q water with 1 μL/mL of ammonia. The solutions
were sonicated for 10 min and allowed to equilibrate for 30 min before
analysis. The complexes were also prepared by addition of a single
crystal of the solid guest to a 100 μM solution of the host, followed by
sonication for 30 min and equilibration overnight. Similar spectral
distributions were obtained for both methods but the signal intensities
are lower when DMSO is used. The CB7 complexes were prepared by
addition of a single crystal of the solid guest to a 100 μM solution of
the host, followed by sonication for 30 min and equilibration
overnight.
aware this is the first report demonstrating that OA is not as
inert as it is believed to be. This observation suggests that one
needs to be watchful when using OA as the host. Along with
OA, we have been employing cucurbiturils as reaction cavities
in our research. In this case, the abstractable hydrogens are at
the exterior. As expected, there was no functionalization of CB
when photolysis of azidoadamantanes included in CB was
carried out. However, we were surprised to note that the
product was noncovalently bound to the host and could not be
extracted by conventional means. The product had to be
displaced with the help of a stronger binding guest. The two
observations we have reported here highlight the challenges in
applying the concepts of supramolecular chemistry in carrying
out selective photoreactions in molecular containers.
EXPERIMENTAL SECTION
■
Materials. Hosts octa acid, cucurbit[7]uril and azidoadamantanes,
and 3-hydroxy-4-azatricyclo[4.3.1.1]undecane (6) were prepared as
reported in literature.6,34,35 All other chemicals of the highest available
purity were purchased from commercial suppliers.
Nature of the Complexes Probed by 1D 1H NMR and 2D
DOSY NMR experiments. To 0.6 mL of a D2O solution of host OA
(1 mM OA in 10 mM Na2B4O7) in an NMR tube were added
increments of 0.25 equiv of guests (2.5 μL of a 60 mM solution in
DMSO-d6) followed by a 5 min sonication after each addition, and 1H
NMR recorded. Formation of the complex was monitored by the
gradual increase in intensity of bound guest signals. After addition of
1.25 equiv of guest, signals due to free, unbound guest were observed.
To a solution of 0.6 mL of 1 mM of CB7 in D2O was added 10 μL of a
60 mM guest solution, the solution was shaken vigorously for 5 min,
and the NMR spectra were recorded.
The diffusion constants (D) of the complexes were determined by
diffusion ordered spectroscopy (DOSY), a technique useful in
separation based on 1H NMR signals and, in particular, as a
supramolecular entity is expected to diffuse slower than the smaller
individual components. All diffusion experiments were performed on a
500 MHz NMR instrument at 25 °C. The concentration of both host
and guest was 1 mM in 10 mM sodium tetraborate buffer. Formation
of the complexes was monitored by adding guests (2.5 μL of a 60 mM
solution in DMSO-d6) to a host solution (1 mM in 10 mM sodium
tetraborate buffer) followed by sonication for 5 min.
Experimental Procedure for Photolysis and Characterization
of Products. Photolysis was performed by irradiation with a medium-
pressure mercury lamp (450 W) of guest@host complex in sodium
tetraborate buffer solution contained in a Pyrex tube (λ > 280 nm),
after purging with nitrogen for 30 min. The conversion was confirmed
1
by comparison of the H NMR spectrum of the solution prior to
1
irradiation and 1D H NMR spectroscopy post irradiation. Complete
ESI-MS System. ESI-MS spectra were obtained using mass
spectrometers equipped with time-of-flight (TOF) and ion trap
analyzers. The ions were continuously generated by infusing the
aqueous sample solution into the source at 4 μL/min. The solutions
were studied in the negative and positive polarities. For the
observation of host−guest complexes of OA soft desolvation
conditions are required. Typical experimental conditions for the
observation of these host−guest complexes in negative polarity were:
capillary voltage (CE): 3.0 kV; capillary exit voltage: −80 V; skimmer
voltage: −15 V; drying gas temperature: 300 °C; drying gas flow: 1 L/
min; nebulizer gas pressure: 40 psi. Host−guest complexes with CB7
can be observed under positive polarity in the following conditions:
capillary voltage (CE): −4.0 kV; capillary exit voltage: 105 V; skimmer
voltage: 35 V; drying gas temperature: 300 °C; drying gas flow: 6 L/
min; nebulizer gas pressure: 30 psi. The photoproducts with OA as
well as the host−guest complexes of the photoproducts with CB7 can
be seen in a with range of spray and ion optics conditions. Typically
the main parameters in the positive polarity were: capillary voltage
(CE): −4.0 kV; capillary exit voltage: 300 V; skimmer voltage: 100 V;
drying gas: 300 °C at 6 L/min; nebulizer gas pressure: 40 psi.
conversion of reactants to products was observed for 1@OA and 2@
OA complexes after irradiation for 40 min (λ > 280 nm), while 1−2@
CB7 solutions were irradiated for 1 h. After the irradiation, the
products were extracted five times using CDCl3 and dried over
anhydrous sodium sulfate. The CDCl3 solution was concentrated to
approximately 0.1 mL and was injected into the GC. The GC program
used for OA samples: starting at 70 °C, held (for 1 min for OA and 2
min for CB7), applying ramp of 10 °C per min, up to 270 °C, held for
10 min. All reactions were repeated (at least three times) for
consistency. The product solution was injected into a GC−MS. The
molecular weight obtained from the molecular ion peak and the
fragmentation pattern confirmed the identity of the products.
Comparison with authentic sample further confirmed the identity of
the products. On the other hand, the aqueous extract was concentrated
to 0.6 mL and further 1D and 2D NMR (COSY, NOESY)
experiments were performed.
Isothermal Titration Calorimetric Study. The equilibrium
constants and thermodynamic parameters for the inclusion of
azidoadamantanes in both OA and CB[7] were determined using an
isothermal titration calorimeter. All titrations were performed at 25 °C
1831
dx.doi.org/10.1021/jo301499t | J. Org. Chem. 2013, 78, 1824−1832

The Journal of Organic Chemistry
Article
Preparation of Compounds 5 and 11. Compound 5.
Compound 5 was obtained after photolysis of 1@OA complexes by
irradiation with a medium-pressure mercury lamp (450 W) in sodium
tetraborate buffer contained in a Pyrex tube (λ > 280 nm), after
purging with nitrogen for 30 min. After irradiation, the products were
extracted five times using CDCl3. The organic layer was discarded and
the aqueous extract was concentrated to 0.6 mL and further 1D and
(11) Masson, E.; Ling, X.; Joseph, R.; Kyeremeh-Mensah, L.; Lu, X.
RSC Adv. 2012, 2, 1213.
(12) Lagona, J.; Mukhopadhyay, P.; Chakrabarti, S.; Isaacs, L. Angew.
Chem., Int. Ed. 2005, 44, 4844.
(13) Lee, J. W.; Samal, S.; Selvapalam, N.; Kim, H.-J.; Kim, K. Acc.
Chem. Res. 2003, 36, 621.
(14) Quast, H.; Eckert, P. Liebigs Ann. Chem. 1974, 1974, 1727.
(15) Dunkin, I. R.; Shields, C. J.; Quast, H.; Seiferling, B. Tetrahedron
Lett. 1983, 24, 3887.
(16) Radziszewski, J. G.; Downing, J. W.; Jawdosiuk, M.; Kovacic, P.;
Michl, J. J. Am. Chem. Soc. 1985, 107, 594.
(17) Jayaraj, N.; Zhao, Y.; Parthasarathy, A.; Porel, M.; Liu, R. S. H.;
Ramamurthy, V. Langmuir 2009, 25, 10575.
(18) Rekharsky, M. V.; Inoue, Y. Chem. Rev. 1998, 98, 1875.
(19) Zadmard, R.; Kraft, A.; Schrader, T.; Linne, U. Chem.Eur. J.
2004, 10, 4233.
(20) Kogej, K.; Schalley, C. A. Analytical Methods in Supramolecular
Chemistry; John Wiley & Sons: Chichester, 2007.
(21) Rozhenko, A. B.; Schoeller, W. W.; Letzel, M. C.; Decker, B.;
Agena, C.; Mattay, J. Chem.Eur. J. 2006, 12, 8995.
(22) Da Silva, J. P.; Kulasekharan, R.; Cordeiro, C.; Jockusch, S.;
Turro, N. J.; Ramamurthy, V. Org. Lett. 2012, 14, 560.
(23) Trott, O.; Olson, A. J. J. Comput. Chem. 2010, 31, 455.
(24) Nocedal, J.; Wright, S. J. Numerical Optomization, 2nd ed.;
Springer Verlag: Berlin, 1999.
(25) Brinker, U. H.; Walla, P.; Krois, D.; Arion, V. B. Eur. J. Org.
Chem. 2011, 2011, 1249.
(26) Warmuth, R. Eur. J. Org. Chem. 2001, 2001, 423.
(27) Lu, Z.; Moss, R. A.; Sauers, R. R.; Warmuth, R. Org. Lett. 2009,
11, 3866.
1
2D NMR (COSY, NOESY) experiments were performed. H NMR
(500 MHz, D2O) δH: −0.94 (1H), −0.65 (6H), −0.4 (3H), 0.05
(3H), 0.18 (3H), 2.1 (8H), 2.5 (8H), 4.4 (2H), 5.7−5.9 (6H), 7.6−7.8
(9H). HRMS (ESI): calcd for the first peak of the isotope series of
C106H80O32N [M + H]+ 1878.4658, found 1878.4543.
Compound 11. Compound 11 was prepared following the
procedure described for compound 5 but using 2@OA complexes.
1H NMR (500 MHz, D2O) δH: −1.7 (2H), −0.9 (1H), −0.85 (2H),
−0.75 (1H), −0.55 (2H), −0.35 (2H), −0.2 (2H), 0.4 (1H), 1.35
(1H), 2.1 (8H), 2.5 (8H), 5.9 (4H). HRMS (ESI): calcd for the first
peak of the isotope series of C106H80O32N [M + H]+ 1878.4658, found
1878.4454.
ASSOCIATED CONTENT
* Supporting Information
■
S
1H NMR titration spectra, ITC data, ESI-MS of host−guest
complexes and photoproducts, 2-D COSY and NOESY spectra
of host−guest complexes, and Cartesian coordinates for docked
host−guest structures. This information is available free of
charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*(J.P.D.S.) Tel: 351 289 800900. E-mail: jpsilva@ualg.pt.
(V.R.) Tel: 305 2841534. E-mail: murthy1@miami.edu.
■
(28) Wagner, G.; Arion, V. B.; Brecker, L.; Krantz, C.; Mieusset, J.-L.;
Brinker, U. H. Org. Lett. 2009, 11, 3056.
(29) Warmuth, R.; Makowiec, S. J. Am. Chem. Soc. 2007, 129, 1233.
(30) Da Silva, J. P.; Jayaraj, N.; Jockusch, S.; Turro, N. J.;
Ramamurthy, V. Org. Lett. 2011, 13, 2410.
Notes
(31) Jayaraj, N.; Porel, M.; Ottaviani, M. F.; Maddipatla, M. V. S. N.;
Modelli, A.; Da Silva, J. P.; Bhogala, B. R.; Captain, B.; Jockusch, S.;
Turro, N. J.; Ramamurthy, V. Langmuir 2009, 25, 13820.
(32) Liu, S.; Ruspic, C.; Mukhopadhyay, P.; Chakrabarti, S.; Zavalij,
P. Y.; Isaacs, L. J. Am. Chem. Soc. 2005, 127, 15959.
(33) Kim, H. J.; Jeon, W. S.; Ki, Y. H.; Kim, K. Proc. Natl. Acad. Sci.
U.S.A. 2002, 99, 5007.
(34) Day, A.; Arnold, A. P.; Blanch, R. J.; Snushall, B. J. Org. Chem.
2001, 66, 8094.
(35) Sasaki, T.; Eguchi, S.; Katada, T.; Hiroaki, O. J. Org. Chem.
1977, 42, 3741.
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
V.R. is grateful to the National Science Foundation for
generous financial support (CHE-0848017) and for funds
toward the purchase of an LC-ESI-MS (CHE-0946858). J.P.S.
thanks the Portuguese Foundation for Science and Technology
(Grant No. REEQ/717/QUI/2005). V.R. thanks U. Brinker
(University of Vienna) for useful discussions.
REFERENCES
■
(1) Ramamurthy, V.; Parthasarathy, A. Isr. J. Chem. 2011, 51, 817.
(2) Parthasarathy, A.; Ramamurthy, V. Photochem. Photobiol. Sci.
2011, 10, 1455.
(3) Sundaresan, A. K.; Ramamurthy, V. Photochem. Photobiol. Sci.
2008, 7, 1555.
(4) Gibb, C. L. D.; Sundaresan, A. K.; Ramamurthy, V.; Gibb, B. C. J.
Am. Chem. Soc. 2008, 130, 4069.
(5) 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.
(6) Gibb, C. L. D.; Gibb, B. C. J. Am. Chem. Soc. 2004, 126, 11408.
(7) Porel, M.; Jayaraj, N.; Kaanumalle, L. S.; Maddipatla, M. V. S. N.;
Parthasarathy, A.; Ramamurthy, V. Langmuir 2009, 25, 3473.
(8) Brinker, U. H.; Buchkremer, R.; Kolodziejczyk, M.; Kupfer, R.;
Rosenberg, M.; Poliks, M. D.; Orlando, M.; Gross, M. L. Angew. Chem.,
Int. Ed. 1993, 32, 1344.
(9) Krois, D.; Brecker, L.; Werner, A.; Brinker, U. H. Adv. Synth.
Catal. 2004, 346, 1367.
(10) (a) Gupta, S.; Choudhury, R.; Krois, D.; Wagner, G.; Brinker, U.
H.; Ramamurthy, V. Org. Lett. 2011, 13, 6074. (b) Gupta, S.;
Choudhury, R.; Krois, D.; Brinker, U. H.; Ramamurthy, V. J. Org.
Chem. 2012, 77, 5155.
1832
dx.doi.org/10.1021/jo301499t | J. Org. Chem. 2013, 78, 1824−1832