Strictly sparse surface modification and its application for endowing nanoparticles with an exact "valency"

Article abstract of DOI:10.1039/d0cc04953k

Sparsely modified surfaces can be used as a general platform for precisely modifying small nanoparticles. However, strictly, rather than statistically, sparse surface modification remains a big challenge. Herein, we report a new and general method for str

Full text of DOI:10.1039/d0cc04953k

ChemComm
View Article Online
View Journal
COMMUNICATION
Strictly sparse surface modification and its
application for endowing nanoparticles with
Cite this:DOI: 10.1039/d0cc04953k
an exact ‘‘valency’’†
Ruiqi Yang,a Jinkang Dou,a Li Jiang
Received 20th July 2020,
Accepted 9th November 2020
ab
a
*
and Daoyong Chen
*
DOI: 10.1039/d0cc04953k
rsc.li/chemcomm
Sparsely modified surfaces can be used as a general platform for in a limited range of NPs. Recently, DNA origami was used to
precisely modifying small nanoparticles. However, strictly, rather form versatile templates for precise modification of NPs, while
than statistically, sparse surface modification remains a big chal- the low stability and insolubility in most organic solvents
lenge. Herein, we report a new and general method for strictly restrict their application for precise modification.5
sparse modification of the surface of relatively large nanoparticles.
On the other hand, the solid-phase chemistry method was
The method is analogous to planting big trees and then removing exploited in several systems for endowing small NPs with an
the big crowns, leaving the stumps on the ground; due to the large exact ‘‘valency’’.6 This approach was based on a ‘‘catch and
exclusive size of the crowns, the stumps are strictly sparsely release’’ mechanism using PS Wang resin that has a low density
distributed. As a proof of concept, strictly sparse modification of of reactive groups as the substrate: a small NP was firstly
surfaces was demonstrated by the successful preparation of anchored to a solid substrate by reacting with one of the
‘‘monovalent’’ and ‘‘divalent’’ golden nanoparticles (AuNPs) with reactive groups, and then it was cleaved off and left together
different sizes. Starting from the ‘‘monovalent’’ and ‘‘divalent’’ with one functional group derived from the reactive group.7 For
AuNPs, AuNP dimers and chain-like AuNP assemblies were pre- the success of the method, the distances between any two of the
pared, respectively.
reactive groups must be well controlled; only when all the
distances are larger than the size of the small NPs can it be
Nanoparticles (NPs) with an exact number of functional guaranteed that each small NP gets exactly one valency from
groups on the surface are desirable because precise reactions one ‘‘catch and release’’ cycle. It is known that the density of
among these NPs with an exact ‘‘valency’’ can lead to precise the reactive groups on Wang resin is only statistically low. In
control over the structures, properties, and functions of the other words, even if the average density is very low, there
superparticles.1 A lot of effort has been devoted to the fabrica- should be a considerable number of reactive groups between
tion of the NPs with a ‘‘valency’’.2 It was reported that using a which the distances are less than the size of the small NPs.
modifier with a relatively large size (such as dendrimer, DNA or Therefore, some NPs may react with two or even more reactive
protein) can endow the NPs with a size smaller than or groups on the substrate and thus get more than one valency
comparable to the modifier with a single valency;3 due to the from one ‘‘catch and release’’ cycle, making the ‘‘valency’’ of
large size, the modifier firstly anchored on the NP could the as-modified small NPs uncertain. Additionally, PS chains
prevent the additional anchoring. Besides, fractionation of near the surface of Wang resin are actually linear or branched,
statistical products of the NPs/DNA or NPs/protein reactions, and the reactive groups locate in different depths in the resin,
polymerization of vinyl ligands on the NP surface, and replace- making the ‘‘catch’’ reaction complicated. The method that can
ment of unstable ligands at two poles of the NPs with ripple- strictly sparsely modify particle surfaces is significant because
like patches was reported for the preparation of ‘‘monovalent’’ it not only represents a challenge in the field of surface
or ‘‘divalent’’ NPs.4 However, these methods were effective only modification but also leads to a general platform for endowing
NPs with an exact ‘‘valency’’.
a The State Key Laboratory of Molecular Engineering of Polymers and Department of
Macromolecular Science, Fudan University, 2005 Songhu Road, Shanghai 200438,
China. E-mail: jiangli@cczu.edu.cn, chendy@fudan.edu.cn
b School of Materials Science and Engineering, Changzhou University, Changzhou
213164, China
Herein, we report strictly sparse surface modification of
silica particles sized close to 100 nm, and their use as a general
platform to endow NPs sized several nanometers with an exact
‘‘valency’’. The strictly sparse surface modification was realized
by grafting a dendrimer-like polymer (DLP) onto the surface
of the silica particles. As depicted in Scheme 1a, DLP was
† Electronic supplementary information (ESI) available. See DOI: 10.1039/
d0cc04953k
Chem. Commun.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Communication
ChemComm
Fig. 1 TEM images of silica particles before (a) and after (b and c) the DLP-
grafting; the DLP grafted silica particles were sufficiently purified to
remove the polymers that are free from the grafting. TEM images of the
silica particles with the surface being strictly sparsely modified with thiol
groups (d), and those tethered with AuNPs (e and f).
weight of B13.5 kg molÀ1) with the polydispersity index of
1.27 (Fig. S6 in the ESI†). The short PEG linker has a Mn of
2 kg molÀ1. For the grafting, DLP was mixed in DMF with the
precursor silica particles modified with coumarin groups on
the surface (S2 in the ESI†); the average size of the silica
particles is 77 nm (Fig. 1a). The grafting of DLP resulted from
photo-dimerization reaction between the coumarin group at
the end of the PEG linker of DLP and the coumarin group on
the surface of the precursor silica particle under UV irradiation
Scheme 1 Schematic illustration of the strictly sparse surface modifica-
tion of silica particles and their use as the template for the preparation of
‘‘monovalent’’ and ‘‘divalent’’ NPs. (a) Simplified chemical structure of the
DLP; (b) a silica particle with the surface being modified by coumarin; (c) a
silica particle with the surface being sparsely grafted by DLP molecules; (d)
a silica particle with the surface being strictly sparsely modified by thiol
groups (SH); (e and g) a silica particle tethering sparse NPs (e) or ‘‘mono-
valent’’ NPs (g) on the surface; (f) the ‘‘monovalent’’ and ‘‘divalent’’ NPs.
composed of 4 long PS arms (PS4) and a short PEG linker: one at 365 nm for 1 h (step i in Scheme 1). After the reaction, the
end of the linker is connected at the focal point of PS4 via a unreacted DLP was completely removed from the mixture
disulfide bond, and the other end of the linker is capped by a solution by five cycles of centrifugation, collection of the silica
coumarin group. The grafting was conducted by a dimerization particles, and redispersion in DMF. The purified DLP-grafted
reaction between the coumarin group at the end of the PEG silica particles were collected and characterized by FTIR
linker and the coumarin group modified on the surface of the (Fig. S11a in the ESI†). Compared with the FTIR spectrum of
silica particle (step i in Scheme 1). Due to the steric repulsion, the precursor silica particles (brown line in Fig. S11a in the
the grafted DLP molecule prohibits subsequent grafting of ESI†), the spectrum of the DLP-grafted silica particles shows
other DLP molecules onto the area close to the grafted site. two new peaks at 3020 and 1490 cmÀ1 (pink line in Fig. S11a in
Since the DLP has a large size, the distance between any two the ESI†), which are assigned to benzene groups of the PS4 in
neighboring grafting points is large; the strictly sparse surface DLP. The FTIR result indicates successful grafting of DLP onto
modification was thus realized, which was clearly confirmed by the silica particle. Also, the purified DLP-grafted silica particles
TEM observations (described below). Imaginably, when one were characterized by thermal gravimetric analysis (TGA). The
disulfide bond between the PS4 and the PEG linker was cleaved TGA curve of the precursor silica particles exhibits a weight loss
(Scheme 1d), exactly one thiol group (SH) was modified onto of 14.9% (brown curve in Fig. S11b in the ESI†), while that of
one grafting point on the surface of the silica particle. Thus, all the purified DLP-grafted silica particles (pink curve in Fig. S11b
the distances between two neighboring SH groups cannot be in the ESI†) shows a weight loss of 17.8%. The additional
remarkably less than the exclusive size of DLP; strictly sparse weight loss is attributed to the grafted DLP, and the weight
surface modification of silica particles by the reactive SH percentage of the grafted DLP was calculated to be 3.4 wt%.
groups was realized. Using such strictly sparsely modified silica Accordingly, the average grafting density of DLP on the surface
particles as the template, ‘‘monovalent’’ and ‘‘divalent’’ AuNPs was calculated to be one DLP chain per 106 nm2 based on the
were prepared (Scheme 1f), respectively, and the valences were method we reported previously (S3 in the ESI†).27 As exhibited
confirmed by the respective dimerization reaction and in Fig. 1c, in a TEM image with a large magnification of the
polymerization.
purified DLP-grafted silica particles stained by RuO4, black dots
The DLP was synthesized by the process described in S1 in that are individually and sparsely dispersed on the surface were
the ESI.† The molecular weight (Mn) of DLP was determined by observed. The average size of the dots is about 7 nm, being
NMR (Mn.NMR) to be 58 kg molÀ1 (each PS arm has a molecular close to the average size of individual DLP observed by TEM
Chem. Commun.
This journal is © The Royal Society of Chemistry 2020

View Article Online
ChemComm
Communication
(Fig. S8 in the ESI†) and the size estimated based on the for 24 h under gentle shaking (step iii in Scheme 1). After the
molecular weight of DLP (assuming the density of the polymer reaction, the silica particles were completely separated from the
is 1.0 g cmÀ3). Therefore, in the TEM images of the purified unreacted AuNPs-1 by five cycles of centrifugation at a speed
DLP-modified silica, each of the dots is one DLP molecule; letting the silica particles precipitate only, collection and then
the DLP molecules were sparsely grafted on the surface based re-dispersion of the precipitates in toluene under ultrasound.
on the TEM observations. The sparse modification is consistent The purified AuNP-modified silica particles were characterized
with the fact that each DLP occupies 106 nm2 (TGA result) on by TEM. As exhibited in Fig. 1e and f, AuNPs-1 were sparsely
the surface, much larger than the size in the dried state (3.14 Â tethered on the surface of the silica particles; in the control
(7/2)2 = 38 nm2; TEM result). Obviously, because the grafting experiment, the precursor silica particles without being mod-
was conducted in the good solvent of DLP, the exclusive size of ified with SH cannot catch AuNPs. Since AuNPs-1 are remark-
DLP (106 nm2), which is much larger than the size in the dried ably smaller than the distance between any two neighboring SH
state (38 nm2), determined the grafting density.
groups, each AuNP-1 can react with only one SH.
Cleavage of the disulfide bond in the structure of DLP
Taking one AuNP-1 tethered on the surface into account
grafted on the silica particles (step ii in Scheme 1) was con- (Scheme 1e), it was tethered by one SH group at one end of the
ducted by adding excess reductant dithiothreitol into the short PEG linker; the other end of the PEG linker was tethered
suspension (Fig. S4 and S12 in the ESI†). Cleavage of the on the surface of the silica particle by a coumarin dimer formed
disulfide bonds was evidenced by the disappearance of the PS during the DLP grafting. Under 254 nm UV irradiation, the
signals in the FTIR spectrum of the silica particles measured coumarin dimer dissociated into two coumarin groups, of
after the cleavage (navy line in Fig. S11a in the ESI†); the which one remains on the surface and the other caps the end
cleavage led to the release of PS4 from the silica particles of the PEG linker and leaves together with the AuNP-1; the PEG
(Scheme 1d). The weight loss of the silica particles after the short linker connects a coumarin group at its one end and an
cleavage measured by TGA is 15.1%, close to that of the AuNP-1 at its other end. From another perspective, the cou-
precursor silica particles (14.9%; navy curve in Fig. S11b in marin group is connected to the AuNP-1 by the short PEG linker;
the ESI†), supporting the almost complete release of PS4 as well the AuNP-1 functionalized with exactly one coumarin group was
as cleavage of the disulfide bonds (496%). Besides, in the TEM obtained (Scheme 1f). After separation from the silica particles
images of the silica particles after the PS4 release (Fig. 1d), no by centrifugation, pure mono-functionalized (‘‘monovalent’’)
small dots were observed even when the particles are suffi- AuNPs-1 were prepared (Fig. S5 and S14 in the ESI†).
ciently stained by RuO4, further confirming the almost com-
plete release of PS4 from the silica particles and cleavage of the tion of the AuNPs-1 through photo-dimerization of the cou-
disulfide bonds. marin groups connected respectively to AuNPs-1 (Fig. 2a). For
The ‘‘monovalency’’ of AuNPs-1 was confirmed by dimeriza-
As illustrated in Scheme 1d, taking one disulfide bond (S–S) the dimerization, ‘‘monovalent’’ AuNPs-1 were dispersed at the
into account, one of the S atoms is connected to the focus point concentration of 0.05 mg mLÀ1 in toluene, and the suspension
of PS4 and the other S atom to the silica surface by the short was irradiated by 365 nm UV for 1 hour. Before the UV
PEG linker. After cleavage of the S–S bond, two SH groups were irradiation, ‘‘monovalent’’ AuNPs-1 were individually dispersed
produced, of which one was removed together with PS4, and the
other SH group is tethered and exposed on the surface of the
silica particle. As mentioned above, DLP molecules were indi-
vidually and sparsely distributed on the silica surface (Fig. 1c).
Therefore, it is concluded that the SH groups were individually
and sparsely distributed on the surface of the silica particles
because each DLP contains only one S–S bond at the focus
point of PS4. In principle, due to the strong repulsion from the
previously grafted DLP, the subsequent grafting of other DLP
molecules onto the area close to the grafted site is prohibited.
Therefore, the distances between any two neighboring grafted
sites cannot be remarkably less than the exclusive size of the
DLP, leading to the strictly sparse modification of the SH.
It is imaginable that a golden NP (AuNP) with a size
remarkably smaller than the exclusive size of DLP can reach
only one SH on the surface (Scheme 1e). In the present study,
AuNPs with an average diameter of 3 nm (AuNPs-1) and 5 nm
(AuNPs-2) were synthesized and used, respectively (Fig. S5 and
S13 in the ESI†).8 Firstly, AuNPs-1 were used. The AuNPs
reacted with the SH groups sparsely modified on the surface
Fig. 2 (a) Schematic illustration of dimerization of ‘‘monovalent’’ AuNPs-1
by mixing the AuNPs with the silica particles in DMF/toluene
by photo-dimerization of coumarin groups. (b) TEM image of AuNP-1
(2/1, v/v) mixture solvent and then keeping the mixture at 35 1C dimers.
Chem. Commun.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Communication
ChemComm
In summary, silica particles with the surface being strictly
(rather than statistically) sparsely modified with reactive groups
were produced. The distribution of the reactive groups is
strictly sparse, while it is only statistically sparse in the cases
of the existing solid substrates. The strictly sparse modification
represents a big challenge in the field of surface modification.
Besides, the strict surface modification can endow small NPs
with not only a precise ‘‘valency’’ but also other tailor-made
properties, which should be desirable for constructing the com-
plexly while precisely structured and functionalized materials.
We are grateful for the financial support from the National
Natural Science Foundation of China (21801046 and 21871057)
and MOST (2016YFA0203302).
Fig. 3 (a) Schematic illustration of polymerization of ‘‘divalent’’ AuNPs-2
by photo-dimerization of coumarin groups. (b) TEM image of AuNP-2
chains. (c) Stacked UV-vis absorption spectra of respective samples of
AuNPs-2.
Conflicts of interest
There are no conflicts to declare.
(Fig. S14 in the ESI†). After the UV irradiation, they are largely
dimerized (Fig. 2b). According to unbiased TEM observations,
the number ratio of the AuNP-1 unimer to the AuNP-1 dimer
was about 1/2; the conversion from ‘‘monovalent’’ AuNPs-1 to
the dimer is as high as 80%. Besides, no trimers or clusters of the
AuNPs-1 were observed. The dimerization of two monovalent
AuNPs-1 resulted from the photo-dimerization of the two coumarin
Notes and references
1 (a) S. C. Glotzer, Science, 2004, 306, 419–420; (b) C. Yi, Y. Yang and
Z. Nie, J. Am. Chem. Soc., 2019, 141, 7917–7925; (c) Z. Tang, Z. Zhang,
Y. Wang, S. C. Glotzer and N. A. Kotov, Science, 2006, 314, 274–278;
(d) R. Fernandes, M. Li, E. Dujardin, S. Mann and A. G. Kanaras,
Chem. Commun., 2010, 46, 7602–7604.
groups connected, respectively, to the two AuNPs-1 (Fig. 2a). 2 (a) M. A. Boles, D. Ling, T. Hyeon and D. V. Talapin, Nat. Mater., 2016,
15, 141–153; (b) S. A. Belhout, J. Y. Kim, D. T. Hinds, N. J. Owen,
J. A. Coulter and S. J. Quinn, Chem. Commun., 2016, 52, 14388–14391;
(c) X. Yao, J. Jing, F. Liang and Z. Yang, Macromolecules, 2016, 49,
Since each coumarin group was connected to an AuNP-1 by a
PEG linker, there are two PEG linkers end-to-end connecting
between the two dimerized AuNPs-1, which leads to a gap of
about 2 nm between the two AuNPs-1 (Fig. 2b).
9618–9625; (d) W. Wang, K. Zhang, Y. Bao, H. Li, X. Huang and
D. Chen, Chem. Commun., 2018, 54, 11084–11087; (e) C. Yi, H. Liu,
S. Zhang, Y. Yang, Y. Zhang, Z. Lu, E. Kumacheva and Z. Nie, Science,
2020, 369, 1369–1374.
AuNPs-2 with a larger size of 5 nm were also used, and
‘‘monovalent’’ and ‘‘divalent’’ AuNPs-2 were prepared, respec- 3 (a) J. P. Hermes, F. Sander, U. Fluch, T. Peterle, D. Thompson,
R. Urbani, T. Pfohl and M. Mayor, J. Am. Chem. Soc., 2012, 134,
14674–14677; (b) Z. Li, E. Cheng, W. Huang, T. Zhang, Z. Yang, D. Liu
and Z. Tang, J. Am. Chem. Soc., 2011, 133, 15284–15287; (c) J. Farlow,
tively. ‘‘Monovalent’’ AuNPs-2 were prepared under the same
conditions through the identical process as those for preparing
monovalent AuNPs-1, except that the AuNPs-1 were replaced by
AuNPs-2. Furthermore, ‘‘divalent’’ AuNPs-2 were prepared by
reaction between ‘‘monovalent’’ AuNPs-2 with the SH strictly
sparsely modified on the silica surface and then cutting the
AuNPs off the surface by dissociation of the coumarin dimers
(Scheme 1g and Scheme S5 in the ESI†). The ‘‘monovalency’’ of
the as-prepared ‘‘monovalent’’ AuNPs-2 was confirmed by their
dimerization reaction due to the photo-dimerization of the
coumarin groups (Fig. S16 in the ESI†). The ‘‘divalency’’ was
proven by their linear polymerization driven by the photo-
dimerization of the coumarin groups (Fig. 3a and b). Theore-
tically, using the strictly sparsely modified silica particles as the
substrate, AuNPs with ‘‘N-valences (N = 3 or 4)’’ can be prepared
by using AuNPs with (N À 1) ‘‘valency’’ as the precursor through
the same process. Furthermore, compared with the spectrum of
D. Seo, K. E. Broaders, M. J. Taylor, Z. J. Gartner and Y. W. Jun, Nat.
Methods, 2013, 10, 1203–1205.
4 (a) P. K. Harimech, S. R. Gerrard, A. H. El-Sagheer, T. Brown and
A. G. Kanaras, J. Am. Chem. Soc., 2015, 137, 9242–9245; (b) C. Krueger,
S. Agarwal and A. Greiner, J. Am. Chem. Soc., 2008, 130, 2710–2711;
(c) G. A. DeVries, M. Brunnbauer, Y. Hu, A. M. Jackson, B. Long,
B. T. Neltner, O. Uzun, B. H. Wunsch and F. Stellacci, Science, 2007,
315, 358–361; (d) K. L. Gurunatha, A. C. Fournier, A. Urvoas,
M. Valerio-Lepiniec, V. Marchi, P. Minard and E. Dujardin, ACS Nano,
2016, 10, 3176–3185.
5 (a) T. G. Edwardson, K. L. Lau, D. Bousmail, C. J. Serpell and
H. F. Sleiman, Nat. Chem., 2016, 8, 162–170; (b) C. R. Laramy,
M. N. O’Brien and C. A. Mirkin, Nat. Rev. Mater., 2019, 4, 201–224;
(c) W. Liu, J. Halverson, Y. Tian, A. V. Tkachenko and O. Gang, Nat.
Chem., 2016, 8, 867–873; (d) J. K. N. Mbindyo, B. D. Reiss,
B. R. Martin, C. D. Keating, M. J. Natan and T. E. Mallouk, Adv.
Mater., 2001, 13, 249–254.
6 (a) K. M. Sung, D. W. Mosley, B. R. Peelle, S. G. Zhang and J. M.
Jacobson, J. Am. Chem. Soc., 2004, 126, 5064–5065; (b) J. G. Worden,
A. W. Shaffer and Q. Huo, Chem. Commun., 2004, 518–519.
individual AuNPs-2, the UV-vis spectra (Fig. 3c) of AuNPs-2 7 (a) J. H. Kim and J. W. Kim, Langmuir, 2008, 24, 5667–5671;
(b) M. X. Xie, L. Jiang, Z. P. Xu and D. Y. Chen, Chem. Commun.,
2015, 51, 1842–1845.
8 M. Brust, M. Walker, D. Bethell, D. J. Schiffrin and R. Whyman,
dimers and the ‘‘polymer’’ have 27 nm and 60 nm red-shifts,
respectively, further supporting the occurrence of the dimeriza-
tion and polymerization. The relative location of ligands on the
surface can be regulated by the molecular weight of the PEG
linker, offering an opportunity to control their intriguing coupling
properties for constructing desired assembly structures.9
J. Chem. Soc., Chem. Commun., 1994, 801–802.
9 (a) Y. Chen, Z. Wang, Y. W. Harn, S. Pan, Z. Li, S. Lin, J. Peng,
G. Zhang and Z. Lin, Angew. Chem., Int. Ed., 2019, 58, 11910–11917;
(b) Y. Chen, Z. Wang, Y. He, Y. J. Yoon, J. Jung, G. Zhang and Z. Lin,
Proc. Natl. Acad. Sci. U. S. A., 2018, 115, 1391–1400.
Chem. Commun.
This journal is © The Royal Society of Chemistry 2020