Controlling the Spatial Organization of Liquid Crystalline Nanoparticles by Composition of the Organic Grafting Layer

Article abstract of DOI:10.1002/chem.201406262

Understanding how the spatial ordering of liquid crystalline nanoparticles can be controlled by different factors is of great importance in the further development of their photonic applications. In this paper, we report a new key parameter to control the mesogenic behavior of gold nanoparticles modified by rodlike thiols. An efficient method to control the spatial arrangement of hybrid nanoparticles in a condensed state is developed by changing the composition of the mesogenic grafting layer on the surface of the nanoparticles. The composition can be tuned by different conditions of the ligand exchange reaction. The thermal and optical behavior of the mesogenic and promesogenic ligands were investigated by using differential scanning calorimetry (DSC) and hot-stage polarized optical microscopy. The chemical structure of the synthesized hybrid nanoparticles was characterized by ¹HNMR spectroscopy, thermogravimetric analysis (TGA), XPS, and elemental analysis, whereas the superstructures were examined by small-angle X-ray diffraction (SAXSRD) analysis. Structural studies showed that the organic sublayer made of mesogenic ligands is denser with an increasing the average ligand number, thereby separating the nanoparticles in the liquid crystalline phases, which changes the parameters of these phases. Softening nanocrystalline gold: An efficient method to control the spatial arrangement of liquid crystalline gold nanoparticles was developed by changing the composition of the mesogenic grafting layer. Structural studies showed that the organic sublayer is more dense with increasing average ligand number, thereby separating the nanoparticles in the liquid crystalline phases, which changes the parameters of these phases (see figure).

Full text of DOI:10.1002/chem.201406262

DOI: 10.1002/chem.201406262
Full Paper
&
Gold Nanoparticles
Controlling the Spatial Organization of Liquid Crystalline
Nanoparticles by Composition of the Organic Grafting Layer
Michał M. Wójcik,* Magdalena Olesi n´ ska, Michał Sawczyk, Józef Mieczkowski, and
[a]
Ewa Górecka*
Abstract: Understanding how the spatial ordering of liquid
crystalline nanoparticles can be controlled by different fac-
tors is of great importance in the further development of
their photonic applications. In this paper, we report a new
key parameter to control the mesogenic behavior of gold
nanoparticles modified by rodlike thiols. An efficient method
to control the spatial arrangement of hybrid nanoparticles in
a condensed state is developed by changing the composi-
tion of the mesogenic grafting layer on the surface of the
nanoparticles. The composition can be tuned by different
conditions of the ligand exchange reaction. The thermal and
optical behavior of the mesogenic and promesogenic li-
gands were investigated by using differential scanning calo-
rimetry (DSC) and hot-stage polarized optical microscopy.
The chemical structure of the synthesized hybrid nanoparti-
1
cles was characterized by H NMR spectroscopy, thermogra-
vimetric analysis (TGA), XPS, and elemental analysis, whereas
the superstructures were examined by small-angle X-ray dif-
fraction (SAXSRD) analysis. Structural studies showed that
the organic sublayer made of mesogenic ligands is denser
with an increasing the average ligand number, thereby sepa-
rating the nanoparticles in the liquid crystalline phases,
which changes the parameters of these phases.
[
11]
Introduction
spacer linking the ligand and metal core, it is possible to
induce the formation of nematic, smectic, and columnar struc-
tures comprising hybrid nanoparticles in which the metal cen-
ters and the organic parts self-segregate in space. In previous
studies we also described dynamic aggregates with thermal
control over the superoganization of NPs (over the superlattice
Liquid crystals (LCs) combine the properties of solids and
[
1]
fluids, thus they can easily be controlled by external stimuli,
which makes them ideal for technological applications.
[2]
Throughout the past few years liquid crystals have appeared
as promising candidates for the construction of inorganic–or-
ganic functional nanomaterials, for example, they allow effi-
[12]
structure of the NPs). It should be noted that the influence
of the ratio between the akyl ligands and the LC ligands on
the formation of columnar or 3D lattices was reported by Mehl
[3]
cient tuning of the bandgap properties. Recently, several sci-
entific groups have dealt with dispersing inorganic nanoparti-
cles within a liquid crystalline matrix, which is generally an at-
tractive method towards improving the liquid crystalline prop-
erties of a material, for example, switching the behavior by
[13]
et al. In contrast to this report, we keep the alkane co-ligand
length and the nanoparticle diameter fixed in all samples. This
approach should enable us to examine the impact of a single
factor (i.e., the ratio of mesogen to alkane ligands) on LC-
phase formation by hybrid nanoparticles.
[
4]
means of electro-optical applications and fabrication of mate-
[
3b,5]
rials with novel properties.
Such materials with unique
Most of the up-to-date reported liquid crystalline NPs are
made from gold clusters covered by alkyl thiols in which a part
of the primary grafting layer is substituted by more complex,
(pro)mesogenic, ligands by using a ligand-exchange reaction.
properties can be achieved by direct coating of nanoparticles
[6]
with liquid crystalline ligands. Organic coating of gold , tita-
[7]
nia, and g-Fe O magnetic nanoparticles was used to induce
2
3
[14,15]
complex spatial arrangements of nanoparticles. It was shown
that with changing the molecular architecture of the meso-
Details of the reaction were studied by Murray et al.
who
showed that the rate of the ligand exchange decreases as the
chain length and/or the steric hindrance of the primary graft-
[
8]
[9]
[10]
morphic ligands (rodlike, bent-core, and dendric ) grafted
on the surface of the nanoparticles (NPs) and the length of the
[16]
1
ing shell increases. Hutchison et al., by using H NMR spec-
troscopy, X-ray photoelectron spectroscopy (XPS), and ther-
mogravimetric analysis (TGA), developed a novel approach for
the preparation of diverse libraries of ligand-stabilized metal
nanoparticles, controlled by the reaction time and the molar
equivalents of thiol versus nanoparticles ratio. The purpose of
this paper is to demonstrate that the ligand exchange reaction
is a powerful tool to control the self-assembly of NPs. We
show that a spatial arrangement of gold nanoparticles is gov-
[
a] Dr. M. M. Wójcik, M. Olesi n´ ska, M. Sawczyk, Prof. J. Mieczkowski,
Prof. E. Górecka
Faculty of Chemistry, University of Warsaw
Pasteura 1, 02-093 Warsaw (Poland)
E-mail: mwojcik@chem.uw.edu.pl
gorecka@chem.uw.edu.pl
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201406262.
Chem. Eur. J. 2015, 21, 10082 – 10088
10082
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
erned by the chemical composi-
tion of the coating layer in
which only the ratio between
alkyl thiols (primary grafting)
and mesogenic thiols (secondary
grafting) was changed. This is
the first report to show the im-
portance of liquid-crystalline
ligand population on nanoparti-
cles of the same diameter and in
formation and control of nano-
particle-ordered aggregate struc-
tures. We also show that ele-
1
mental analysis H NMR spec-
troscopy and XPS can be simul-
taneously used to estimate the
number of ligands with high
precision.
Results and Discussion
In order to investigate composi-
tion–structure relationships, two
mesogenic ligands, having the
same linear mesogenic core and
branched (i.e., ligand II) or
kinked terminal chains (i.e.,
ligand I) were tested. These mol-
ecules have been synthesized
according to the Scheme 1a. De-
tailed information of intermedi-
ate synthesis and the two final
mesogenic ligands I and II, in-
1
cluding yields, as well as H/
Scheme 1. a) Synthesis scheme of rodlike mesogenic thiol derivatives and schematic representation of the final
13
C NMR spectroscopic and ele-
molecules I and II. Reaction conditions: i) Br
CO , 4,-4’-dihydroxybiphenol, DMF; vi) P(Ph
DIAD), 11-bromoundecan-1-ol, THF; vii) bis(trimethylsilyl)sulfide (HMDST), tetrabutylammonium fluoride (TBAF),
2
, hn; ii) SOCl
2
3
, CHCl , heating to reflux; iii) 2 equiv dioctylamine, CHCl
3
;
mental analysis (EA) data, are iv) oleyl alcohol, pyridine; v) KI, K
2
3
3 3
) , diisopropyl azodicarboxylate
(
provided in the Experimental
Section and the Supporting In-
formation. These ligands were
used to prepare two series of
hybrid nanoparticles, H1 to H5
THF, RT. b) Schematic illustration of the partial ligand-exchange reaction on the surface of gold nanoparticles and
graphical models of the hybrids H-1, H-2, and H-3.
1
Table 1. Hybrid composition and number of constituent ligands obtained from elemental analysis, H NMR
(
by using ligand I) and H6 to H7
by using ligand II).
spectroscopy, and TGA.
(
1
Sample
Elemental analysis
TGA
H NMR
The composition of the hy-
[a]
[b]
[c]
[d]
1
[f]
[
%C]
N
L-EA
D
1
D
2
N
L-TGA
Int. 3.98
ppm]
molar ratio
nL- H NMR
brids is described in Table 1. For
the samples studied Au309(LX)m(I
or II)n the relative ratio m/n of
the primary alkyl ligands LX to
the incoming mesogenic ligands
I (or II in case of hybrids H6 and
H7) was controlled by the reac-
[e]
[
n/m
H1
H2
H3
H4
H5
H6
H7
17.27
19.32
24.19
27.05
30.39
19.61
34.22
15
20
31
39
50
19
60
10.24
9.21
7.32
5.61
4.34
9.03
2.81
12.64
15.93
21.76
25.93
32.48
15.41
35.88
15
20
29
39
50
19
60
11.79
8.83
6.09
4.44
3.63
9.41
2.98
0.20
0.29
0.49
0.82
1.23
0.27
2.04
15
21
30
41
51
20
61
[
17]
tant
concentrations
(see
[a] Number of mesogenic ligands calculated from the results of the elemental analysis. [b] Weight loss [%] asso-
Table 1 in the Supporting Infor-
mation). In all studied samples
the average number of ligands
at the surface, n+m, was esti-
mated at approximately 91,
ciated with thermal desorption of primary alkyl thiols. [c] Weight loss [%] associated with thermal desorption
and decomposition of mesogenic thiols. [d] Number of incoming ligands calculated from the results of the
thermogravimetric analysis. [e] Calculated as the ratio of the area under the curve of the signal at d=3.98 ppm
(
[
assumed to be 2) to the difference in the areas under the curves of the signals at d=2.69 and 3.98 ppm.
f] Number of incoming ligands calculated from the results of the H NMR spectroscopic measurements.
1
Chem. Eur. J. 2015, 21, 10082 – 10088
www.chemeurj.org
10083
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
based on experimental measurements (Small angle X-ray dif-
fraction (SAXRD) and TEM) correlated with theoretical calcula-
tions. To correlate the liquid-crystalline properties of nanoparti-
cles to the relative population of alkane and mesogenic li-
gands it was necessary to establish the nanoparticle organic
corona composition with high precision. For this purpose, four
complementary methods were used, that is, elemental analysis,
1
H NMR spectroscopy, TGA, and XPS. We also investigated the
structural and thermal properties of ligands I and II, used as
liquid-crystalline behavior-inducing agents.
Elemental analysis was used to follow the ligand-exchange
reaction. Two factors contribute to the high sensitivity of this
method, that is, the substitution of alkane to mesogenic ligand
causes a large increase of the carbon-atom content in the
sample, and additionally, for ligand I, a new element (nitrogen)
is introduced. Based on the metal core size and the total
number of ligands, a calibration curve (Plot 1 in the Supporting
Information) showing the number of secondary ligands versus
the carbon or nitrogen content in the samples of the hybrid
materials (H1 to H5, Table 1), was calculated. A similar calibra-
tion curve (n vs. %C) was also obtained for samples H6 and
H7 modified with the ligand II (Plot 2 in the Supporting Infor-
mation). According to the calibration curve, an increase of the
number of mesogenic ligands by one changes the carbon con-
centration in the hybrid by at least 0.25%, a value that is an
order of magnitude higher than the standard deviation of the
EA method, which is 0.023%. Therefore, the average number
of mesogenic ligands can be precisely determined with the ac-
curacy of one molecule. However, the disadvantage of EA is
the need for large sample quantity.
Formation of liquid crystalline phases is very sensitive to mo-
lecular factors of the attached ligands, for example, even small
changes in the length of the terminal chains might have a cru-
cial effect on the behavior of the system. They differ in the
type, length, and branching of the terminal chains. The phase-
transition temperatures and structural changes of used organic
ligands are shown in Figure 1. Molecule II (bromide precursor,
for the molecular formula, see the Supporting Information)
forms a variety of smectic phases. The phase sequence Cr-
HexF-HexI-SmC-Iso was evidenced by X-ray studies. The lower-
ing temperature increases the layer spacing, the transition be-
tween the hexatic phases HexI-HexF is accompanied by a small
discontinuity in the layer spacing and change in the position
of higher angle signals (see the Supporting Information, Fig-
ures 1–3). The amide analogue is monotropic LC. Usually when
considering LC phases made of nanocrystals, their properties
are directly compared with mercapto-functionalized ligands.
However, it should be noted that the presence of a ÀSH
moiety in the mesogen structure often causes narrowing of
the LC-phase temperature range and simplifies polymorphism.
The same tendency can be observed for compounds II and II-
bromine. Compound II exhibits a Sm phase between 75 and
Figure 1. a) DSC thermograms for compounds I (black) and II-bromide (red,
for the molecular formula, see the Supporting Information), heating/cooling
scans. b) Microphotographs of the textures obtained in the polarized optical
microscope on planar samples of compound II-bromide, the Sm C phase (on
the right) and the Hex I phase (on the left). c) Variation of d spacing ob-
tained from the main X-ray signal as a function of temperature (sample of
compound II-bromide; for more X-ray data, see the Supporting Information
Figures 1–3).
more similar to the properties of compound II-bromide than
to the properties of compound II (thiol). Therefore, when de-
signing LC nanocrystals the mesogenic properties of the bro-
mine derivatives of the ligands should be considered to be
more informative and comparable to properties of the final
material.
To confirm the discussed methodology of ligands number
1
determination, H NMR studies were carried out by using a Var-
ianUnityPlus 200 MHz equipment, in two ways: direct measure-
ment of the hybrid nanoparticles and I -decomposed gold
2
1
nanoparticles. In comparison with H NMR spectra of alkane-
coated nanoparticles (GNP@C6, Figure 4 in the Supporting In-
formation), the spectra of the hybrid material shows broad-
ened signals in the aromatic regions, which suggests the pres-
ence of secondary ligands covalently attached to the metal
1018C, whereas for compound II-bromine a substantially wider
1
and more phase-rich diagram (based on differential scanning
calorimetry (DSC) measurements) was evidenced. As it will be
discussed later, hybrid nanoparticles based on compound II
form stable long-range-ordered structures. This behavior is
surface. Typical H NMR spectra of the hybrid material samples
H1, H3, and H4, are presented in Figure 2 and show different
areas under the curves depending on the average ligand pop-
1
ulation. Because the H NMR experiments were run using the
Chem. Eur. J. 2015, 21, 10082 – 10088
www.chemeurj.org
10084
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
The analysis of monolayer-protected nanoparticles by XPS
can provide information regarding both the nature of the
chemical bond between the adsorbed head groups and the
metal surface as well as the chemical composition of the
hybrid material. Examination of the XPS spectra films of com-
pounds H2 and H4 is of particular important for our study of
nanomaterial composition. Figures 9 and 10 in the Supporting
Information show the XPS survey spectra for the mesogen-
capped nanocrystals synthesized according to Scheme 1 and
listed in the Table 1. For both these films, bands of the binding
energy, E , for the Au, S, N, and C atoms are distinguished in
B
their survey spectra. A high-resolution scan from 390.0 to
410.0 eV revealed the nitrogen atom 1s peak at 399.56 eV.
Recall the synthetic scheme in Scheme 1a, the hybrid nanopar-
ticles H1 to H5 were formed from the mesogenic ligand I,
which introduces nitrogen atoms. In both spectra, signals for
gold (4f , 83.7 eV and 4f , 87.2 eV), carbon (1s from 282.6 to
1
Figure 2. H NMR spectra of the three representative hybrids composed of
the same primary nanoparticles and mesogenic ligand (i.e., ligand I), but
having a different degree of mesogenic substitution (H1=red line,
H3=green line, and H4=blue line).
7/2
5/2
288.3 eV, with a maximum at 284.3 eV), sulfur (2p peak at
same acquisition time and hybrid nanoparticle concentration,
the result confirms a different degree of substitution of secon-
dary ligands. Non-homogenous broadening of the signals
a binding energy of 162.1 eV along with a broad shoulder for
sample H4 or doublet at 161.8 eV and for sample H2 at
162.8 eV, see Figure 3) and oxygen (1s at 531.9 eV) were ob-
1
limits the precision of quantitative H NMR data analysis. Nev-
1
ertheless, H NMR spectroscopy is very useful to confirm the
purity of the hybrid material. The absence of sharp signals
proves complete removal of unsubstituted secondary ligands
(i.e., ligand I) in the purification procedure, and the presence
of unbound ligands might affect the structure of the liquid
1
crystalline phases made of gold nanoparticles. The H NMR sig-
nals from hybrids are significantly broadened in comparison to
those registered for a free ligand and, moreover, this broaden-
ing is different for different parts of the ligand molecules, de-
pending on the distance from the metal surface. It is well
1
known that H NMR line broadening for proteins and polymers
is caused by their slow rotation in solution and the same situa-
tion is observed for ligands immobilized at the gold surface.
NPs can be regarded as having a densely packed, solid-like or-
ganic region near to the gold surface (mainly built of the alkyl
thiol chains and the linkers of the mesogenic ligands) and rela-
tively more mobile terminal chains of mesogenic ligands.
[
18]
Figure 3. XPS characterization of H2 and H4 nanoparticles. Comparison of
the A) Au 4f, B) C 1s, C) N 1s, and D) S 2p normalized core-level spectra (HR-
XPS) recorded from a mesogen-stabilized gold nanoparticle film grown on
a Si (111) substrate. Curve 1 (green)=film of a H2 sample with lower meso-
gen content, curve 2 (magenta)=film of a H4 sample with higher mesogen
content.
Iodine-mediated cluster decomposition, which quantitatively
liberates all the attached ligands as disulfides, was used to
more precisely calculate the ligand-exchange ratios, and gave
almost identical results to the EA and TGA measurements
(
Table 1). The composition of the organic corona was calculat-
1
ed by using H NMR data of I -decomposed materials. Fig-
2
1
ures 5–8 in the Supporting Information show the H NMR spec-
tra of I -decomposed gold nanoparticles with ligand I. The n/m
ratio was estimated from the H NMR data by comparing the
served. The film was not sufficiently thick; therefore, an addi-
tional signal was measured from the substrate (Si 2p core level
at 105.4 eV). In addition, the oxygen 1s signal was influenced
by the composition of the substrate, which made it impossible
to take the oxygen signals into account in the calculation.
These peaks are labeled in the bottom survey scans (Figures 9
and 10 in the Supporting Information).
2
1
intensities of the signals coming from methyl groups at the
sulfur atoms (peak at d=2.69 ppm present in the primary alkyl
surfactant and the mesogenic ligand) to the signal coming
from methylene groups at the oxygen atom present in the
mesogenic ligand (peak at d=3.98 ppm). The results were con-
sistent with the data obtained from the EA and TGA measure-
The binding energy of the major component (Au 4f7/2
83.7 eV) is in agreement with results reported for alkane thiols
assembled on gold nanoparticles and/or bulk gold sub-
1
ments. It should be noted that direct H NMR analysis of gold
nanoparticles, without I decomposition, is not sufficient for
2
[19]
[20]
the determination of the exchange ratio with high precision.
strates as well as for bulk gold. The sulfur 2p signal was
Chem. Eur. J. 2015, 21, 10082 – 10088
www.chemeurj.org
10085 ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
fitted with two sum functions. According to previously pub-
lished data, we assign the peak at 162.1 eV (2p3/2) to sulfur
[
21]
bound to Au nanoparticles (SAu). Furthermore, we did not
observe any remaining signal at 163.7 eV (S 2p3/2), which is as-
signed to sulfur of free thiol groups (SÀH), which are not
[
22]
bound to the gold particles. In the E range of the C 1s elec-
B
trons, there are three and four different forms of carbon for
each part of the mesogenic and alkyl thiol molecules, respec-
tively. By fitting the experimental data we found bands at the
EB values of 284.3, 286.3, 287.7 eV with different relative-inten-
sity ratios. The most intense band at 284.3 eV is assigned to
carbon atoms of the alkyl group and aromatic moieties of the
mesogenic compounds. Other smaller bands presumably origi-
nate from carbon atoms of carbonyl groups and other carbon
species. The origin of the other three small bands is most likely
the same as that for typical complex organic molecules. The
energy positions found for the nitrogen 1s XPS peak for com-
pounds H2 and H4 correspond to the energy positions found
[
23]
for amide-functionalized compounds. Quantitative XPS anal-
yses of the hybrids H2 and H4 give gold-to-sulfur, gold-to-
carbon, and gold-to-nitrogen ratios as summarized in the
Table 2 in the Supporting Information, depending on the
ligand concentration on the gold surface. In all cases, these
ratios are in good agreement with tendencies obtained from
1
TGA, EA, H NMR spectroscopy, and theoretical calculations. We
can observe increasing atomic ratios between C/Au atoms and
N/Au atoms. However, direct approximation of the number of
ligands with high accuracy is not possible because quantitative
analysis of XPS requires the use of additional calculations de-
scribed in reference [10] that would take the thickness of the
organic layer into account, which would require many addi-
tional analyses. However, XPS analysis further confirmed the
absence of free thiol molecules (together with the absence of
1
sharp CH SH peaks in the H NMR spectrum). Finally, we have
2
turned to the thermogravimetric analyses; the results shown in
Figure 4 indicate that the mesogen-capped gold nanoparticles
start losing mass at about 2108C, which is approximately 508C
above the boiling point of hexanethiol. Previous TGA studies
of alkyl thiol capped gold nanoclusters show that the weight
loss occurs gradually, but more rapidly above the boiling point
Figure 4. a) TGA data recorded for samples H1 to H5. b) First derivative TGA
profile of the hybrid nanoparticle samples. The large change at low temper-
atures is due to the release of primary thiols (C H13SH). The second double
6
peak is attributed to the decomposition of mesogenic thiols. c) Hybrid nano-
particles sample to weight loss in the first step (black line) and second step
[
24]
of the ligand. The second double-peak weight loss, starting
at about 3408C, is due to mesogenic thiol detachment and de-
composition. The composites with the larger content of meso-
genic thiols not only show greater weight losses above ap-
proximately 3408C, but also have a lower remaining mass
upon complete thermolysis (plateau level above 5008C). No
further thermal events can be observed in Figures 4a and
b until the sample reaches a temperature of 5108C, leaving
nonvolatile metallic gold. The average population of alkyl
thiols and mesogenic molecules can be found by comparison
of theoretical calculations (Plot 1 and Plot 2 in the Supporting
Information) and the weight losses occurring in two distinct re-
gions from 130 to 2508C and from 250 to 5108C. When analyz-
ing samples H1 to H5, that is, with a growing mesogenic
ligand content, the weight loss in the first step decreased from
(
red line).
were recorded for sample H6 and H7 and their auxiliary
sample (Plot 3 in the Supporting Information).
The structural analyses of ordered aggregates formed by
nanoparticles were performed by using small-angle X-ray dif-
fraction (Figure 5, SAXRD 2D patterns) method. For hybrid H4
with a high exchange ratio (n/mꢀ1.0) a lamellar phase has
been found at room temperature. The structure can be recog-
nized from the characteristic X-ray pattern (see Figures 11 and
12 in the Supporting Information), for an aligned sample (ob-
tained by a rubbing technique), up to three commensurate,
sharp reflections located along the same azimuthal direction
(layer normal) related to a layered structure and a diffused
signal from the in-plane structure along the direction perpen-
10.2 to 4.3%, whereas in the second step the weight loss in-
creased from 12.6 to 32.5% (Figure 4c). The same observations
Chem. Eur. J. 2015, 21, 10082 – 10088
www.chemeurj.org
10086
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
54 Š. This indicates that, for these samples, metal cores are
separated by a single organic layer in which ligands attached
to the metal cores from neighboring layers interdigitate (sche-
matic representation in Figure 5, layer spacing d ). Evidently,
1
such interdigitation is necessary to ensure proper organic sub-
layer density. The sample also has a lower clearing point, and
the direct lamellar–isotropic phase transition in these samples
was observed at approximately 1008C. In the case of hybrids
comprising ligand II (i.e., compounds H6 and H7), either a la-
mellar structure with large layer spacing (ꢀ100 Š) is observed
for n/m>0.1 or, for a smaller number of mesogenic ligands
(
i.e., n/m<0.1), only amorphous aggregates with no long-
range structure were detected. The lamellar spacing of hybrids
H6 and H7 does not change much with an increase of the con-
centration of the mesogenic ligand. This observation indicates
that the mesogenic ligand II has very strong smectogenic abili-
ty. The X-ray data also show that the structure gradually
changes from a simple smectic phase observed for compound
H2 with a low ligand population (Figure 5a) to a more com-
plex polymorphic structure in the case of the highly loaded
sample H4 (Figure 5b). For samples H6 and H7, independent
of the ligand population and temperature (i.e., below decom-
position temperature, which occurs at ꢀ1708C), the smectic
structure is observed. For this ligand, at very low exchange
ratio, no LC phase is formed.
Figure 5. Characterization of sample H2 (a), H4 (b) and H6 (c) (top section).
Temperature evolution of SAXRD patterns evidences a significant change in
X-ray signal positions for different grafting layer composition (H2 and H4)
and nanoparticles covered with different mesogenic ligands. (bottom sec-
tion). Schematic illustration of smectic phase made of H2 (T=958C) with
layer spacing d
=82 Š.
1
=54 Š and smectic phase made of H4, with layer spacing
d
2
dicular to the layer normal were registered. The observed X-ray
pattern is typical for a lamellar phase with no long-range order
in the metal sublayer. The layer spacing is approximately 82 Š
and the mean in-plane distance is about 33 Š. The measured
layer spacing shows that the lamellar structure is constructed
from the sublayers of metal cores separated by the organic bi-
layers made of mesogenic ligands, in which ligands from each
NP form a separate organic sublayer (schematic representation
Conclusion
In summary, the above-described results indicate that the aver-
age concentration of mesogenic ligands in the grafting layer is
an important factor driving the superorganization of NPs. The
organic sublayer separating the metal cores can be built from
interdigitated ligands attached to the metal NPs from neigh-
boring layers; for a small exchange ratio, or for a larger con-
centration of the mesogenic ligands in the organic corona, it
can be of bilayer-type in which the mesogenic ligands do not
interdigitate. The type of soft layer also influences the clearing
temperature: the larger the average number of ligands at-
tached to the metal surface is, the stronger are interactions be-
tween the NPs and the higher the phase-transition tempera-
ture to the isotropic phase.
in Figure 5, layer spacing d ). To further confirm the SAXRD
2
measurements, we have performed transmission electron mi-
croscopy (TEM) analysis. Investigation of the H4 sample re-
vealed mean distances between the smectic layer made of the
Au core of about 8 nm (Figure 6), showing excellent correlation
to the pattern observed in the SAXRD experiment. The phase
sequence of this material was also established by using
SAXRD. The measured isotropization temperature was approxi-
mately 1408C, but before the isotropization event a small tem-
perature interval of columnar phase appeared. On the other
hand, for hybrid nanoparticles with low mesogenic ligands
content, structure H-2 (n/m <0.5), a different type of lamellar
phase was found, having a much lower layer spacing of about
Experimental Section
The primary gold nanoparticles (GNP@C6) were synthesized by
[8]
using a slightly modified literature procedure. Briefly, hydrogen
tetrachloroaurate (1.68 mL of 30% AuCl3 diluted with HCl; dis-
solved with 88.3 mL of distilled water) was extracted into toluene
by using the phase-transfer reagent tricaprylmethylammonium
chloride (4.84 g dissolved in 234.5 mL of toluene). The resulting or-
ganic phase was then stirred with hexanethiol (C H SH, 0.4 mL) for
6
13
1
h, until a pale yellow color appeared and the mixture was re-
duced by an excess of sodium borohydride (1.14 g). After stirring
at room temperature for 1.5 h, the organic phase was collected,
washed two times with distilled water (2”30 mL), and the solution
was concentrated to approximately 3 mL. The dark solution was
suspended in ethanol (280 mL), stored at 48C for 12 h, and centri-
fuged (3 min, 13000 rpm). The washing procedure was repeated
Figure 6. TEM image of annealed hybrid material H4, showing narrow size
distribution and smectic organization of hybrid nanoparticles (core size
ꢀ
2 nm) modified with mesogen I.
Chem. Eur. J. 2015, 21, 10082 – 10088
www.chemeurj.org
10087
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
twice. After that the refinement was carried out. Briefly, a small
amount of toluene (25 mL) was added to the sediment, after that
the mixture was precipitated with acetone (100 mL) and centri-
fuged with a speed of 2500 rpm for 3 min. The collected precipi-
tate was washed with a small amount of acetone. A fraction of
small gold nanoparticles (diameter less than 1.6 nm was separated
with the supernatant). Finally, the nanoparticles were dissolved in
toluene (20 mL) and centrifuged for 5 min to remove insoluble ag-
gregates. A sample with C H SH, as the primary ligand was used
[4] M. Urbanski, B. Kinkead, H. Qi, T. Hegmann, H.-S. Kitzerow, Nanoscale
2
010, 2, 1118–1121.
5] Q. Liu, M. G. Campbell, J. S. Evans, I. I. Smalyukh, Adv. Mater. 2014, 26,
178–7184.
6] a) B. Donnio, P. García-Vµzquez, J. L. Gallani, D. Guillon, E. Terazzi, Adv.
Mater. 2007, 19, 3534–3539; b) M. Draper, I. M. Saez, S. J. Cowling, P.
Gai, B. Heinrich, B. Donnio, D. Guillon, J. W. Goodby, Adv. Funct. Mater.
[
[
7
2
011, 21, 1260–1278; c) J. Paczesny, M. Wojcik, K. Sozanski, K. Nikiforov,
C. Tschierske, A. Lehmann, E. Gorecka, J. Mieczkowski, R. Holyst, J. Phys.
Chem. C 2013, 117, 24056–24062.
6
13
for the ligand-exchange reaction. The core diameter and the com-
position of the ligand were determined by TEM, H NMR spectros-
[7] K. Kanie, T. Sugimoto, J. Am. Chem. Soc. 2003, 125, 10518–10519.
1
[
8] a) L. Cseh, G. H. Mehl, J. Mater. Chem. 2007, 17, 311 –315; b) M. Wojcik,
M. Kolpaczynska, D. Pociecha, J. Mieczkowski, E. Gorecka, Soft Matter
copy, TGA, XPS spectroscopy, and SAXRD (see the Supporting Infor-
mation). A nearly monodisperse AuNP sample (309 atoms per NP
as estimated from SAXRD and TEM—Figure 1 in the Supporting In-
formation and calculations according to Ref. [25]), with a nearly
spherical metal core of (2.1Æ0.2) nm diameter and about 91 hexa-
nethols (LX) attached to the metal surface, were used as an entry
point to obtain hybrids. The exchange reaction leading to the
hybrid nanoparticles and synthesis of mesomorphic ligands is illus-
trated in Scheme 1 and described in detail in the Supporting Infor-
mation. The small angle X-ray diffraction (SAXRD) patterns for the
powder as well as partially aligned samples were obtained with
a Bruker Nanostar system. The nanoparticle samples were aligned
by shearing of a small amount of the material placed on the
Kapton tape at a temperature of approximately 808C. The same
Bruker Nanostar system was also used for the scattering experi-
ments (SAXRD). The scattering data from nanoparticle solutions in
toluene were analyzed assuming a spherical form factor for non-in-
teracting gold particles (structure factor S=1) and a Shultz distri-
bution (zero particles with zero diameter) of the particle sizes. The
gold cluster size was also evaluated from the broadening of the X-
2
010, 6, 5397–5400.
9] V. M. Marx, H. Girgis, P. A. Heiney, T. Hegmann, J. Mater. Chem. 2008, 18,
983–2994.
[
2
[
[
10] a) K. Kanie, M. Matsubara, X. Zeng, F. Liu, G. Ungar, H. Nakamura, A. Mar-
umatsu, J. Am. Chem. Soc. 2012, 134, 808–811; b) S. Mischler, S. Guerra,
R. Deschenaux, Chem. Commun. 2012, 48, 2183–2185; c) S. Frein, J.
Boudon, M. Vonlanthen, T. Scharf, J. Barberµ, G. Süss-Fink, T. Bürgi, R.
Deschenaux, Helv. Chim. Acta 2008, 91, 2321–2337.
11] a) W. Lewandowski, K. Jatczak, D. Pociecha, J. Mieczkowski, Langmuir
2
013, 29, 3404–3410; b) W. Lewandowski, D. Constantin, K. Walicka, D.
Pociecha, J. Mieczkowski, E. Gorecka, Chem. Commun. 2013, 49, 7845–
847.
7
[
[
12] M. M. Wojcik, M. Gora, J. Mieczkowski, J. Romiszewski, E. Gorecka, D. Po-
ciecha, Soft Matter 2011, 7, 10561–10564.
13] X. Mang, X. Zeng, B. Tang, F. Liu, G. Ungar, R. Zhang, L. Cseh, G. H. Mehl,
J. Mater. Chem. 2012, 22, 11101–11106.
[14] M. J. Hostetler, A. C. Templeton, R. W. Murray, Langmuir 1999, 15, 3782–
3789.
[
[
[
[
[
15] A. Dass, A. Stevenson, G. R. Dubay, J. B. Tracy, R. W. Murray, J. Am. Chem.
Soc. 2008, 130, 5940–5946.
16] G. H. Woehrle, L. O. Brown, J. E. Hutchison, J. Am. Chem. Soc. 2005, 127,
2172–2183.
17] Calculations according to: M. J. Hostetler, MJ. E. Wingate, C. J. Zhong,
CJ. E. Harris, R. W. Vachet, M. R. Clark, Langmuir 1998, 14, 17–30.
18] A. C. Templeton, M. J. Hostetler, C. T. Kraft, R. W. Murray, J. Am. Chem.
Soc. 1998, 120, 1906–1911.
[26]
ray signals from gold crystal lattice by using the Schererr model.
Acknowledgements
19] a) D. Li, Q. He, Y. Cui, K. Wang, X. Zhang, J. Li, Chem. Eur. J. 2007, 13,
2
224–2222; b) C. N. R. Rao, G. U. Kulkarni, P. J. Thomas, P. P. Edwards,
The authors greatly acknowledge financial support from the
Ministry of Science and Higher Education Republic of Poland,
grant Iuventus Plus no. 0618/IP3/2011/71 and the National Sci-
ence Centre (Poland), grant no. DEC-2013/08M/ST5/00781.
Chem. Eur. J. 2002, 8, 28–35.
[
[
20] Z. Wu, J. Chen, R. Jin, Adv. Funct. Mater. 2011, 21, 177–183.
21] a) E. E. Foos, A. W. Snow, M. E. Twigg, M. G. Ancona, Chem. Mater. 2002,
14, 2401–2408; b) T. Kong, J. Zeng, X. Wang, X. Yang, J. Yang, S.
McQuarrie, A. McEwan, W. Roa, J. Chen, J. Z. Xing, Small 2008, 4, 1537–
1543.
Keywords: gold · hybrid materials · ligand exchange · liquid
crystals · nanoparticles
[22] D. G. Castner, K. Hinds, D. W. Grainger, Langmuir 1996, 12, 5083–5086.
[
[
23] J. M. Pringle, O. Winther-Jensen, C. Lynam, G. G. Wallace, M. Forsyth,
D. R. MacFarlane, Adv. Funct. Mater. 2008, 18, 2031–2040.
24] a) D. V. Leff, L. Brandt, J. R. Heath, Langmuir 1996, 12, 4723–4730;
b) R. H. Terrill, T. A. Postlethwaite, C. Chen, J. E. Hutchison, C. Poon, A.
Tarzis, A. Chen, M. R. Clark, C. S. Johnson, E. T. Samulski, J. M. Desimone,
R. W. Murray, Abstr. Pap. Am. Chem. Soc. 1995, 210, 380.
[
[
1] Handbook of Liquid Crystals, Vols. 1–3 (Eds.: D. Demus, J. Goodby, G. W.
Gray, H.-W. Spiess, V. Vill), Wiley-VCH, Weinheim, 1998.
2] a) H. K. Bisoyi, S. Kumar, Chem. Soc. Rev. 2011, 40, 306–319; b) M. St e˛ -
pie n´ , B. Donnio, J. L. Sessler, Angew. Chem. Int. Ed. 2007, 46, 1431–
[
[
25] M. Wojcik, W. Lewandowski, J. Matraszek, J. Mieczkowski, J. Borysiuk, D.
Pociecha, E. Gorecka, Angew. Chem. Int. Ed. 2009, 48, 5167–5169;
Angew. Chem. 2009, 121, 5269–5271.
1435; Angew. Chem. 2007, 119, 1453–1457; c) Q. Liu, Y. Cui, D. Gardner,
X. Li, S. He, I. I. Smalyukh, Nano Lett. 2010, 10, 1347–1353; d) K.-H. Lee,
S.-M. Kim, H. Jeong, Y. Pak, H. Song, J. Park, K.-H. Lim, J.-H. Kim, Y. S.
Kim, H. C. Ko, I. K. Kwon, G. Y. Jung, Adv. Mater. 2013, 25, 3209–3214.
3] a) C. Rockstuhl, T. Scharf, J. Microsc. 2008, 229, 281; b) D. Coursault, J.
Grand, B. Zappone, H. Ayeb, G. LØvi, N. FØlidj, E. Lacaze, Adv. Mater.
26] A. Patterson, Phys. Rev. 1939, 56, 978–982.
[
Received: November 27, 2014
2012, 24, 1461–1465.
Published online on June 2, 2015
Chem. Eur. J. 2015, 21, 10082 – 10088
www.chemeurj.org
10088
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim