Visible light excitation of CdSe nanocrystals triggers the release of coumarin from cinnamate surface ligands

Article abstract of DOI:10.1021/ja063562c

The photochemical properties of organic ligands on the surface of nanocrystalline CdSe particles were examined. A number of thiols carrying a substituted cinnamate tail was synthesized. In solution, these cinnamate compounds undergo light-induced (374 nm) E-Z isomerization, followed by a nonphotolytic lactonization to give highly fluorescent coumarin. The cinnamate-thiols were successfully exchanged onto the CdSe nanocrystal, and the photochemical behavior of these conjugates was studied. Upon aerobic photolysis at 374 nm, the surface cinnamates released coumarin accompanied by rapid nanocrystal degradation. This degradation was not observed under similar anaerobic conditions or when the organic ligands did not contain the cinnamate group. Surprisingly, very similar results were obtained upon irradiation at visible wavelengths at which the cinnamate has no absorption. With the aid of UV-visible absorption spectroscopy, fluorescence spectroscopy, and electrochemistry, a unified theory for both the increased photoinstability of the nanocrystal as well as the coumarin release was proposed. It involves cinnamate radical anions on the CdSe surface, formed upon electron transfer from the excited nanocrystal to the surface cinnamate, undergoing E-Z isomerization. Practically, this results in the remarkable ability to release coumarin from nanocrystal ligands simply by exciting the nanocrystal with visible light. This new photorelease protocol not only aids in the understanding of fundamental nanocrystal-ligand interactions but may also offer new opportunities in the areas of drug delivery and imaging.

Full text of DOI:10.1021/ja063562c

Published on Web 08/11/2006
Visible Light Excitation of CdSe Nanocrystals Triggers the
Release of Coumarin from Cinnamate Surface Ligands
Maikel Wijtmans,*,†,‡,§ Sandra J. Rosenthal,† Binne Zwanenburg,‡ and
Ned A. Porter*,†
Contribution from the Department of Chemistry, Vanderbilt UniVersity, 1822 Station B,
NashVille, Tennessee, and Department of Organic Chemistry, Radboud UniVersity Nijmegen,
ToernooiVeld 1, 6525 ED, Nijmegen, The Netherlands
Received May 22, 2006; E-mail: wijtmans@few.vu.nl; n.porter@vanderbilt.edu.
Abstract: The photochemical properties of organic ligands on the surface of nanocrystalline CdSe particles
were examined. A number of thiols carrying a substituted cinnamate tail was synthesized. In solution, these
cinnamate compounds undergo light-induced (374 nm) E-Z isomerization, followed by a nonphotolytic
lactonization to give highly fluorescent coumarin. The cinnamate-thiols were successfully exchanged onto
the CdSe nanocrystal, and the photochemical behavior of these conjugates was studied. Upon aerobic
photolysis at 374 nm, the surface cinnamates released coumarin accompanied by rapid nanocrystal
degradation. This degradation was not observed under similar anaerobic conditions or when the organic
ligands did not contain the cinnamate group. Surprisingly, very similar results were obtained upon irradiation
at visible wavelengths at which the cinnamate has no absorption. With the aid of UV-visible absorption
spectroscopy, fluorescence spectroscopy, and electrochemistry, a unified theory for both the increased
photoinstability of the nanocrystal as well as the coumarin release was proposed. It involves cinnamate
radical anions on the CdSe surface, formed upon electron transfer from the excited nanocrystal to the
surface cinnamate, undergoing E-Z isomerization. Practically, this results in the remarkable ability to release
coumarin from nanocrystal ligands simply by exciting the nanocrystal with visible light. This new photorelease
protocol not only aids in the understanding of fundamental nanocrystal-ligand interactions but may also
offer new opportunities in the areas of drug delivery and imaging.
Introduction
aggregating into larger insoluble agglomerates. In the classical
CdSe syntheses, the trioctyl-phosphine oxide (TOPO) ligand
Semiconductor nanocrystals (NCs) have received extensive
interest in the past decades.1,2 The existence of the quantum
confinement effect gives rise to spectacular size-dependence of
the nanocrystals properties,3 such as their optical4 and electronic
properties.5 CdSe nanocrystals have been popular objects of
study as well-defined syntheses for these species exist,4,6,7 and
a variety of exciting applications has been disclosed.8 CdSe NCs
have to be stabilized with surface ligands to prevent them from
serves this purpose, providing steric protection of the nanocrystal
surface. Furthermore, it enhances the solubility of the NCs in
organic solvents. However, with the increasingly specialized
applications of NCs, the need for tailor-made and functionalized
surface ligands becomes more and more evident. The original
TOPO ligands offer an excellent starting point toward this
further functionalization. Methodology has been developed to
replace TOPO ligands with compounds such as alkylamines,9
pyrimidines,10 and thiols. The latter have been popular ligands
owing to the strong binding of the SH group to the surface Cd
atoms. For example, many groups have prepared NCs that are
covered with biomolecules through thiol linkers.11 A rapidly
expanding area is the binding of small organic ligands, which
are prepared through multistep synthesis. The attachment of
small organic ligands provides a unique means to direct the
† Vanderbilt University.
‡ Radboud University Nijmegen.
§ Current address: Department of Medicinal Chemistry, Vrije Universiteit
Amsterdam, De Boelelaan 1083, 1081 HV Amsterdam, The Netherlands.
(1) For reviews see: (a) Trindade, T.; O’Brien, P.; Pickett, N. L. Chem. Mater.
2001, 13, 3843. (b) Murray C. B.; Kagan, C. R.; Bawendi, M. G. Annu.
ReV. Mater. Sci. 2000, 30, 545.
(2) For a recent report on the status quo of CdSe NC: Scher, E. C.; Manna L.;
Alivisatos, A. P. Philos. Trans. R. Soc. London Ser. A 2003, 361, 241.
(3) (a) Alivisatos, A. P. J. Phys. Chem. 1996, 100, 13226-13239. (b) Alivisatos,
A. P. Science 1996, 271, 933.
(4) Murray, C. B.; Norris, D. J.; Bawendi, M. G. J. Am. Chem. Soc. 1993,
115, 8706.
(9) Tapalin, D. V.; Rogach, A. L.; Kornowski, A.; Haase, M.; Weller, H. Nano
Lett. 2001, 1, 207.
(5) For an example, see: Ingram, R. S.; Hostetler, M. J.; Murray, R. W.;
Schaaff, T. G.; Khoury J. T.; Whetten, R. L.; Bigioni, T. P.; Guthrie, D.
K.; First, P. N. J. Am. Chem. Soc. 1997, 119, 9279.
(6) Peng, X.; Wickham, J.; Alivisatos, A. P. J. Am. Chem. Soc. 1998, 120,
5343.
(7) Qu, L.; Peng, Z. A.; Peng, X. Nano Lett. 2001, 1, 333.
(8) Michalet, X.; Pinaud, F. F.; Bentolila, L. A.; Tsay, J. M.; Doose, S.; Li, J.
J.; Sundaresan, G.; Wu, A. M.; Gambhir, S. S.; Weiss, S. Science 2005,
307, 538.
(10) Trindade, T.; O’Brien, P.; Zhang, X. J. Mater. Res. 1999, 14, 4140.
(11) (a) Watson, A.; Wu, X.; Bruchez, M. Biotechniques 2003, 34, 303. (b)
Chan, W. C. W.; Shuming, N. Science 1998, 281, 2016. (c) Chunyang, Z.;
Hui, M.; Shuming, N.; Yao, D.; Lei, J.; Dieyan, C. Analyst 2000, 125,
1029. (d) Bruchez, M., Jr.; Moronne, M.; Gin, P.; Weiss, S.; Alivisatos,
A. P. Science 1998, 281, 2013. (e) Mitchell, G. P.; Mirkin, C. A.; Letsinger,
R. L. J. Am. Chem. Soc. 1999, 121, 8122. (f) A° kerman, M. E.; Chan, W.
C. W.; Laakkonen, P.; Bhatia, S. N.; Ruoslahti, E. Proc. Natl. Acad. Sci.
U.S.A. 2002, 99, 12617.
9
11720
J. AM. CHEM. SOC. 2006, 128, 11720-11726
10.1021/ja063562c CCC: $33.50 © 2006 American Chemical Society

Release of Coumarin from Cinnamate Surface Ligands
A R T I C L E S
Scheme 1. Mechanism of Photocleavage of Cinnamate Esters
Table 1. Different Types of Nanocrystals Prepared
properties of the NC without altering its overall size signifi-
cantly. Several examples using thiols, amines, and phosphine
oxides have been reported over the last years, with applications
ranging from the fields of medicinal to polymer chemistry.12
However, in most cases the focus has been on the interaction
of the organic surface ligand with the exterior environment rather
than with the NC itself. Systematic studies on the mutual
dynamic interaction of CdSe NCs with small but functionalized
organic surface ligands have mostly been lacking. Yet, these
interactions will prove relevant in the chemistry of NCs covered
with larger surface (bio)molecules. Moreover, it affords the
chance to probe the intrinsic properties of the NC from a
different point of view, i.e., from its surface layer. Triggered
by these opportunities, we set out to investigate the chemistry
of functionalized organic thiols on the CdSe surface, with a
clear focus on the interaction between the ligand and the NC.
We decided to focus on photoactive organic surface ligands,
as this combines well with the interesting photonic properties
of CdSe NC. The design of the ligands was based on photoactive
serine protease inhibitors, in which our group has had an interest
over the past decades.13,14 The chromophore in these inhibitors
is an o-hydroxy E-cinnamate ester unit (λmax ) 374 nm) that
upon photolysis yields the Z-isomer, which subsequently lac-
tonizes to release a coumarin unit and an alcohol or phenol14
(Scheme 1). The p-diethylamino substituted cinnamate proved
to be most efficient as serine protease inhibitors, giving
7-diethylaminocoumarin (1, hereafter denoted as “coumarin”)
as product.15 Conveniently, coumarin is highly fluorescent (λex
) 374 nm, λem ) 438 nm) whereas the cinnamate precursors
are not, providing excellent means to follow this chemistry in
situ by fluorescence spectroscopy. Inspired by this, we selected
the cinnamate unit depicted in Scheme 1 as the key feature of
a synthetic thiol that was attached to the surface of CdSe NC.
In this paper, we report on the synthesis and photochemical
studies of the NC covered with this photoactive thiol, confirm
the existence of photointeraction between the cinnamate unit
and the NC and propose a mechanism for this interaction.
Remarkably, we demonstrate that isomerization and subsequent
NC type
size NC (Å)
ligand(s)
I
II
III
IV
V
31
31
31
31
31
31
2,6
3,6
4,6
5,6
6
VI
7
lactonization of the surface ligand to release coumarin can be
initiated simply by exciting the nanocrystal.
Results and Discussion
Synthesis and Characterization of NC-Thiol Conjugates.
The ideal ligand consists of a thiol group that attaches to the
CdSe surface, a spacer to provide steric flexibility and the
photoactive cinnamate unit. Thiol 2 satisfies these demands. It
was prepared from a functionalized Wittig reagent and a
substituted salicylaldehyde in 5 steps (see Supporting Informa-
tion). For purposes discussed later, derivatives 3-5 were
prepared analogously. All cinnamates thus made had UV-
visible absorption maxima at 374-376 nm. The ligands were
attached to the NC by stirring pyridine-coated CdSe nanocrystals
(py-NC)16 in deoxygenated HCl-free chloroform with ligands
2-5 and 3 equiv of 6.17 Inert coligand 6 was applied to induce
better solubility and to ensure high dilution of the cinnamates
on the surface, so that direct interaction between two cinnamates,
once on the surface, could be minimized. The exchange was
followed by precipitation, removal of any free ligand by
extensive washing, and drying (see Table 1). NCs coated with
just 6 or mercapto-undecanol (7) were prepared for reference
purposes.
(12) (a) Peng, X.; Wilson, T. E.; Alivisatos, A. P.; Schultz, P. G. Angew. Chem.,
Int. Ed. Engl. 1997, 36, 145. (b) Wang, Y. A.; Li, J. J.; Chen, H.; Peng, X.
J. Am. Chem. Soc. 2002, 124, 2293. (c) Rosenthal, S. J.; Tomlinson, I.;
Adkins, E. M.; Schroeter, S.; Adams, S.; Swafford, L.; McBride, J.; Wang,
Y.; DeFelice, L. J.; Blakely, R. D. J. Am. Chem. Soc. 2002, 124, 4586. (d)
Tomlinson, I. D.; Mason, J.; Burton, J. N.; Blakely, R.; Rosenthal, S. J.
Tetrahedron 2003, 59, 8035. (e) Guo, W.; Li, J. J.; Wang, A.; Peng, X. J.
Am. Chem. Soc. 2003, 125, 3901. (f) Guo, W.; Li, J. J.; Wang, A.; Peng,
X. Chem. Mater. 2003, 15, 3125. (g) Sill, K.; Emrick, T. Chem. Mater.
2004, 16, 1240. (h) Skaff, H.; Emrick, T. Angew. Chem., Int. Ed. 2004,
43, 5383. (i) Javier, A.; Yun, C. S.; Sorena, J.; Strouse, G. F. J. Phys.
Chem. B 2003, 107, 435. (j) Liu, J.; Tanaka, T.; Sivula, K.; Alivisatos, A.
P.; Frechet. J. M. J. J. Am. Chem. Soc. 2004, 126, 6550. (k) Querner, C.;
Reiss, P.; Bleuse, J.; Pron, A. J. Am. Chem. Soc. 2004, 126, 11574. (l)
Uyeda, H. T.; Medintz, I. L.; Jaiswal, J. K.; Simon, S. M.; Mattoussi, H.
J. Am. Chem. Soc. 2005, 127, 3870.
All NCs show continuous absorption far into the visible region
(Figure 1A), a property typical for CdSe NC,4 with the first
excitation peak at 560 nm. A key feature from these UV-visible
absorption spectra is the increased absorption around 374 nm
of Type I NC as compared to Type V. The differential
absorption peak centered at 374 nm (see inset of Figure 1A)
(13) (a) Porter, N. A.; Bruhnke, J. D.; Koenigs, P. Bioorg. Photochem. 1993, 2,
197. (b) Thuring, J. W.; Li, H.; Porter, N. A. Biochemistry 2002, 41, 2002.
(14) (a) Turner, A. D.; Pizzo, S. V.; Rozakis, G.; Porter, N. A. J. Am. Chem.
Soc. 1988, 110, 244. (b) Turner, A. D.; Pizzo, S. V.; Rozakis, G.; Porter,
N. A. J. Am. Chem. Soc. 1987, 109, 1274.
(16) Kuno, M.; Lee, J.; Dabbousi, O.; Mikulec, F.; Bawendi, M. G. J. Chem.
Phys. 1997, 106, 9869.
(17) This same ligand has been used to coat gold nanocrystals: Zheng, M.; Li,
Z.; Huang, X. Langmuir 2004, 20, 4226.
(15) Porter, N. A.; Bruhnke, J. D. J. Am. Chem. Soc. 1989, 111, 7616.
9
J. AM. CHEM. SOC. VOL. 128, NO. 35, 2006 11721

A R T I C L E S
Wijtmans et al.
at 2.8 ppm, analogously to shifts reported in the literature.21,24
Evidently, upon dissolving the NC some 25% of surface thiols
were released and migrated into solution. This was confirmed
by subjecting freshly dissolved Type IV NC to centrifugation
through a size-exclusion membrane, which resulted in recovery
of ∼40% of the cinnamate ligands in the filtrate. These findings
indicate that upon NC dissolution, an equilibrium between free
and bound thiol ligand was established, which is in agreement
with a recent report describing reversible ligand detachment of
a variety of ligands on CdSe NCs.22
Combining the RBS spectrum (see Supporting Information)
1
and the H NMR spectrum, we were able to roughly estimate
the stoichiometry on the surface of Type IV NC. It followed
that, on average, the surface of 1 nanocrystal before dissolution
was covered with ∼160 molecules of TOPO, ∼60 of PEG-thiol
6, and ∼15 of cinnamate 5, affording almost 100% total
coverage based on available surface Cd atoms.20 Clearly, the
single py-exchange was only moderately effective in replacing
TOPO, as observed before.16 The coverage numbers for Type
IV NC obtained from RBS/NMR and UV-visible absorption
spectroscopy (15 and 5 cinnamates per NC, respectively) both
confirm the important notion that there is low coverage of the
dry nanocrystals by the cinnamates of interest. The quantitative
discrepancy between the two can be attributed to the relatively
large errors in the RBS/NMR estimation, and as such coverage
numbers obtained from UV-visible absorption spectroscopy
were used throughout the rest of the work. The low coverage
likely holds true for all types of NC, because the same exchange
procedure was used for all NC and incorporation of an
iodobenzyl-group, located on the outer spheres of the surface
monolayer, would have only little effect on the stacking
efficiency.
Photolysis Studies: Photostability. Photolysis studies were
performed on all types of NC in DMF solution at similar
concentrations (optical density at the first excitation peak
∼0.15). It is well-known that aerobic photolysis of CdSe NC
results in decomposition.23,24 Peng et al performed an extensive
study on this process using thiol-coated CdSe NC and irradiation
at 254 nm.24 Figure 2A shows the evolution of the UV-visible
absorption spectrum of Type I NC upon aerobic irradiation at
374 nm. After an incubation period (ca. 1 h), the first excitation
peak shifts from 560 to 542 nm over 6 h accompanied by general
washing out of absorption pattern and precipitation of NC,
indicating NC decomposition through oxidation of surface Cd
and Se atoms.23,25,26 Surprisingly, Type I NC decomposed
dramatically faster than Type V NC (Figure 2B) under identical
conditions; the latter only started to decompose after >15 h of
irradiation. We think that Type V NC shows the moderate
photoinstability reported for NC coated with long-chain thiols24
and that Type I NC actually undergoes unusually rapid
photodegradation. Superoxide (O2-) was found to be a reactive
oxygen species involved in this photodegradation.27 This was
Figure 1. (A) Representative UV-visible absorption spectra for Types I
and V CdSe NC in DMF. (Inset) 325-425 nm spectrum obtained after
subtraction of spectrum of Type V from spectrum of Type I. (B) 1H NMR
of Type IV NC in CDCl3 with integration of key-peaks. Peaks marked with
X are residual solvent peaks (CHCl3 and a trace of CH2Cl2).
stems from cinnamate ligands, either surface bound or released
to the solvent (vide infra). Using the ratio of this differential
absorption peak versus the NC absorption at 560 nm in
conjunction with the ꢀ-values for CdSe NC18 and for cinnamates
2-5 at 374 nm (see Supporting Information), it was determined
that 5-8 cinnamates were on the surface of Types I-IV NC
before dissolution.19 NCs covered with iodinated ligand 5 (Type
1
IV) were prepared for detailed analysis with H NMR spec-
troscopy and Rutherford Back Scattering (RBS).20 The 1H NMR
spectrum of Type IV NC in CDCl3 (Figure 1B) showed proton
resonances for the cinnamate downfield protons (vinylic and
aromatic, 6.0-8.0 ppm) and the alkyl protons (ca. 3.5 ppm) of
the glycol coligand (6). None of the protons of the cinnamate
π-system (6.0-8.0 ppm) broadened extensively, which indicated
that the NC surface monolayers were reasonably homogeneous.
A signal for the unbound cinnamate thiol (5), a sharp quartet B
for CH2SH group, is visible and it contributed only 25% to the
“total” amount of 5 as established by comparison to the
integration of the aromatic protons and the benzylic group at
5.1 ppm. The same observation was made for coligand (6,
quartet A). The shifted CH2S signals for the corresponding
bound thiols were buried under the multitude of peaks starting
(21) Majetich, S. A.; Carter, A. C.; Belot, J.; McCullough, R. D. J. Phys. Chem.
1994, 98, 13705.
(22) Kalyuzhny, G.; Murray, R. W. J. Phys. Chem. B 2005, 109, 7012.
(23) Xu, L.; Wang, L.; Huang, x.; Zhu, J.; Chen, H.; Chen. K. Physica E 2000,
8, 129.
(24) Aldana, J.; Wang, Y. A.; Peng, X. J. Am. Chem. Soc. 2001, 123, 8844.
(25) Peng, X.; Schlamp, M. C.; Kadavanich, A. V.; Alivisatos, A. P. J. Am.
Chem. Soc. 1997, 119, 7019.
(26) Bowen-Katari, J. E.; Colvin, V. L.; Alivisatos, A. P. J. Phys. Chem. 1994,
98, 4109.
(27) Spanhel, L.; Haase, M.; Weller, H.; Henglein, A. J. Am. Chem. Soc. 1987,
109, 5649.
(18) Yu, W. W.; Qu, L.; Guo, W.; Peng, X. Chem. Mater. 2003, 15, 2854;
Chem. Mater. 2004, 16, 560. The following ꢀ-value for 31 Å CdSe NC
was taken from these papers: 110 000 cm-1 M-1 at 560 nm.
(19) The UV subtraction method was verified for Type IV NC by recording the
1H NMR spectrum in the presence of an internal standard and by measuring
a UV-visible absorption spectrum. The number of ligands present on the
dry nanocrystals, as calculated by these two techniques, matched within
10-20%.
(20) Taylor, J.; Kippeny, T.; Rosenthal, S. J. J. Clust. Sci. 2001, 12, 571.
9
11722 J. AM. CHEM. SOC. VOL. 128, NO. 35, 2006

Release of Coumarin from Cinnamate Surface Ligands
A R T I C L E S
points toward electron transfer from excited NC (NC*) to O2
as the key-step.27,30 A striking finding was that the accelerated
photodegradation of Type I NC occurred at wavelengths all the
way to 560 nm and only then when O2 was present (see
Supporting Information).31 The cinnamate ligands having no
absorption >440 nm, this excluded excited surface cinnamates
as triggers in the NC decomposition.
NC covered with various cinnamates related to 2 showed the
same behavior. Thus, replacement of the phenolic -OH in Type
I NC by a -OMe (Type III NC) or reduction of the spacer
length to 6 carbons (Type II NC) afforded NC that also
photodegraded easily at 374 or 560 nm, albeit with a longer
incubation time for Type III NC (see Supporting Information).
In sharp contrast to Types I-III, cinnamate-lacking Type V and
VI NC did not show significant photodecomposition under the
employed conditions. Importantly, no degradation occurred
either upon irradiation of Type V NC in the presence of free
cinnamate 8 (see Supporting Information), emphasizing that the
enhancement in photodegradation for Types I-III is solely a
result of bound cinnamates and that cinnamates released to the
solvent (vide supra) do not contribute appreciably. Conceivably,
cinnamate surface ligands enhanced the key step in the degrada-
tion, i.e., electron transfer to O2.32 Proof for this was achieved
by adding the external electron donor N,N,N′,N′-tetramethyl-
phenylenediamine (TMPD) or acceptors tetranitromethane
(TNM) and CBr4, the latter both being better acceptors than
O2.33 In the presence of TNM, Type I NC photodegraded
dramatically faster than with just air (Figure 4A). Similar results
were observed for Type II NC when CBr4 (1.0 mM) was used
(not shown). On the other hand, 1 mM of TMPD completely
inhibited photodegradation of Type I, even after 24 h of
irradiation.34 Interestingly, Type V NC photodegraded very fast
in the presence of 100 µM TNM, in sharp contrast to their
insignificant degradation with just air (Figure 4B). An experi-
ment in which light, then air, were introduced stepwise showed
that only light and not air was necessary for Type V NC to
rapidly degrade in the presence of TNM, i.e., NC* reacted
directly with TNM. Thus, the instability of the NC is strictly
regulated by the ease with which an electron from NC* can be
transferred to any electron acceptor in solution.
Figure 2. Evolution of absorption spectra during aerobic irradiation of
Type I NC and Type V NC in DMF at 374 nm. (A) Three-dimensional
representation of full-scale UV-visible absorption spectra for Type I. (B)
Time-evolved first excitation peaks for Types I and V.
Photolysis Studies: (Photo)Chemistry of Surface Ligands.
Having established a dramatic effect of the cinnamate ligands
on the photostability of the NC, we turned to an investigation
of the chemistry of the cinnamate ligands to probe if this
photoeffect also revealed itself in the organic coating. Toward
this end, we analyzed for coumarin 1 by in situ fluorescence
spectroscopy during all photolyses described previously.35 Upon
(29) Li, M. Y.; Cline, C. S.; Koker, E. B.; Carmichael, H. H.; Chignell, C. F.;
Bilski, P. Photochem. Photobiol. 2001, 74, 760.
(30) Uchihara, T.; Kamiya, T.; Maedomari, S.; Ganeko, M.; Kinjo, S.; Tamaki,
Y.; Kinjo, A. J. Photochem. Photobiol. A 2000, 130, 63.
(31) Minor differences in rates of decomposition were attributed to the spectral
intensities of the mercury lamp that was used.
(32) A similar explanation has been reported for the high tendency of excited
CdS microcrystals, covered with viologen-functionalized thiols, to donate
electrons to solution acceptors: Torimoto, T.; Maeda, K.; Maenaka, J.;
Yoneyama, H. J. Phys. Chem. 1994, 98, 13658.
(33) (a) Hiller, K. O.; Asmus, K. D. J. Phys. Chem. 1983, 87, 3687. (b) Sawyer,
D. T.; Chiericato, G.; Angelis, C. T.; Nanni, E.; Tsuchiya, T. Anal. Chem.
1982, 54, 1720.
(34) A related substance, phenylenediamine, is known to be a scavenger of
photogenerated CdSe holes: Sharma, S. N.; Pillai, Z. S.; Kamat, P. V. J.
Phys. Chem. B 2003, 107, 10088.
(35) The fluorescence counts for coumarin could only be used qualitatively since
we found that part of the coumarin emission was quenched by the NC (not
shown).
Figure 3. Irradiation of Type I NC at 374 nm in the presence or absence
of 2.2 mM KO2-crownether complex and/or air. For the curve with open
triangles (no air, KO2), air was let in the system after 4 h.
deduced from the increased oxidation of epinephrine28 to
-
adrenochrome by formed O2 (see Supporting Information).
-
However, O2 itself did not induce NC photodegradation.
Rather, molecular oxygen (O2) was necessary to initiate the
photodecomposition (Figure 3). Furthermore, singlet oxygen
(1O2) was found not to be actively involved as its quencher NaN3
did not inhibit photodegradation (results not shown).29 All this
(28) Misra, H. P.; Fridovich, I. J. Biol. Chem. 1972, 247, 3170.
9
J. AM. CHEM. SOC. VOL. 128, NO. 35, 2006 11723

A R T I C L E S
Wijtmans et al.
Figure 5. (A) Formation of coumarin from Type I NC upon (an)aerobic
irradiation at 374 nm. (B) Formation of coumarin from Type I NC upon
anaerobic irradiation at 374 nm in the presence of 2.2 mM KO2/crownether
complex. Light source was turned on after 1 h in Graph B. Coumarin
fluorescence was measured with λex ) 374 nm and λem ) 438 nm.
Figure 4. (A) Aerobic 560 nm irradiation of Type I NC in the absence or
presence of additives TNM or TMPD. Beyond 20 min of photolysis time
for the TNM experiment, decomposition was too extensive for accurate
analysis. (B) 374 nm irradiation of Type V NC in the presence and absence
of 0.1 mM TNM. In the stepwise experiment, light was turned on after 80
min, followed by the introduction of air after 180 min. (C) Structures of
model cinnamate compounds.
photolysis of Type I NC, the following observations were
made: (1) Coumarin was formed at 374 nm and at a faster rate
aerobically than anaerobically (Figure 5A). (2) Epinephrine, a
superoxide quencher, reduced the rate of coumarin formation
at 374 nm (not shown). (3) Coumarin was formed upon
irradiation with a variety of wavelengths, i.e., from 374 to 560
nm (Figure 6). The first two observations can simply be
explained by assuming that the rate-determining dark lacton-
ization step (see Suppporting Information) is accelerated by the
strongly basic superoxide anion. Indeed, adding KO2 dramati-
cally increased coumarin formation at 374 nm (Figure 5B).36
The third observation, however, is of a considerably more
intriguing nature. This coumarin formation at visible light was
reproducible for Type II NC, whereas Type III, lacking the
Figure 6. Coumarin formation during aerobic irradiation of Type I NC at
various wavelengths. Coumarin fluorescence was measured with λex ) 374
nm and λem ) 438 nm.
phenol group, failed to give any increase in fluorescence at 438
nm. The cinnamate unit has no absorption >440 nm, and hence,
these findings necessarily imply that the isomerization of the
surface cinnamates >440 nm proceeds by a mechanism different
from the photochemical isomerization depicted in Scheme 1.
Air-free photolyses of Types I and II at 560 nm followed by
separation and subsequent quantitative analysis of NC and
filtrate were carried out to estimate the efficiency of the
coumarin formation (see Supporting Information). The NC cores
were still perfectly intact as the first excitation peak had not
shifted. The 374 nm absorption highlighted in Figure 1A,
however, is no longer found in the NC but can be fully recovered
in the filtrate. Amounts of coumarin in the filtrates were
quantified using both fluorescence spectroscopy and GC analy-
sis, which matched acceptably. In all cases, GC analysis also
(36) It has been known that the superoxide anion has reasonable stability in
DMF (Boon, P. J.; Olm, M. T.; Symons, M. C. R. J. Chem. Soc., Faraday
Trans. 1988, 84, 3341) The authors do mention that part of the superoxide
anion adds to the carbonyl group of DMF, but this species is presumably
also strongly basic.
9
11724 J. AM. CHEM. SOC. VOL. 128, NO. 35, 2006

Release of Coumarin from Cinnamate Surface Ligands
A R T I C L E S
Scheme 2. Photodegradation of All Types of NC in the Presence
revealed substantial amounts of free coligand 6 in both dark
and irradiated samples, usually in 20-30% range. This is in
full agreement with the previously discussed equilibrium
between bound and free ligand. The “photolysis efficiency” was
approximated by dividing the amount of coumarin in the filtrate
by the coverage numbers deduced from the UV-visible
absorption spectra (Figure 1A), i.e., the percentage of cinnamate
ligands that were converted to coumarin. Types I and II gave
high photolysis efficiencies of 93 ( 5% and 83 ( 3% (n ) 2),
respectively, upon 24 h anaerobic irradiation at 560 nm.37
of Excellent Electron Acceptorsa
a R ) ligand. A ) electron acceptor.
Mechanism of Double Bond Isomerization. Intuitively, the
mechanisms for cinnamate-accelerated NC decomposition and
coumarin formation at >440 nm seem intimately related. The
observation that coumarin formation at 560 nm was inhibited
almost completely by adding the electron donor TMPD (see
Supporting Information) solidly indicates that electron transfer
also plays a key role in the rotation of the double bond. Electron
transfer from NC* to nearby (organic) acceptors has been
reported for CO2,38 nitroxides,39 Au NC connected by small
organic spacers,40 and a series of quinones.41 In our case,
electron transfer leads to a surface cinnamate radical anion.
Gratifyingly, we were able to prove by electrochemistry that
such cinnamate radical anions can rotate around the double bond
without decomposition (see Supporting Information). Some
electrochemical isomerizations of Z-isomers of dimethylmaleate
and 1,2-diarylsulfonylethenes via radical anions are known.42
Cyclic voltammograms (CVs) of model cinnamates 8 and 9a,b
recorded in DMF43 showed an oxidation peak around 0.8-0.9
V and a reduction peak around -1.3 to -1.5 V. We performed
dark bulk electrolyses in DMF in conjunction with HPLC
analysis.43 Electrolysis of 8 at -1.5 V gave little formation of
coumarin in 3 h, meaning that either the E-Z isomerization or
lactonization is not favored under these particular conditions.
Importantly, no decomposition of the radical anion of 8 was
observed.44 A clearer picture was obtained from the electrolysis
of Z-MeO-model 9b at -1.5 V, in which no lactonization step
is involved. After 3 h, 4% E-isomer (9a) had formed whereas
there was no decomposition of either 9a or 9b. Prolonged
electrolysis led to ca. 10% isomerization after 20 h.
Scheme 3. Decomposition of NC Covered with Cinnamate
Ligands and Possible Mechanisms for NC-Induced Isomerization
of the Cinnamate Ligandsa
a “Cinn” represents both Z- and E-isomers.
chemistry of the surface ligands at wavelengths > 440 nm.45
Despite the fact that the bound ligands are in equilibrium with
solution ligands (eq 1), we believe all relevant steps take place
with bound ligands, i.e., at the NC surface.
Upon irradiation of a nanocrystal (Scheme 2), an electron
and hole are formed in the excited state of the NC (eq 2). Part
of the time, these two will recombine to emit fluorescence light.
If an excellent electron acceptor A is in solution (TNM or CBr4,
to a much lesser extent O2), an electron can be transferred from
the NC* to A leaving an unpaired hole in the NC (eq 3). This
NC cation (NC+) induces ligand oxidation and cleavage and
subsequent surface oxidation leading to decomposition (eq 4).25
If cinnamate ligands cover the NC surface, an additional
triggering mechanism can take place even without excellent
electron acceptors in solution (Scheme 3).
General Mechanism for Enhanced Photodegradation and
Coumarin Release. We propose the following general mech-
anism to explain both the enhanced photodecomposition of the
NC upon attachment of cinnamate ligands as well as the
After absorption of light (eq 5), the excited NC transfers an
electron to the nearby cinnamates π-system (eq 6) at a rate much
higher than transfer to O2 in solution. The resulting cinnamate
radical anion ([cinn]•-) has several pathways available to it.
Most importantly, it can rotate around its double bond during
its lifetime (eq 7). Furthermore, it can transfer its electron, either
back to the NC (eq 6) or into solution provided a reasonable
electron acceptor such as O2 is present (eq 8). In the first
scenario, the NC hole and electron will recombine and no net
electrons or atoms are lost. However, in the second case, a NC
hole is left unpaired, and this will eventually lead to degradation
(eq 4). Most evidence presented by us points toward this
(37) We found similar results with 72 Å NC covered with ligand 2. Thus,
coumarin was efficiently formed from these NC using 620 nm light.
However, the tendency for the thiol ligands to detach upon dissolution was
dramatically higher for these NC than for the 31 Å NC. Therefore, 72 Å
NC are of limited use and are not explicitedly discussed in this paper.
(38) Wang, L. G.; Pennycook, S. J.; Pantelides, S. T. Phys. ReV. Lett. 2002, 89,
075506/1.
(39) Laferrie`re, M.; Galian, R. E.; Maurel, V.; Scaiano, J. C. Chem. Commun.
2006, 257.
(40) Bakkers, E. P. A. M.; Marsman, A. W.; Jenneskens, L. W.; Vanmaekel-
bergh, D. Angew. Chem., Int. Ed. 2000, 39, 2297.
(41) Burda, C.; Green, T. C.; Link, S.; El-Sayed, M. A. J. Phys. Chem. B 1999,
103, 1783.
(42) (a) Doherty, A. P.; Scott, K. J. Electroanal. Chem. 1998, 442, 35. (b)
Mabon, G.; Simonet, J. Electrochim. Acta 1992, 37, 2467.
(43) Platinum electrodes as working and counter electrode; silver electrode as
reference electrode; nBu4N+PF6- as supporting electrolyte. Platinum itself
was shown not to induce isomerization of the cinnamate by coordination
chemistry.
(44) Interestingly, electrolysis of 8 at +0.9 V was accompanied by vast
decomposition of the radical cation, thereby excluding any mechanisms
that involve the radical cation in explaining the coumarin formation from
the NC.
(45) At wavelengths < 440 nm, the same proposed mechanisms will hold true
but now isomerization can also occur by direct absorption of the light by
the surface cinnamates. The latter pathway was predictable and not of high
interest.
9
J. AM. CHEM. SOC. VOL. 128, NO. 35, 2006 11725

A R T I C L E S
Wijtmans et al.
involvement of electron transfer in the cinnamate isomerization
step. However, as of yet, it cannot be fully excluded that transfer
of excited energy between NC and ligand (eq 9) was occurring
in some form. This leaves the cinnamate ligand in the excited
state, at which point it can isomerize as well (eq 10). A strong
argument against energy transfer from the NC* to the ligand is
the fact that this involves an energetically uphill pathway,
although nonvertical energy transfer directly to the triplet
cinnamate state may significantly reduce the energetic require-
ments for such a transfer.46 We also considered a 2-photon
mechanism, but this was deemed unlikely due to insufficient
power of the light source to achieve 2-photon excitation.47
Irrespective of the exact isomerization mechanism, nonphotolytic
lactonization of any Z-cinnamate can take place at any time.
The latter step will produce coumarin 1 which migrates into
solution, and hydroxyundecyl ligands are left on the NC (as in
Type VI NC).
released coumarin from ligands on a CdSe nanocrystal carrier
without decomposition of the NC core. That is, the organic-
chemical cascade is initiated simply by exciting the NC.
The presented findings further progress the fundamental
understanding of nanocrystal-ligand composites by revealing
dynamic core-ligand interactions that may be invisible on many
occasions. Moreover, the photorelease concept itself could prove
useful in a variety of areas, such as drug delivery and
bio-imaging. However, some critical issues need to be addressed
first: most importantly, the inherent (in)stability of the NC. First,
the photostability of the cinnamate-covered NC in the presence
of O2, a given parameter in biological media, has to be enhanced,
which will likely be a significant challenge. Major advances in
this area have been made recently.12 Second, the equilibrium
between bound and unbound ligands upon dissolution will be
unacceptable for biological applications. Ligand headgroups that
bind tighter could overcome this hurdle. Last, the rate of
lactonization has to be improved substantially. This problem
might solve itself when biological aqueous systems are used.48
Conclusion & Outlook
We have synthesized and characterized 31 Å CdSe nanoc-
rystals with E-cinnamate ligands bound to the NC surface by
thiol groups. Upon dissolution of these NC in organic solvents,
some surface thiols are released into solution. The NCs have
dramatically increased aerobic photoinstability as compared to
NC without cinnamate ligands. We propose this is because the
surface cinnamate facilitates electron transfer from excited NC*
to O2 in solution by acting as an electron carrier. Even in the
absence of air, this reversible electron transfer between NC*
and cinnamate occurs. While carrying the electron in its radical
anion form, the double bond of the cinnamate unit partially
rotates, giving small amounts of the Z-isomer. This Z-isomer
undergoes lactonization and releases coumarin into solution.
Thus, using visible light and anaerobic conditions, we efficiently
Acknowledgment. We thank Prof. David Cliffel and Dr. Sven
Eklund for help with the electrochemistry experiments. The
advice of Prof. David Piston on the 2-photon excitation is greatly
acknowledged. Funding from NSF 4-20-430-3381 is greatly
appreciated.
Supporting Information Available: General materials and
methods; procedures for ligand exchange, photolyses, and
electrochemistry; synthetic scheme and procedures; GC- and
HPLC-retention times; selected NMR spectra; RBS spectrum
of Type V NC; UV-visible absorption spectrum of filtrate and
NC after Type II NC 560 nm photolysis; plots for CV and bulk
electrolysis; additional time-evolved photolysis graphs. This
material is available free of charge via the Internet at
http://pubs.acs.org.
(46) Nonvertical energy transfer is well-known for stilbenes: Brennan, C. M.;
Caldwell, R. A.; Elbert, J. E.; Unett, D. J. J. Am. Chem. Soc. 1994, 116,
3460.
(47) Larson, D. R.; Zipfel, W. R.; Williams, R. M.; Clark, S. W.; Bruchez, M.
P.; Wise, F. W.; Webb, W. W. Science 2003, 300, 1434.
JA063562C
(48) The rate of coumarin formation from cinnamates in solution can be increased
as much as 1000-10 000-fold by simply switching from DMF to water or
ethanol instead. This is a result of acid-catalysis of the lactonization step.
9
11726 J. AM. CHEM. SOC. VOL. 128, NO. 35, 2006