[56] makes our rods ‘‘softer’’ and more highly charged, but nonetheless we do observe concentration-dependent ordering.
In our experiments [55], gold nanorods of aspect ratios 13–18 were concentrated
and separated from spherical nanoparticles by centrifugation. Thermogravimetric
analysis showed that @20% of the total mass of the nanorods was associated with
the CTAB surfactant. This is significantly larger than that calculated for monolayer
coverage (3 wt%) on the basis of particle surface area, and suggests that the nanorods are covered with multiple CTAB layers; indeed, El-Sayed has spectroscopic
evidence for CTAB bilayers on gold nanorods [56]. We found that the presence of
the surfactant coating was of key importance, not only for hydrophilic stabilization
of the nanorods in water but also for controlling long-range self-assembly in concentrated dispersions; we found that the optimum conditions required for in situ
liquid crystalline ordering involved redispersing the nanorods in 1–100 mM CTAB.
Above this surfactant concentration, the nanorods precipitated instantaneously and
were unable to be redispersed to image in the electron microscope. Below this
CTAB concentration, nanorods were also unable to be redispersed in water. Although not fully explored, we believe these effects are not simply due to the ionic
nature of CTAB, as @1 mM NaCl precipitates the gold nanorods.
In general, the 18 aspect ratio nanorod solutions were dark brown in color (when
concentrated) and had a weak absorbance maximum in the visible at @530 nm, in
addition to a near-infrared absorbance at @1700 nm [27]. Thin films of the concentrated dispersions supported on glass slides showed iridescent droplets @0.1
mm in diameter under polarizing light microscopy (Figure 9.6). The observed textures were indicative of localized regions of liquid crystalline ordering and are
similar to nematic droplets observed in boehmite nanoneedle solutions [52]. Significantly, the liquid crystalline droplets were stable up to 200
C in air, after which
the surfactant began to degrade, although the nanorods remained unchanged in
size and shape. Similar experiments with concentrated surfactant solution alone
showed much smaller ‘‘speckles’’ in the polarizing microscope and no liquid crystalline textures.
Small-angle X-ray scattering (SAXS) experiments were undertaken to determine
the extent of long-range ordering in concentrated (@5–10 wt% of solids) and diluted (by a factor of @1000) dispersions of the gold nanorods [55]. The scattering
curves show ripples in the scattered X-ray intensity due to particle shape and interparticle interactions. The data were fitted to a model consisting of core–shell
cylinders stacked with a Gaussian distribution of interparticle distances using a
method of non-linear least-squares fitting. The fitting parameters included the radius of the nanorods, the number of particles in a stack, the width of the Gaussian
distribution of interparticle distances in the stack, the surfactant layer thickness
and the major radius of the elliptical impurities. Attempts to fit the SAXS data to
isolated rods failed; rod stacks were required. The fits suggest that the concentrated
solutions contained self-assembled stacks of ca. 200 nanorods, each of which had a
surfactant coating 3.9 nm in thickness, consistent with a CTAB bilayer [56]. In
contrast, smaller clusters of @30 rods were present in the more dilute sample. The
9 Synthesis, Assembly and Reactivity of Metallic Nanorods
294
In our experiments [55], gold nanorods of aspect ratios 13–18 were concentrated
and separated from spherical nanoparticles by centrifugation. Thermogravimetric
analysis showed that @20% of the total mass of the nanorods was associated with
the CTAB surfactant. This is significantly larger than that calculated for monolayer
coverage (3 wt%) on the basis of particle surface area, and suggests that the nanorods are covered with multiple CTAB layers; indeed, El-Sayed has spectroscopic
evidence for CTAB bilayers on gold nanorods [56]. We found that the presence of
the surfactant coating was of key importance, not only for hydrophilic stabilization
of the nanorods in water but also for controlling long-range self-assembly in concentrated dispersions; we found that the optimum conditions required for in situ
liquid crystalline ordering involved redispersing the nanorods in 1–100 mM CTAB.
Above this surfactant concentration, the nanorods precipitated instantaneously and
were unable to be redispersed to image in the electron microscope. Below this
CTAB concentration, nanorods were also unable to be redispersed in water. Although not fully explored, we believe these effects are not simply due to the ionic
nature of CTAB, as @1 mM NaCl precipitates the gold nanorods.
In general, the 18 aspect ratio nanorod solutions were dark brown in color (when
concentrated) and had a weak absorbance maximum in the visible at @530 nm, in
addition to a near-infrared absorbance at @1700 nm [27]. Thin films of the concentrated dispersions supported on glass slides showed iridescent droplets @0.1
mm in diameter under polarizing light microscopy (Figure 9.6). The observed textures were indicative of localized regions of liquid crystalline ordering and are
similar to nematic droplets observed in boehmite nanoneedle solutions [52]. Significantly, the liquid crystalline droplets were stable up to 200
C in air, after which
the surfactant began to degrade, although the nanorods remained unchanged in
size and shape. Similar experiments with concentrated surfactant solution alone
showed much smaller ‘‘speckles’’ in the polarizing microscope and no liquid crystalline textures.
Small-angle X-ray scattering (SAXS) experiments were undertaken to determine
the extent of long-range ordering in concentrated (@5–10 wt% of solids) and diluted (by a factor of @1000) dispersions of the gold nanorods [55]. The scattering
curves show ripples in the scattered X-ray intensity due to particle shape and interparticle interactions. The data were fitted to a model consisting of core–shell
cylinders stacked with a Gaussian distribution of interparticle distances using a
method of non-linear least-squares fitting. The fitting parameters included the radius of the nanorods, the number of particles in a stack, the width of the Gaussian
distribution of interparticle distances in the stack, the surfactant layer thickness
and the major radius of the elliptical impurities. Attempts to fit the SAXS data to
isolated rods failed; rod stacks were required. The fits suggest that the concentrated
solutions contained self-assembled stacks of ca. 200 nanorods, each of which had a
surfactant coating 3.9 nm in thickness, consistent with a CTAB bilayer [56]. In
contrast, smaller clusters of @30 rods were present in the more dilute sample. The
9 Synthesis, Assembly and Reactivity of Metallic Nanorods
294