Support curvature and conformational freedom control chemical reactivity of immobilized species

Article abstract of DOI:10.1021/ja411573a

We show that bimolecular reactions between species confined to the surfaces of nanoparticles can be manipulated by the nature of the linker, as well as by the curvature of the underlying particles.

Full text of DOI:10.1021/ja411573a

Communication
pubs.acs.org/JACS
Support Curvature and Conformational Freedom Control Chemical
Reactivity of Immobilized Species
Tino Zdobinsky, Pradipta Sankar Maiti, and Rafal Klajn*
Department of Organic Chemistry, Weizmann Institute of Science, Rehovot 76100, Israel
S
* Supporting Information
ABSTRACT: We show that bimolecular reactions
between species confined to the surfaces of nanoparticles
can be manipulated by the nature of the linker, as well as
by the curvature of the underlying particles.
or many decades, enzymes have impressed scientists with
Fthe elegance with which they control chemical reactions.
With their abilities to activate, preorganize, and increase local
concentrations of substrate molecules, they have inspired
chemists to design novel systems in which improved reactivities
emerge.1 In one example, Liu and co-workers induced dramatic
acceleration of bimolecular reactions between substrates by
preorganizing them on nucleic acid templates.2 Other templates
that were used to bring molecules together include self-
assembled cages3 as well as surfaces of inorganic materials.4
However, increasing effective molarity alone is not enough to
induce efficient bimolecular reactions; as we demonstrate in
this study, lack of sufficient conformational freedom can
drastically suppress chemical reactivity despite greatly increased
local concentrations of the reactive moieties. Further, we show
that reactivity of surface-confined species can be controlled by
tailoring the flexibility of the linker chains, as well as by
manipulating the curvature of the underlying nanoparticle (NP)
surface.
Ethynylanthracenes5−7 are attractive substrates for studying
the effect of molecular confinement on chemical reactivity. In
the absence of the CC triple bond, anthracenes undergo the
well-known8−11 [4+4] dimerization when exposed to long-wave
(λ ∼ 365 nm) UV irradiation. The presence of the triple bond
at the 9-position, however, alters the reaction pathway so as to
induce a selective [4+2] Diels−Alder dimerization12−14 under
the same reaction conditions. Recently, Weiss and co-workers
have demonstrated that upon binding to Au(111) surfaces,
thiol 1 dimerizes in the [4+4] fashion despite the presence of
the triple bond;14,15 that is, the “normal” behavior of
unsubstituted anthracenes is restored. These researchers argued
that the planar surface acts as a template favoring molecular
arrangements ideal for the [4+4], but not for the [4+2],
cycloaddition. Inspired by these findings and motivated by the
many advantages of NPs over planar surfaces,16 we synthesized
thiolated 9-ethynylanthracenes 1−3 and investigated their
photoreactivities within self-assembled monolayers on metallic
nanospheres.
Figure 1. (a) Structural formula of 9-(4-mercaptophenyl-ethynyl)-
anthracene 1. (b−d) TEM images of 1-functiona-lized 2.5 nm, 5.5 and
7.5 nm Au NPs, respectively. (e) UV−vis spectra of small-molecule 1
(in the form of a thioacetate; pure 1 is highly unstable), and the NPs
shown in (b). The broad band at ∼550 nm is due to surface plasmon
resonance. (f) UV−vis spectra of 1 released from 5.5 nm (dark gray)
and 7.5 nm (black) Au NPs preirradiated under the same conditions (t
= 21 h). Also shown (light gray) is a spectrum of 1 released from
nonirradiated 2.5 nm NPs. The spectra were normalized with respect
to the same initial concentration of 1.
other NPs investigated in this study were prepared via ligand
exchange reaction using preformed, dodecylamine (DDA)-
capped NPs20 and excess of free thiol, followed by
Our initial experiments were based on 2.5 nm gold NPs
functionalized with the previously reported14,17−19 9-(4-
mercaptophenylethynyl)anthracene (Figure 1b). These and
Received: November 13, 2013
Published: December 9, 2013
© 2013 American Chemical Society
2711
dx.doi.org/10.1021/ja411573a | J. Am. Chem. Soc. 2014, 136, 2711−2714

Journal of the American Chemical Society
Communication
precipitation, thorough washing to remove any nonadsorbed
thiols, and finally redispersion in pure toluene. As shown in
Figure 1e, binding to Au NPs alters the optical properties of 1,
red-shifting and significantly broadening the characteristic
absorption band at ∼400 nm. We then exposed the samples
to λ ∼ 365 nm UV light at ∼0.7 mW/cm2, an intensity
sufficient to induce both [4+2] and [4+4] dimerization
reactions14,21 while not affecting the stability of thiolate
monolayers22 on metallic NPs under investigation. Both
dimerization reactions can conveniently be followed by UV−
vis spectroscopy: the Diels−Alder reaction results in the blue-
shift of the anthracene absorption bands,12 whereas the [4+4]
dimerization manifests itself by a practically complete
disappearance of absorption above ∼300 nm.23 While
unadsorbed 1·Ac in solution underwent facile [4+2] Diels−
Alder dimerization upon UV irradiation (compare Figures S1
and S2 in the Supporting Information), no changes in the UV−
vis spectra of NP-anchored 1 were observed even after
continuous exposure of the samples to 24 h of UV light
under the same illumination conditions (I ∼ 0.7 mW/cm2).
This result suggests that the large curvature of the 2.5 nm NPs
prohibits two anthracene moieties from assuming orientation
required for dimerization to occur. Interestingly, our observa-
tions are in agreement with the recent findings that the [4+4]
reaction on rough Au surfaces is greatly suppressed as
compared to the atomically flat ones.21
To confirm the important role of the substrate curvature, we
also prepared and functionalized larger, ∼5.5 and ∼7.5 nm Au
NPs (1·Au-5.5 nm and 1·Au-7.5 nm in Figure 1c,d).24
Unfortunately, we found that once precipitated (during the
purification procedure), these larger particles were insoluble in
toluene or other solvents, presumably due to poor solvation of
the π−π-stacked ligands. However, the on-surface reactivity
could still be investigated; we UV-irradiated the stable solutions
of these NPs in the presence of free 1 and, following selective
removal and washing of NPs, released the surface-bound
molecules using an optimized cyanide etching protocol25 (see
Supporting Information, page S9).26 UV−vis spectra of thus
released 1 are shown in Figure 1f, whereby the decreased
absorption in the anthracene region is an indication of the
[4+4] dimerization; by comparing peak intensities, we conclude
that the yield of the reaction is ∼9% and ∼26% on 5.5 and 7.5
nm NPs, respectively.27
In addition to decreasing NP curvature, the [4+4]
dimerization might be encouraged by providing the anthracene
groups with extra conformational freedom, for example, by
incorporating alkylene bridges. To test this hypothesis, we
synthesized 9-ethynylacetylenes 2 and 3 (equipped with
additional −(CH2)2− and −(CH2)5− linkers, respectively;
Figure 2a). Again, gold particles decorated with these ligands
could not be solubilized: as soon as 2 and 3 were added to
solutions of differently sized DDA-protected Au NPs, rapid
precipitation commenced. However, we found, somewhat
surprisingly, that palladium NPs with diameters nearly identical
to those of the smallest Au NPs were readily soluble in toluene
when protected with 1, 2, 3, or any combinations thereof (cf.
Figure 2b for monodisperse 2.6 nm Pd NPs coated with 1).
Moreover, the absence of surface plasmon resonance in
palladium made it possible to decouple the optical response
of the ligands from that of Pd, and to quantify the coverage of
the particles with 1−3 (Figure 2c−f and page S10 of the
Supporting Information). The results are rather unexpected:
while both 2 and 3 formed densely packed monolayers on Pd
Figure 2. (a) Structural formulas of ethynylanthracenes 2 and 3. (b)
TEM image of 1-functionalized 2.6 nm Pd NPs. (c) UV−vis spectra of
Pd NPs functionalized with 1, 2, 3, and dodecanethiol (DDT). (d)
Schematic representation of a Pd NP covered with 9 molecules of
ligand 1. The particle surface is likely cofunctionalized with residual
DDA molecules. (e) Schematic representation of a Pd NP covered
with 58 molecules of ligand 2 and (below) no changes in the UV−vis
spectra of these particles upon UV irradiation.
(57.9 and 57.8 molecules per NP; corresponding to ∼0.364
nm2 per molecule; compare to ∼0.214 nm2 occupied by a single
alkyl thiolate ligand within densely packed monolayers on
planar gold), as little as 8.6 molecules of 1 could be
accommodated on a single Pd particle, corresponding to an
average surface area of ∼2.51 nm2 taken by each molecule. On
the basis of these numbers, it is reasonable to propose two
different binding modes of these thiols to Pd NPs: 2 and 3 are
oriented normal to the particle surface, whereas 1 maximizes
contact with the surface,28,29 as schematically presented in
Figure 2d−f. We hypothesize that this transition is due to
lateral van der Waals interactions between alkylene chains in 2
and 3, since no such interactions are possible in 1.
2712
dx.doi.org/10.1021/ja411573a | J. Am. Chem. Soc. 2014, 136, 2711−2714

Journal of the American Chemical Society
Communication
Among all Pd NPs subjected to UV irradiation, only 3·Pd-2.6
nm showed changes in the UV−vis spectra (Figure 2f, bottom),
where the intensity of the anthracene bands decreased by ∼21%
after 18 h of UV light. Exposure to an additional 20 h of UV did
not result in any further decrease of the signal, indicating that
the number of anthracene units able to assume orientation
necessary for the dimerization to take place does not increase
with time (at least within our time scales), e.g., as a result of
ligand migration.
Finally, we considered the possibility of an on-nanoparticle
Diels−Alder reaction between the CC triple bond dienophile
of a shorter ligand (1 or 2), and the anthracene diene of a
longer ligand (2 or 3) (Figure 3e). Toward this end, we
low for the product (isolated anthracene)) to that at λ = 374
nm (low for the substrate; high for the product; compare with
Figure S3 in the Supporting Information). As Figure 3d shows,
no indication of the [4+2] cycloaddition between NP-bound
ligands was found, even in (2+3)·Pd, where both ligands are
expected to be oriented perpendicular to the surface as
schematically presented in Figure 3e. On the other hand,
both types of NPs functionalized with mixed monolayers
containing 3 showed a decrease of the overall anthracene
absorbance, suggesting that the most flexible 3 can dimerize in
the [4+4] fashion even in the presence of additional, unreactive
ligands 1 and 2 on the same NPs (based on the UV−vis
spectra, we estimate that the conversion was ∼16% and ∼9%
for (1+3)·Pd and (2+3)·Pd, respectively).
In sum, we investigated chemical reactivity of arylethynylan-
thracenes on the surfaces of metallic nanoparticles. We found
that (i) although no [4+4] dimerization is observed on 2.5 nm
NPs functionalized with the rigid 1, the reaction can be induced
by decreasing the NP curvature; (ii) the mode of binding of 1−
3 to Pd NPs is critically dependent on the presence of the
alkylene linker; (iii) the [4+4] reaction can also be induced by
introducing a flexible (and appropriately long) linker between
the anthracene moiety and the particle surface; and (iv) the
Diels−Alder reaction was blocked on the surfaces of NPs,
highlighting the importance of substrate preorganization in
bimolecular reactions. We believe these results contribute to
our understanding of chemical reactivity within confined
environments, and are of interest in the context of the
development of new sensing and catalytic systems.
ASSOCIATED CONTENT
* Supporting Information
■
S
Synthesis and characterization of 1−3; preparation, function-
alization and characterization of Au and Pd nanoparticles. This
material is available free of charge via the Internet at http://
pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
rafal.klajn@weizmann.ac.il
■
Funding
This work was supported by the Israel Science Foundation
(Grant No. 1463/11).
Notes
The authors declare no competing financial interest.
REFERENCES
■
(1) Meeuwissen, J.; Reek, J. N. H. Nat. Chem. 2010, 2, 615−621.
(2) Li, X. Y.; Liu, D. R. Angew. Chem., Int. Ed. 2004, 43, 4848−4870.
(3) Yoshizawa, M.; Klosterman, J. K.; Fujita, M. Angew. Chem., Int.
Ed. 2009, 48, 3418−3438.
(4) Fruchtel, J. S.; Jung, B. Angew. Chem., Int. Ed. 1996, 35, 17−42.
(5) Godinez, C. E.; Zepeda, G.; Garcia-Garibay, M. A. J. Am. Chem.
Soc. 2002, 124, 4701−4707.
(6) Goldsmith, R. H.; Vura-Weis, J.; Scott, A. M.; Borkar, S.; Sen, A.;
Ratner, M. A.; Wasielewski, M. R. J. Am. Chem. Soc. 2008, 130, 7659−
7669.
Figure 3. (a) Changes in the UV−vis spectra of NPs protected with
the different mixed monolayers upon UV irradiation. (d) Following
the progress of the Diels−Alder reaction in solution and on NPs. (e)
Schematic representation of the behavior of mixed monolayers.
prepared 2.6 nm Pd NPs functionalized with mixed monolayers
of 1+2 (Figure 3a), 1+3 (Figure 3b), and 2+3 (Figure 3c).
Notably, the 1:1 ratio of thiols used for surface modification
was similar to that on the resulting NPs,30 although we were
unable to accurately determine the composition of the mixtures
on NPs due to the different absorption coefficients of 1, 2 and 3
(Figures S5−S10 in the Supporting Information). The progress
of the Diels−Alder reaction can be followed using UV−vis
spectroscopy by comparing the absorbance at λ = 422 nm (high
for the substrate (anthracene coupled to the C−C triple bond);
(7) Fudickar, W.; Linker, T. J. Am. Chem. Soc. 2012, 134, 15071−
15082.
(8) Bouas-Laurent, H.; Castellan, A.; Desvergne, J. P.; Lapouyade, R.
Chem. Soc. Rev. 2000, 29, 43−55.
(9) Grimme, S.; Diedrich, C.; Korth, M. Angew. Chem., Int. Ed. 2006,
45, 625−629.
(10) Fox, M. A.; Wooten, M. D. Langmuir 1997, 13, 7099−7105.
2713
dx.doi.org/10.1021/ja411573a | J. Am. Chem. Soc. 2014, 136, 2711−2714

Journal of the American Chemical Society
Communication
(11) Bhola, R.; Payamyar, P.; Murray, D. J.; Kumar, B.; Teator, A. J.;
Schmidt, M. U.; Hammer, S. M.; Saha, A.; Sakamoto, J.; Schluter, A.
̈
D.; King, B. T. J. Am. Chem. Soc. 2013, 135, 14134−14141.
(12) Becker, H. D.; Andersson, K. J. Photochem. 1984, 26, 75−77.
(13) Yuasa, J.; Ogawa, T.; Kawai, T. Chem. Commun. 2010, 46,
3693−3695.
(14) Kim, M.; Hohman, J. N.; Cao, Y.; Houk, K. N.; Ma, H.; Jen, A.
K. Y.; Weiss, P. S. Science 2011, 331, 1312−1315.
(15) Zheng, Y. B.; Pathem, B. K.; Hohman, J. N.; Thomas, J. C.; Kim,
M.; Weiss, P. S. Adv. Mater. 2013, 25, 302−312.
(16) It is possible to probe a large number of surface-bound
molecules in a single experiment; the molecules can readily be
investigated using techniques such as NMR and UV−vis spectroscopy;
one can etch the NP templates to release the organic molecules; the
effect of curvature (i.e., differently sized nanospheres) can be
investigated; etc.
(17) Zareie, M. H.; Ma, H.; Reed, B. W.; Jen, A. K. Y.; Sarikaya, M.
Nano Lett. 2003, 3, 139−142.
(18) Dou, R. F.; Ma, X. C.; Xi, L.; Yip, H. L.; Wong, K. Y.; Lau, W.
M.; Jia, J. F.; Xue, Q. K.; Yang, W. S.; Ma, H.; Jen, A. K. Y. Langmuir
2006, 22, 3049−3056.
(19) Fang, L.; Park, J. Y.; Ma, H.; Jen, A. K. Y.; Salmeron, M.
Langmuir 2007, 23, 11522−11525.
(20) Jana, N. R.; Peng, X. G. J. Am. Chem. Soc. 2003, 125, 14280−
14281.
(21) Zheng, Y. B.; Payton, J. L.; Song, T.-B.; Pathem, B. K.; Zhao, Y.;
Ma, H.; Yang, Y.; Jensen, L.; Jen, A. K. Y.; Weiss, P. S. Nano Lett. 2012,
12, 5362−5368.
(22) Witt, D.; Klajn, R.; Barski, P.; Grzybowski, B. A. Curr. Org.
Chem. 2004, 8, 1763−1797.
(23) Wu, D. Y.; Zhang, L. P.; Wu, L. Z.; Wang, B. J.; Tung, C. H.
Tetrahedron Lett. 2002, 43, 1281−1283.
(24) Chovnik, O.; Balgley, R.; Goldman, J. R.; Klajn, R. J. Am. Chem.
Soc. 2012, 134, 19564−19567.
(25) Zhu, T.; Vasilev, K.; Kreiter, M.; Mittler, S.; Knoll, W. Langmuir
2003, 19, 9518−9525.
(26) As a control experiment, we subjected free 1 (protected as a
thioacetate) to etching conditions and verified that it did not affect the
UV−vis spectra.
(27) Notably, we find no indication (cf. Figure 3) of the Diels−Alder
reaction between the NP-bound 1 and the same molecules moving
around freely in solution.
(28) Likely by coordinating with the CC triple bond to the Pd
surface.
(29) Schramm, A.; Stroh, C.; Dossel, K.; Lukas, M.; Fischer, M.;
Schramm, F.; Fuhr, O.; von Lohneysen, H.; Mayor, M. Eur. J. Org.
Chem. 2013, 70−79.
(30) We found that the total numbers of anthracene moieties per NP
were in the following ranges: (1+2)·Pd, 29.3−31.7 ligands/NP; (1+3)·
Pd, 23.1−28.6 ligands/NP; (2+3)·Pd, 48.2−55.5 ligands/NP.
2714
dx.doi.org/10.1021/ja411573a | J. Am. Chem. Soc. 2014, 136, 2711−2714