Melding Caged Compounds with Supramolecular Containers: Photogeneration and Miscreant Behavior of the Coumarylmethyl Carbocation

Article abstract of DOI:10.1021/acs.orglett.7b01572

By merging well-established concepts of supramolecular chemistry, protecting group strategy, and photochemistry, we have solubilized in water hydrophobic organic molecules consisting of a photoactive protecting group and masked carboxylic acids, released the desired acid, and confined a reactive carbocation intermediate within a capsule. Confinement of the photogenerated carbocation brought out the latent radical-like behavior. This observation is consistent with the recent theoretical prediction of the 7-(diethylamino)coumarinyl-4-methyl carbocation having a triplet diradical ground-state electronic contribution.

Full text of DOI:10.1021/acs.orglett.7b01572

Letter
pubs.acs.org/OrgLett
Melding Caged Compounds with Supramolecular Containers:
Photogeneration and Miscreant Behavior of the Coumarylmethyl
Carbocation
Nareshbabu Kamatham,† Jose P. Da Silva,*,‡ Richard S. Givens,*,§ and V. Ramamurthy*,†
́
†Department of Chemistry, University of Miami, Coral Gables, Miami, Florida 33124, United States
‡CCMAR - Centre of Marine Sciences, University of Algarve, Campus de Gambelas, 8005-139 Faro, Portugal
§Department of Chemistry, University of Kansas, Lawrence, Kansas 66045, United States
S
* Supporting Information
ABSTRACT: By merging well-established concepts of supramolecular chemistry, protecting
group strategy, and photochemistry, we have solubilized in water hydrophobic organic molecules
consisting of a photoactive protecting group and masked carboxylic acids, released the desired
acid, and confined a reactive carbocation intermediate within a capsule. Confinement of the
photogenerated carbocation brought out the latent radical-like behavior. This observation is
consistent with the recent theoretical prediction of the 7-(diethylamino)coumarinyl-4-methyl
carbocation having a triplet diradical ground-state electronic contribution.
Scheme 1. Structures of Water-Soluble Octa Acid Cavitand, 7-
Diethylaminocoumaryl-4-methyl Esters 1−4, Released
Carboxylic Acids 7−10, and Photoproducts 5 and 6
ight control of vectorial, spatial, and timely delivery of
Lmolecules of chemical and biological interest have attracted
intense interest during the past decade.1−3 The strategy conceals
the molecule of interest with a protecting group that releases it
upon exposure to light. Since the molecule of interest is released
with a photon instead of a chemical agent, the group is termed a
photoremovable protecting group (PPG). Thus, a PPG is
covalently linked to the molecule to be released, i.e., the leaving
group (LG).2 Ideally, a PPG−LG should undergo rapid, efficient
photoreaction, be water soluble, absorb in the visible−near-IR,
release the LG at a given location, and discard the spent PPG
after the release. Among the various known PPGs, 7-
(diethylamino)coumaryl-4-methyl (DEACM) derivatives are
currently under intense scrutiny.3−8 In this paper, we explore
the added value of supramolecular concepts to enhance the utility
of the PPG strategy.9,10 As a model, we experiment with the
delivery of carboxylic acids of different sizes in an aqueous
medium using the DEACM derivative as the PPG and water-
soluble octa acid (OA)11 as the supramolecular host. In this
approach, we encapsulate a PPG−LG molecule in the hydro-
phobic interior of an OA capsule and release the carboxylic acid
(LG) therein by exciting the host−guest complex above 400 nm.
The supramolecular encapsulation has enabled us to solubilize
even hydrophobic PPG−LG molecules in aqueous media, and
the host retains the coumarin remains following the release of the
carboxylic acid. More importantly, this approach revealed a thus
far unobserved property of the DEACM carbocation, namely the
triplet diradical.12 Structures of the octa acid host, the PPG−LGs,
the released carboxylic acids, and the coumarin byproducts are
shown in Scheme 1.
provided in the SI. Host−guest complexation was monitored by
UV−vis, fluorescence, and 1H NMR spectra. The PPG
investigated here, 7-(diethylamino)-4-methylcoumarin (5), is
well-known to display solvent-dependent emission and
absorption spectral features including maxima, lifetimes, and
fluorescence efficiencies.13 Therefore, the PPG linked to the
leaving groups in 1−4 serves as a sensitive probe for revealing the
location and orientation of the ester ensemble prior to, during,
and after photolysis.
As illustrated in Figure 1 and Figures S17−S24, the extent of
the shift in absorption and fluorescence maxima in an aqueous
solution between OA complexed and free 1−4 depends on the
structure of PPG−LG. 1H NMR spectral titrations (Figure S32)
as well as DOSY data (Figure S33) suggested that trigger 1,
similar to 5, formed a 1:2 capsuleplex with OA.14 The signals due
Procedures for the synthesis of DEACM-LG 1−4 (Scheme 1),
their spectral data (Figures S1−S16), methods for their host−
guest complexation with OA in phosphate/borate buffer, and
procedures for irradiation and analyses of photoproducts are
Received: May 24, 2017
© XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.7b01572
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
observed in solvent water. It is further possible that longer guests
exceed the OA capsule capacity, as in the case of 4, and expose the
7-(diethylamino)coumaryl group, the less hydrophobic part of
the guest, to external water molecules (for a cartoon visualization
see Figure S48).
For adamantyl ester 4, the emission maximum appears at 490
nm (λmax in water 470 nm), suggesting that the coumaryl
chromophore is in the most highly polar environment compared
with the other three. Consistent with this, the 1H NMR spectrum
revealed 4 to form a 1:1 host−guest cavitandplex with the
adamantyl portion fully within the OA and the chromophore
portion exposed to the solvent H2O (Figure 3).16 Adamantyl
Figure 1. (i) Absorption spectra of 1@(OA)2 (red), 2@(OA)2 (blue),
4@OA (black), and 5@(OA)2 (green); ([guest] = 2.5 × 10−5 M, [OA]
= 5 × 10−5 M in phosphate buffer/H2O, pH = 7.4). (ii) Representative
normalized emission spectra of 1@(OA)2, 2@(OA)2), 4@OA, and
5@(OA)2. Observed shifts with respect to aqueous media absent the OA
are included in parentheses. For spectra of all four triggers, see Figure
S27.
to the methyl groups of OCOCH3 and NCH2CH3 are shifted
upfield (δ ∼1.0 to −2.5 ppm) (Figure 2, see the dotted red lines
Figure 3. 1H NMR spectra (500 MHz) of (i) 4 in CDCl3, (ii) 4@OA
([OA] = 1 mM and [4] = 1.0 mM in 10 mM phosphate buffer buffer/
D2O, pH = 8.7), (iii) 20 min irradiation of (ii), (iv) 10@OA ([OA] = 1
mM and [10] = 1.0 mM in 10 mM Na2B4O7 buffer/D2O, pH = 8.7). The
colored stars “***” indicate the upfield shifted signals due to the guests
included within OA. The green star “*” in (ii) indicates the signals due to
NCH2CH3 exposed to water.
hydrogens were shifted upfield (δ 0.8 to −1 ppm), whereas the
diethylamino hydrogens at δ 3.3 and 1.1, were similar to their
positions in CDCl3 (see the dotted red and blue lines connecting
i and ii). Excess OA had little effect on the positions of the
NCH2CH3 signals, but did broaden them due to the dynamic
equilibria from capping-decapping of the OA cover on the
diethylamino end. Among the four esters, the adamantyl ester 4 is
unusual: (a) The extended length of the adamantyl carboxylate
exceeds the length of the capsule and (b) the more hydrophobic
adamantyl group prefers the OA and exposes the less
hydrophobic DEACM group to water (Figure 3 and Figure S37).
Based on absorption, emission and 1H NMR spectra (Figure 1
and Figures S18, S19, S22, S23, S35, and S36) we conclude that 2
and 3 formed 1:2 elongated capsuleplexes less compact than 1. In
these cases, the emitting DEACM group is partially exposed to
the aqueous media at the middle of the capsule. The dynamic
nature of the varying exposure of these two complexes is revealed
through the broadening of the 1H NMR signals when excess OA
is present (Figures S35 and S36). Thus, the OA complexes of 1-4
provided ample opportunity for irradiation and thus examination
of the photochemistry of DEACM-LG in a confined space while
monitoring the systematic variation of the polarity of their
environment. The effects of these variations on the photo-
chemistry of 1@OA2 (1:2 complex) and 4@OA (1:1 complex)
are described below, whereas that of 2@OA2 and 3@OA2 are
only briefly mentioned (for spectral details, see the SI).
1
Figure 2. H NMR spectra (500 MHz, in 10 mM phosphate buffer/
D2O, pH = 8.7) of (i) OA ([OA] = 1 mM), (ii) 1 ([1] = 0.5 mM), (iii)
1@(OA)2 ([OA] = 1 mM and [1] = 0.5 mM), (iv) 5 h irradiation of (iii),
(v) 6@(OA)2 ([OA] = 1 mM and [6] = 0.5 mM), (vi) 5@(OA)2 ([OA]
= 1 mM and [5] = 0.5 mM), and (vii) acetic acid. The colored stars
“***” indicate the upfield-shifted signals due to the guests included
within OA. The green star “*” indicates the methyl signal due to free
acetic acid in aqueous medium.
connecting (ii) and (iii)), suggesting that these two groups are
deeply submerged in the OA cavities, extending end to end
between the two interior poles of the capsule through C−H---π
interactions.15 Interestingly, the absorption and emission
maxima for OA complexes of 1−4 are not the same (Figure 1
and Figures S17−S24). Had the emitting chromophore in 1−4
been located in the same environment within the OA capsule,
one would expect the absorption and emission maxima to be the
same for all guests. The difference in the maxima is a
consequence of the exposure of the PPG’s fluorophores to
differing populations of water molecules. As the size and length of
the ester group increases (methyl, ethyl and propyl) in 1−3, the
emission maxima shift to longer wavelengths, closer to those
1
Progress of the photoreactions of 1-4 was followed by H
1
NMR, LC-DAD-MS and fluorescence. H NMR provided the
product identities and quantitative estimate of released
B
DOI: 10.1021/acs.orglett.7b01572
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
carboxylic acids, LC-DAD-MS permitted quantitative estimates
of the two coumarin derived products 5 and 6 and fluorescence
monitored the change in the environment around the coumarin
moiety. Although carboxylic acids were quantitatively released,
we were concerned with our inability to obtain a good mass
balance with respect to the coumarin PPG. Early literature
dealing with 5 as a laser dye had identified at least five
photoproducts of dimerization, oxidation and radical reac-
tions.17−19 We suggest that the poor mass balances could be
due to side reactions resulting from 5 and 6 that, as they are
formed, begin to compete for incident light with 1. Although
several small peaks were observed by LC-DAD-MS, sufficient
amounts for spectral identification could not be isolated or
characterized.
products in small amounts. Surprisingly, formation of 5 is
suppressed when the irradiation was conducted under air-
saturated conditions (Figure S42(ii)). We conclude that
formation of 5 only occurs within OA and must be ascribed to
the influence of the capsule on the photochemical divergence
observed.
The photochemical outcome of the 1:1 OA complex of 4 is
clearly different from the 1:2 complexes of 1−3 with OA. A
1
comparison of the H NMR spectrum of the irradiated sample
with that of the authentic 10@OA confirmed that, following the
release, the more hydrophobic adamantyl carboxylic acid still
prefers the hydrophobic OA cavity rather than water (Figure
3(iii) and (iv); see the dotted red lines connecting (iii) and (iv)).
Interestingly, in this case even under nitrogen-saturated
conditions, 6 continues to be the preferred product (Figure
4(iii) and Table S1). Since 4 forms an open OA cavitandplex
(Figure S43) the trigger is exposed to water and probably
undergoes the same reaction as when it is in aqueous media. This
is consistent with our postulate that 5 is formed only within a
hydrophobic closed capsule. Thus, the photobehaviors of 1 and 4
within OA bring out two extremes of the influence of
confinement on the photorelease process.
Plots of product yields vs time as monitored by LC-DAD-MS
are provided in Figure 4 and Figures S40 and S41. Analysis of
Formation of the photoproducts 5 and 6 from the photo-
triggers of 1−3 within OA under nitrogen suggests that both
radical and carbonium ion intermediates are generated during
their photochemistry. But in aqueous solution in the absence of
OA as outlined above only 6 is produced (Figure S45(iii) and
(iv)). To identify whether the triplet (T1) is involved in the
photorelease, the fluorescence intensities were compared under
nitrogen and oxygen saturated conditions (Figure S31). Oxygen
had no effect on the fluorescence intensity of 1@OA2 nor on the
production of the carboxylic acid. However, unexpectedly the
oxygen altered the ratio of 1@OA2 conversion to 5 and 6. LC-
DAD-MS analysis under aerated and nitrogen-saturated
conditions confirmed that 5 is not formed in the presence of
oxygen (Figure S47(ii)). Thus, it is clear that the presence of
oxygen selectively favors the formation of 6. At low conversions,
the initial degradation rates are the same within the experimental
error (Tables S2 and S3). However, at higher conversions,
secondary products formed in the presence of oxygen in small
amounts could not be identified.
The differences in photobehavior of DEACM-LG between
isotropic and confined media resembles the chemistry of
diradical and carbene intermediates in solution and within
OA.23,24 Assuming that the ion pair resulting from DEACM-LG
would also show a difference in behavior between a confined
capsule and an isotropic aqueous solution, we sought the origin
of the unusual formation of 5 on the basis of the confinement
effect. Albright and Winter, through UB3LYP calculations,
proposed that the disjointed DEAC−CH2 cation exists as a
closed-shell singlet in equilibrium with its open-shell triplet.12
They calculated that the open-shell triplet diradical is, in fact, 9
kcal/mol more stable than its closed-shell singlet. In our hands
and as suggested,12 the closed-shell singlet cation in isotropic
aqueous solution is more rapidly trapped by the solvent to give 6.
However, when confined within the hydrophobic OA capsule
and in the absence of nucleophilic agents, the singlet has
sufficient time to intersystem cross to the lower energy open shell
triplet (Scheme 2 and Figure S49).12 While reactions of the
ground-state triplet DEACM carbocation have not been
reported, reduction and hydrogen abstraction are possible
processes that would lead to 5. The absence of this product in
the presence of oxygen could be due to trapping of the triplet
Figure 4. Solvolysis of 1 and 4 under different conditions: (i)
photosolvolysis of 1 in aqueous buffer with OA (λ > 300 nm), N2
atmosphere; (ii) photosolvolysis of 1 in aqueous buffer with OA (λ >
300 nm), under air; (iii) photosolvolysis of 4 in aqueous buffer with OA
(λ > 300 nm), N2 atmosphere; (iv) photosolvolysis of 4 in aqueous
buffer with OA (λ > 300 nm), air equilibrated. [1] = 100 μM, [OA] =
200 μM in 10 mM Na2B4O7).
1
these results as well as H NMR spectra leads to the following
conclusions: (a) The DEACM-LG 1 thermally hydrolyzes very
slowly resulting in the release of acetic acid and formation of 6 in
the buffer solution; ∼ 50% conversion in 10 h (Figure S45(i)).
(b) In contrast, OA protects 1 from thermal hydrolysis and
release of 6 and acetic acid are reduced to ∼5% conversion in 10 h
in the dark (Figure S45(ii)). This confirmed that access of water
1
to the encapsulated 1 is restricted by OA. (c) According to H
NMR, irradiation of 1@OA2 quantitatively releases acetic acid
within 5 h (Figure 2(iv)). (d) 1H NMR spectra of irradiated 1@
OA2 under nitrogen saturated conditions surprisingly revealed an
unexpected photoreduction product 5 along with the expected
PPG hydrolysis product 6 and acetic acid (Figure 2(iv), (v), (vi)
and (vii)). (e) It is clear from the fluorescence maxima (Figure
S43 and S44) and 1H NMR spectra (Figure 2(iv), (v), (vi), and
(vii)) that 5 and 6 formed from the photolysis of 1@OA2 had
remained within the capsule. (f) 1H NMR spectra, and product
analyses by LC-DAD-MS of irradiation of 2 and 3 (Figures S39,
S40 and S46) further confirmed reactivity similar to 1@OA2.
The photochemistry literature on DEACM-LG esters in
aqueous media report hydrolysis to 6 is the major or exclusive
photoproduct.6,7,20−22 In contrast, as illustrated in Figure S47(i),
excitation of 1@OA2 under nitrogen-saturated conditions
resulted in 5 as well as 6 along with many other undetermined
C
DOI: 10.1021/acs.orglett.7b01572
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Notes
Scheme 2. Suggested Mechanism for the Photorelease of
Carboxylic Acids from DEACM Acetate within the Octa Acid
Capsule
The authors declare no competing financial interest.
a
ACKNOWLEDGMENTS
■
The National Science Foundation (CHE-1411458), FCT-
Foundation for Science and Technology (UID/Multi/04326/
2013), and Kansas University Endowment Association,
respectively, are thanked by V.R., JPDS and RSG for financial
support.
REFERENCES
■
(1) Dynamic Studies in Biology: Phototriggers, Photoswitches and Caged
Molecules; Goeldner, M., Givens, R., Eds.; Wiley-VCH: New York, 2005.
(2) Klan, P.; Solomek, T.; Bochet, C. G.; Blanc, A.; Givens, R.; Rubina,
M.; Popik, V.; Kostikov, A.; Wirz, J. Chem. Rev. 2013, 113, 119.
(3) Chitose, Y.; Abe, M.; Furukawa, K.; Lin, J.-Y.; Lin, T.-C.; Katan, C.
Org. Lett. 2017, 19, 2622.
a
For a detailed scheme, see Figure S44.
(4) van Wilderen, L. J. G. W.; Neumann, C.; Rodrigues-Correia, A.;
Kern-Michler, D.; Mielke, N.; Reinfelds, M.; Heckel, A.; Bredenbeck, J.
Phys. Chem. Chem. Phys. 2017, 19, 6487.
(5) Givens, R. S.; Matuszewski, B. J. Am. Chem. Soc. 1984, 106, 6860.
(6) Schmidt, R.; Geissler, D.; Hagen, V.; Bendig, J. J. Phys. Chem. A
2007, 111, 5768.
(7) Suzuki, A. Z.; Watanabe, T.; Kawamoto, M.; Nishiyama, K.;
Yamashita, H.; Ishii, M.; Iwamura, M.; Furuta, T. Org. Lett. 2003, 5,
4867.
(8) Givens, R. S.; Rubina, M.; Wirz, J. Photochem. Photobio. Sci. 2012,
11, 472.
(9) Turro, N. J.; Ramamurthy, V.; Scaiano, J. C. Modern Molecular
Photochemistry of Organic Molecules; University Science Books:
Sausalito, CA, 2010.
(10) Jayaraj, N.; Jagadesan, P.; Samanta, S. R.; Da Silva, J. P.;
Ramamurthy, V. Org. Lett. 2013, 15, 4374.
(11) Gibb, C. L. D.; Gibb, B. C. J. J. Am. Chem. Soc. 2004, 126, 11408.
(12) Albright, T. R.; Winter, A. H. J. Am. Chem. Soc. 2015, 137, 3402.
(13) Jones, G., II; Jackson, W. R.; Choi, C.; Bergmark, W. R. J. Phys.
Chem. 1985, 89, 294.
(14) Porel, M.; Chuang, C.; Burda, C.; Ramamurthy, V. J. Am. Chem.
Soc. 2012, 134, 14718.
diradical by oxygen. Further support for the above rationale
comes from the photobehavior of 4@OA. In this case, 6 is the
major product and 5 is formed only in very minor amounts
(Figure 4(iii) and (iv)), consistent with the model that the
DEACM carbocation is exposed to the aqueous environment
outside the open end of the cavitand. While it is quite possible
that 5 could be formed by a homolytic process, especially within a
nonpolar cavity, our earlier studies with 4-methoxyphenacyl10
and 7-methoxycoumaryl-4-methyl25 derivatives, which are
known to include homolytic components (or intermediates),
formed OA adducts. The absence of such products in the current
system suggested that the cleavage occurs preferentially, if not
exclusively, by a heterolytic process as observed with 4-
hydroxyphenacyl esters.26 The triplet radical most likely abstracts
the benzylic hydrogen present in the interior of the capsule to
yield 5.
By incorporating supramolecular chemistry concepts, we have
been able to expand the utility of the photoprotecting group
strategy to hydrophobic systems and provide a unique method
for removing the remains following the delivery of the desired
molecular target. Our ability to enclose and isolate the PPG−LG
and photogenerate carbocation within a “dry” capsule has created
a miniature single molecule hydrophobic “laboratory” for us to
probe the computational predictions12 regarding the electronic
nature of the DEACM carbocation.
(15) Choudhury, R.; Barman, A.; Prabhakar, R.; Ramamurthy, V. J.
Phys. Chem. B 2013, 117, 398.
(16) Jayaraj, N.; Zhao, Y.; Parthasarathy, A.; Porel, M.; Liu, R. S. H.;
Ramamurthy, V. Langmuir 2009, 25, 10575.
(17) Von Trebra, R.; Koch, T. H. Appl. Phys. Lett. 1983, 42, 129.
(18) Winters, B. H.; Mandelberg, H. I.; Mohr, W. B. Appl. Phys. Lett.
1974, 25, 723.
(19) Jones, I. G.; Bergmark, W. R.; Jackson, W. R. Opt. Commun. 1984,
50, 320.
(20) Shembekar, V. R.; Chen, Y.; Carpenter, B. K.; Hess, G. P.
Biochemistry 2007, 46, 5479.
(21) Schonleber, R. O.; Bendig, J.; Hagen, V.; Giese, B. Bioorg. Med.
Chem. 2002, 10, 97.
(22) Hagen, V.; Bendig, J.; Frings, S.; Eckardt, T.; Helm, S.; Reuter, D.;
Kaupp, U. B. Angew. Chem., Int. Ed. 2001, 40, 1045.
(23) Ramamurthy, V. Acc. Chem. Res. 2015, 48, 2904.
(24) Gupta, S.; Choudhury, R.; Krois, D.; Wagner, G.; Brinker, U. H.;
Ramamurthy, V. Org. Lett. 2011, 13, 6074.
ASSOCIATED CONTENT
* Supporting Information
■
S
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.orglett.7b01572.
Experimental procedures; H NMR, 13C NMR, UV, and
1
ESI-MS spectra for all new compounds; irradiation
procedures; COSY, DOSY, and 1H NMR titration spectra
of host−guest complexes; progress of photoreactions as
1
monitored by H NMR, LC-DAD-MS, and fluorescence
(PDF)
(25) Kamatham, N.; Mendes, D. C.; Da Silva, J. P.; Givens, R. S.;
Ramamurthy, V. Org. Lett. 2016, 18, 5480.
(26) Jagadesan, P.; Da Silva, J. P.; Givens, R. S.; Ramamurthy, V. Org.
Lett. 2015, 17, 1276.
AUTHOR INFORMATION
Corresponding Authors
■
*E-mail: jpsilva@ualg.pt.
*E-mail: givensr@ku.edu.
*E-mail: murthy1@miami.edu.
ORCID
V. Ramamurthy: 0000-0002-3168-2185
D
DOI: 10.1021/acs.orglett.7b01572
Org. Lett. XXXX, XXX, XXX−XXX