Subscriber access provided by HACETTEPE UNIVERSITESI KUTUPHANESI
Article
Sulfate Mediated End-to-End Assembly of Gold Nanorods Seyyed Mohammad Hossein Abtahi, Nathan D. Burrows, Fred A. Idesis, Catherine J. Murphy, Navid B. Saleh, and Peter J. Vikesland Langmuir, Just Accepted Manuscript • DOI: 10.1021/acs.langmuir.6b04114 • Publication Date (Web): 18 Jan 2017 Downloaded from http://pubs.acs.org on January 25, 2017
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Langmuir is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 33
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
1
Sulfate Mediated End-to-End Assembly
2
of Gold Nanorods
3 4
S.M.H. Abtahi,1, 2, 3 Nathan D. Burrows,4 Fred A. Idesis,4 Catherine J. Murphy,4
5
Navid B. Saleh,5 and Peter J. Vikesland*1, 2, 3 1.
6
Blacksburg, VA, USA
7 8
2.
Virginia Tech, Institute for Critical Technology and Applied Science (ICTAS) Center for Sustainable Nanotechnology (VTSuN), Blacksburg, VA, USA
9 10
3.
Center for the Environmental Implications of Nanotechnology (CEINT), Duke University, Durham, NC, USA
11 4.
12
15
Department of Chemistry, University of Illinois at Urbana-Champaign, 600 S. Matthews Ave., Urbana, IL, 61801, USA
13 14
Virginia Tech, Department of Civil and Environmental Engineering,
5.
The University of Texas at Austin, Department of Civil, Architectural and Environmental Engineering, Austin, TX 78712, USA
16
ACS Paragon Plus Environment
1
Langmuir
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 2 of 33
17 18
Abstract
19
There is interest in the controlled aggregation of gold nanorods (GNRs) for the
20
production of extended nanoassemblies. Prior studies have relied upon chemical
21
modification of the GNR surface to achieve a desired final aggregate structure. Herein,
22
we illustrate that control of electrolyte composition can facilitate end-to-end assembly of
23
cetyltrimethylammonium bromide (CTAB) coated GNRs. By adjusting either the sulfate
24
anion concentration or the exposure time it is possible to connect GNRs in chain-like
25
assemblies. In contrast, end-to-end assembly was not observed in control experiments
26
using monovalent chloride salts. We attribute the end-to-end assembly to the localized
27
association of sulfate with exposed quaternary ammonium head groups of CTAB at the
28
nanorod tip. To quantify the assembly kinetics, visible-near infrared extinction spectra
29
were collected over a pre-determined time period and the colloidal behavior of the GNR
ACS Paragon Plus Environment
2
Page 3 of 33
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
30
suspensions was interpreted using plasmon band analysis. Transmission electron
31
microscopy and atomic force microscopy results support the conclusions reached via
32
plasmon band analysis and the colloidal behavior is consistent with Derjaguin-Landau-
33
Verwey-Overbeek (DLVO) theory.
34
Introduction
35
Gold and silver nanoparticles are of interest because of the unique electro-optical
36
properties (e.g., localized surface plasmon resonance [LSPR])1-5 that originate from the
37
collective behavior of their surface electrons. In the case of LSPR, visible-infrared light
38
interacts with a noble metal nanoparticle smaller than the incident radiation wavelength1,
39
6
40
conduction band electrons. The LSPR is a function of the local dielectric environment as
41
well as the size and shape of the nanoparticle.1,
42
extinction spectrum at the wavelength where the frequency of the incident light matches
43
the frequency at which the conduction electrons oscillate.1,
44
dependent on nanoparticle size and shape and shifts to longer wavelengths with an
45
increase in particle size.14-16 Anisotropic nanoparticles, such as gold nanorods (GNRs),
46
exhibit two different plasmon oscillations: one along the short axis (the transverse
47
oscillation) and the other along the long axis (the longitudinal oscillation).5, 17-21 Each of
48
these oscillations exhibits its own specific LSPR band. The transverse LSPR band
49
typically occurs at a wavelength between 500-550 nm and shifts to longer wavelengths
50
with an increase in diameter. The longitudinal LSPR band is aspect ratio (AR:
51
length/diameter) sensitive, and occurs at longer wavelengths for GNRs with higher
52
AR.19, 22, 23 The extinction (absorbance + scattering) of these peaks is linearly related to
and resonant interactions result in the coherent oscillation of the nanoparticle's
2, 7-12
ACS Paragon Plus Environment
A peak is observed in the
6, 8, 13
Peak location is
3
Langmuir
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 4 of 33
53
the concentration of nanoparticles in suspension. Changes in the number concentration,
54
the AR, or the local dielectric environment can be detected by a change in extinction or
55
by the shift of these two plasmon peaks.14, 24
56
The electro-optical properties of GNRs can be significantly altered by nanoparticle
57
assembly. We use the term controlled assembly to refer either to end-to-end, end-to-
58
side, or side-by-side assembly.25,
59
plasmon preferentially and are of particular interest for sensing technologies that rely on
60
large electric fields.26-28 Extended end-to-end GNR assemblies have previously been
61
produced by manipulation of GNR surface chemistry,29-32 by alteration of the chemistry
62
of the solution in which the GNRs are suspended,13, 30, 33, 34 by applying polystyrene and
63
surface enhanced Raman spectroscopy (SERS) markers,35 or by attachment of pH-
64
responsive DNA36 to the GNR surface. Side-by-side assemblies of GNRs have been
65
prepared by cysteine conjugated GNRs in the presence of lead37 and in adiptic acid
66
treated GNR suspensions.38 Interchangeable end-to-end and side-by-side assemblies
67
of GNRs have been prepared by addition of tetrahydrofuran,39 dithiol polyethylene
68
glycol,40 bifunctional poly(ethylene glycol, PEG),26 or antibodies,25 onto GNR surfaces
69
where the final assembled configuration is controlled by the concentration of the organic
70
molecule. Each of these approaches requires extensive nanoparticle functionalization.
26
End-to-end assemblies couple the longitudinal
71
In this work, we illustrate a simple method in which sulfate ions are used to facilitate
72
end-to-end GNR assembly in aqueous suspension. The coordinating activity of sulfate
73
anions enables conversion of >80% of well-dispersed GNRs in a suspension into
74
extended structures (chains of dimers, trimers, tetramers, etc.). For these experiments
75
we used cetyltrimethylammonium bromide capped GNRs (CTAB-GNRs) because CTAB
ACS Paragon Plus Environment
4
Page 5 of 33
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
76
is the de facto GNR capping agent in GNR synthesis5, 18, 41-44. Visible-near infrared (Vis-
77
NIR) extinction spectra were collected over a pre-determined time period and the
78
colloidal behavior of the GNR suspensions was interpreted using plasmon band
79
analysis. Transmission electron microscopy (TEM) and atomic force microscopy (AFM)
80
were used to further characterize the samples. These results are self-consistent and
81
can be explained by Derjaguin-Landau-Verwey-Overbeek (DLVO) theory.
82
Materials and Methods
83
Preparation of CTAB-coated GNRs
84
GNRs were synthesized via the well-established seed-mediated surfactant-directed
85
method.45, 46 In brief, 3-4 nm spherical gold nanoparticle seeds were produced through
86
the reduction of chloroauric acid (HAuCl4•3H2O, 2.5×10-4 M in aqueous solution) in the
87
presence of CTAB ([CH3(CH2)15N(CH3)3]+Br-, 0.10 M) by addition of ice-cooled and
88
freshly prepared sodium borohydride (NaBH4, 0.010 M). A rapid color change to light
89
brown indicated seed formation. In the next step, GNRs of aspect ratio 4.4 were grown
90
from an aqueous growth solution consisting of 0.10 M CTAB, 5.0×10-4 M HAuCl4,
91
8.0×10-5 M AgNO3, and 5.5×10-4 M ascorbic acid, which became colorless upon acid
92
addition. Once the gold seeds were added to the growth solution, it changed color to
93
dark brown over 1-2 h indicating GNR formation. Spheres, cubes, and other shaped
94
byproducts
95
centrifugation (25 min at 8000 rcf), repeated 5 times with pellet resuspension in 800 µM
96
CTAB, yielding CTAB-GNRs. ICP-MS measurements indicated that the GNR stock had
97
a 41.0±0.3 mg/L atomic gold concentration. Accordingly, a concentration of 2.1×1014
(minor
component)
and
chemical
byproducts
ACS Paragon Plus Environment
were
removed
by
5
Langmuir
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 6 of 33
98
GNRs/L was calculated based on the density of gold (19.3 g/cm3)47 and the TEM
99
determined average GNR size of 28.8 ± 0.1 nm length and 6.6 ± 0.1 nm diameter.
100
End-to-end linkage of CTAB-GNRs in CaSO4 and MgSO4
101
Suspensions of CTAB-GNRs (1.1×1014 GNRs/L) were incubated over a range of
102
CaSO4 and MgSO4 concentrations (1-5 mM) for 24 h. To initiate an experiment, CTAB-
103
GNRs were washed by centrifugation (20 min at 11000 rcf) and 90% of the supernatant
104
was replaced with DI water, followed by bath sonication for 1 min. The resuspended
105
GNRs were readily dispersed and were colloidally stable. This step removed excess
106
CTAB from the GNR suspension. Following a second centrifugation step (20 min at
107
11000 rcf), 90% of the supernatant was replaced with a given CaSO4 or MgSO4 solution
108
followed by 1 min of bath sonication. The GNR suspensions were then transferred to
109
Vis-NIR cuvettes that were used for colloidal stability monitoring over 24 h. During the
110
first hour, 10 min sampling intervals were used. Subsequently, one-hour intervals were
111
used for the remaining 23 h. Vis-NIR extinction spectra were collected between 400-
112
1200 nm. TEM samples were prepared after 3 and 10 h of exposure to salt solutions. To
113
quantitatively capture the temporal changes in the extinction of the GNR suspensions,
114
individual peaks within the Vis-NIR spectrum were fit with Lorentzian distributions48,
115
using Grams/AI software (Thermo-Fisher, Waltham, MA).
49
116
TEM samples were prepared by submerging 200 mesh carbon film coated copper
117
grids in a given GNR suspension and then fast dried by putting the grid on top of an
118
80 °C hot plate for 10 s. An AFM sample was prepared following 3 h GNR exposure to
119
salt solutions. Negatively charged AFM metal specimen discs were submerged
ACS Paragon Plus Environment
6
Page 7 of 33
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
120
vertically in a petri dish containing 5 mL GNR suspension sample for half an hour
121
followed by DI water rinsing prior to AFM measurement.
122
Instrumentation
123
An Agilent Cary 5000 spectrometer (Santa Clara, CA) and disposable polystyrene
124
cuvettes (10 mm pathlength) were used for Vis-NIR extinction spectroscopy. TEM
125
images were taken using a Philips (Thermo-Fisher, Waltham, MA) EM420 conventional
126
transmission electron microscope. A Malvern (Worcestershire, UK) Zetasizer Nano-ZS
127
was used for electrophoretic mobility measurements. A Bruker (Billerica, MA)
128
Nanoscope IIIa AFM and Ted Pella (Redding, CA) 15 mm metal specimen discs were
129
used for atomic force microscopy. Sonication was performed using a Fisher Scientific
130
FS20H bath sonicator. DI water (18.2 MΩ) was produced by a Thermo Scientific
131
Barnstead nanopure system.
132
Results and Discussion
133
GNR characterization
134
The GNRs were characterized via TEM, Vis-NIR spectroscopy, electrophoretic
135
mobility, and ICP-MS measurements. The GNRs (inset to Figure 1; Figure S1) have an
136
average length of 28.8 ± 0.1 nm and an average diameter of 6.6 ± 0.1 nm (the noted
137
errors reflect 95% confidence interval based on n = 2104 measurements obtained using
138
ImageJ). These relative dimensions result in an average AR of 4.4 ± 0.1. These GNRs
139
exhibit a transverse extinction band at 518 nm and a longitudinal band at 756 nm, which
140
are the expected wavelengths for GNRs of these dimensions.45 The electrophoretic
141
mobility of the GNRs in DI water was determined to be +3.58 µm•cm/V•s. Applying
ACS Paragon Plus Environment
7
Langmuir
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 8 of 33
142
Henry’s equation while considering Smoluchowski equation for Henry’s constant
143
resulted in calculated ζ-potential of +45.6 mV. We note that for the GNRs and the
144
range of ionic strengths tested in this study,
145
the nanoparticle effective diameter in nm) varies from 4.09 to 9.13. Because
146
the Smoluchowski equation is an acceptable approximation of ζ-potential (details and
147
example calculations are presented in SI; pages 2-3).50,
148
consistent with the literature and illustrate that the rods have a positive effective surface
149
charge.44, 51
150 151 152 153
Figure 1. Vis-NIR extinction spectrum for the synthesized CTAB-GNRs after centrifugal purification and washing. Inset: Transmission electron micrograph of CTABGNRs.
ૂa (ૂ is 1/Debye length in 1/nm and “a” is
51, 52
ૂa>>1,
These values are
154
ACS Paragon Plus Environment
8
Page 9 of 33
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
155
Langmuir
End-to-end assembly of CTAB-coated gold nanorods in CaSO4 and MgSO4
156
The in situ development of GNR assemblies was probed via Vis-NIR extinction
157
spectroscopy. It has been shown previously that during side-to-side and end-to-side
158
assemblies, the extinction of the initial transverse LSPR band increases and the band
159
slightly red shifts to higher wavelengths; while the extinction of the longitudinal band
160
decreases and blue shifts to lower wavelengths, thus resulting in the convergence of the
161
two bands.26 The calculated change in extinction intensity versus the incident light
162
wavelength and the corresponding transverse and longitudinal bands of these two
163
assemblies are presented in Table S3 and Figure S6. These calculations were
164
developed based on our experiment parameters, while employing extended Mie theory
165
modeling. In end-to-end assembly the bands behave quite differently. For aggregates
166
with this configuration, neither the transverse nor the longitudinal band shifts; however,
167
the extinction of the longitudinal band significantly decreases as the total average length
168
of the GNR chain increases. Under these conditions, a third plasmon band appears at
169
wavelengths larger than that of the original longitudinal band (Figure S5). Because of
170
extended plasmonic coupling this new plasmon band red shifts to longer and longer
171
wavelengths as the GNR chain length increases.22, 26, 53-55
172
To evaluate the aggregation behavior of the GNRs, CTAB-GNR suspensions were
173
exposed to 1, 2, and 5 mM CaSO4 or MgSO4. We observed salt concentration
174
dependent changes in the visual characteristics of the suspensions. In the presence of
175
DI water the suspension is brownish in color, while at higher salt concentrations the
176
color darkens. At a salt concentration of 5 mM the colloidal stability of the suspension
177
was lost within 24 h (Figure S2). As illustrated in representative plots for 2 mM CaSO4
ACS Paragon Plus Environment
9
Langmuir
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 10 of 33
178
and 2 mM MgSO4 (Figure 2) there are readily observable declines in the intensities of
179
both the transverse and longitudinal bands and additional bands appear at longer
180
wavelengths. Such behavior is consistent with the previously described end-to-end
181
aggregation process.
182
concentrations.
Qualitatively similar results were observed at the other salt
183
To quantify the declines in peak intensity we plotted the change in extinction for
184
both the transverse and longitudinal bands for each salt concentration. As shown in
185
Figure 2C-F, a sharp drop in the longitudinal peak intensity was observed within the first
186
hour of exposure of the GNRs to all concentrations of CaSO4 and MgSO4, accompanied
187
by a slower decrease in intensity over the next 23 h. In contrast, the transverse band
188
intensity did not change as noticeably. To interpret the temporal changes in the
189
longitudinal and transverse bands we fitted our collected data using a simple second
190
order aggregation model56, 57 (details are presented in the SI; page 4):
191
ௗ[ௗ] ௗ௧
= −݇[]݀ݎଶ
(1)
192
Where [rod] is the number concentration of individual, fully dispersed rods and k is the
193
observed second order rate constant. Other more complex aggregation models may be
194
more appropriate for this system due to its complexity; however, our purpose in using
195
this model is to capture the differential behavior of the two extinction bands. Number
196
concentration-based rate constants and associated 95% confidence intervals for single
197
CTAB-GNRs in the presence of CaSO4 and MgSO4 are presented in Table 1. We note
198
that for 5 mM CaSO4 the second order model did not fit the collected data as well as it
199
did for other experimental conditions. We attribute this behavior to the more rapid
200
aggregation observed with CaSO4 and the greater degree of colloidal instability.
ACS Paragon Plus Environment
10
Page 11 of 33
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
201 202 203 204 205 206 207 208
Langmuir
Figure 2. Vis-NIR extinction spectra of 1.1×1014 GNRs/L exposed to A) 2 mM CaSO4 and B) 2 mM MgSO4 from zero hours (red spectra) to 24 hours (blue spectra). Changes in the average normalized transverse band and longitudinal band extinction maxima of CTAB-GNRs (1.1×1014 GNRs/L) in 1-5 mM CaSO4 and MgSO4 solutions as a function of exposure time. (C) Transverse band-CaSO4 (D) Longitudinal band-CaSO4 (E) Transverse band-MgSO4 (F) Longitudinal bandMgSO4.
209
ACS Paragon Plus Environment
11
Langmuir
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
210 211 212 213
Table
1.
Second
order
number
Page 12 of 33
concentration
rate
constants
L ) for single CTAB-GNRs in the presence of different 1.07 × 10 ( goldnanoro ds ). S concentrations of CaSO4 and MgSO4. Both the changes in the transverse band and the longitudinal band are provided. Errors are reported at the ±95% confidence intervals.
(
14
Salt Concentration (mM) 1 2 5
Transverse band CaSO4 MgSO4 0.0006 ± 0.0001 0.0007 ± 0.0001 0.0065 ± 0.0000 0.0032 ± 0.0001 0.22 ± 0.049 0.012 ± 0.0006
Longitudinal band CaSO4 MgSO4 0.025 ± 0.0020 0.019 ± 0.0021 0.05 ± 0.0051 0.028 ± 0.0026 0.49 ± 0.072 0.051 ± 0.0053
214 215
An increase in ionic strength significantly increased the rate at which the GNRs
216
assembled. In all cases, the rate of change of the longitudinal band was greater than
217
that of the transverse band (Table 1). Over the 24 h period that the GNRs were exposed
218
to 1 mM and 2 mM of CaSO4 and MgSO4 there was minimal change in the transverse
219
band intensity, while simultaneously there was a 40-60% decline in the longitudinal
220
band intensity. If the assemblies were colloidally unstable and were precipitating out of
221
suspension and leaving only single GNRs behind then the rate constants for the change
222
in the transverse and longitudinal bands of single GNRs should be equal (Table S1 and
223
Figure S3). However, the much slower change in the transverse band relative to the
224
longitudinal band supports the conclusion that the GNRs are assembling into extended
225
structures. In end-to-end nanoparticle assembly the transverse band is not expected to
226
shift, while the longitudinal band will decrease in intensity and red-shift.
227
Following exposure of the GNRs to CaSO4 or MgSO4 a new shoulder with an
228
extinction maximum at a wavelength greater than the longitudinal band appears. The
229
intensity of this shoulder increases and red-shifts to longer wavelengths over time and
230
at increased salt concentrations.
231
assembly of GNRs to produce new structures with elongated AR (i.e., end-to-end
Formation of this shoulder is consistent with the
ACS Paragon Plus Environment
12
Page 13 of 33
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
232
assemblies). Past studies have illustrated that chain-like assemblies of GNRs display
233
the formation of an extinction band at longer wavelengths compared to individual
234
rods.25, 26, 30, 34 The peak wavelength is strongly related to the number of linked rods and
235
shifts to longer wavelengths as the chains become increasingly extended over time.
236
TEM and AFM images consistent with formation of end-to-end assemblies are
237
shown in Figure 3. Both the TEM and AFM images illustrate the formation of end-to-end
238
linkages. The AFM images unlike the TEM images were collected in situ without sample
239
drying. Observation of an extended end-to-end assembly in the AFM image supports
240
the argument that the presence of these assemblies in the TEM images is not
241
necessarily a result of drying mediated artifacts. TEM images taken at different times
242
and salt concentrations suggest that the end-to-end chains of GNRs grow longer over
243
time and with an increase in ionic strength (Figure 3B-C, Figure S4).
ACS Paragon Plus Environment
13
Langmuir
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 14 of 33
A
B
C
50 nm 244 245 246 247
Figure 3. A) Representative AFM image of CTAB-GNRs after 3 h exposure to 2 mM CaSO4, B) TEM image of CTAB-GNRs after 3 h exposure to 2 mM CaSO4, C) TEM image of CTAB-GNRs after 10 h exposure to 2 mM CaSO4.
248
ACS Paragon Plus Environment
14
Page 15 of 33
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
249
Langmuir
Extinction spectra peak fitting Prior studies have established that there is a linear relationship between the peak
250 251
wavelength location and AR.22,
58
252
assemblies of different length can be estimated by application of extended Mie theory
253
(see SI for details).22, 58-61 Knowing the physical and chemical properties of the GNRs
254
and the solution chemistry employed herein, we were able to calculate the expected
255
location of the extinction band for each end-to-end configuration (see SI for details of
256
calculations). Our GNRs have an AR of 4.4, therefore the calculated λmax values for
257
dimer (AR=8.8), trimer (AR=13.2), and tetramer (AR=17.6) end-to-end linked AuNRs
258
are 860, 965, and 1070 nm, respectively (Figure 4A). These values are qualitatively
259
similar to those theoretically calculated for end-to-end assemblies of GNRs of similar
260
AR.
Theoretical peak wavelengths for end-to-end GNR
ACS Paragon Plus Environment
15
Langmuir
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 16 of 33
261 262 263 264 265 266
Figure 4. A) Plasmon band wavelengths of different GNR end-to-end assemblies as calculated by extended Mie theory. B) Exemplary figure illustrating quality of fit for experimental vs. fitted Vis-NIR spectrum for GNRs exposed to 5 mM MgSO4 after 1 hr. The magnitudes of individual extinction bands are illustrated.
267
ACS Paragon Plus Environment
16
Page 17 of 33
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
268
Based upon these wavelengths we used a non-linear least squares iterative62
269
procedure (within GRAMS/AI) to fit the collected Vis-NIR spectra for each experiment as
270
the summation of five separate Lorentzian bands that correspond to the transverse
271
band and the longitudinal band for single, dimer, trimer, and tetramer GNRs. The
272
Lorentzian distribution can be explained as a Fourier-transform of exponentially
273
decaying oscillations63. We used extended Mie theory modeling to calculate the change
274
in extinction intensity versus the incident light wavelength and to locate the plasmon
275
bands of different end-to-end assemblies as has been done previously22 (refer to SI
276
page 7 for calculation details). As a result, Lorentzian distributions were used for curve
277
fitting. In this exercise, the center of each peak is fixed while the amplitude (height) and
278
width are unknown parameters. Using this constrained fitting approach, we are able to
279
fit and reproduce the collected Vis-NIR spectra quite readily (Figure 4B). By fitting each
280
of the spectra collected over the course of an experiment we can probe the relative
281
numbers of single, dimer, trimer, and tetramer assemblies present at any time. Figure 5
282
illustrates the temporal changes in each Lorentzian distribution band for the CaSO4
283
experiments, while the results for the MgSO4 experiments are shown in Figure S7.
ACS Paragon Plus Environment
17
Langmuir
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 18 of 33
284 285 286 287 288 289
Figure 5: Extinction intensities of different Lorentzian distribution band end-to-end assemblies of GNRs exposed to three different CaSO4 concentrations A) Transverse bands of single rod B) Longitudinal bands of single rod C) Longitudinal bands of dimer D) Longitudinal bands of trimer E) Longitudinal bands of tetramer
290
With an increase in salt concentration increasing numbers of GNRs participate in
291
end-to-end assembly and there is a concomitant decrease in the extinction of the
ACS Paragon Plus Environment
18
Page 19 of 33
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
292
longitudinal band that represents single GNRs (Figure 5B). At 1 and 2 mM CaSO4, the
293
extinction of the dimer band at 860 nm increases rapidly and then plateaus, while at 5
294
mM it increases and then decreases with time (Figure 5C). At 5 mM CaSO4, the dimer,
295
trimer, and tetramer extinction intensities rapidly increase but then decrease over
296
extended time. These results collectively show that at high salt concentrations the end-
297
to-end GNRs assemblies that are formed initially, grow further to be colloidally unstable,
298
and finally undergo precipitation. This precipitation trend is consistent with visual
299
observation of large dark pellets at the bottom of the sample cuvette (Figure S2).
300
The validity of our curve fitting technique was studied by analyzing 36 TEM images
301
taken from one sample (GNRs exposed to 2 mM CaSO4) at three different time periods
302
(3, 10, and 24 h). A total of 270 GNRs were counted via TEM image analysis. Figure 6
303
shows the change in percentage of these assemblies as a function of time. As
304
expected, the length of the end-to-end assemblies generally increase over time. As an
305
example, after 3 h almost 50% of GNRs are in the dimer form and no tetramers were
306
found; but after 24 h 50% of GNRs are in the tetramer form while only 12% single GNRs
307
were observed. Because of the potential for drying mediated artifacts and the potential
308
effects of the TEM substrate on aggregate formation the results presented in Figure 6
309
cannot be directly compared to the plasmon band analyses; however, the general trend
310
is the same with longer and longer assemblies forming with extended time.
ACS Paragon Plus Environment
19
Langmuir
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 20 of 33
311 312 313 314
Figure 6. Percentage of GNRs forming different end-to-end assemblies. Analysis of 36 TEM images taken from GNRs exposed to 2 mM CaSO4 at three different time periods. Representative TEM images are shown Figure S4.
315 316
Anion identity and valence dictate assembly
317
Past studies have suggested that CTAB forms a dense 3.5 nm thick vesicle-like
318
bilayer on the GNR side facets, while it may not fully coat the ends.44, 64, 65 The presence
319
of the positively charged vesicle-like bilayer around the side facets protects the GNRs
320
from undergoing side-by-side aggregation. To investigate the chemistry underlying the
321
observed end-to-end assembly we exposed CTAB-GNRs to CaCl2, MgCl2, NaCl, and
322
Na2SO4 salts at the same ionic strength used in the CaSO4 and MgSO4 experiments.
ACS Paragon Plus Environment
20
Page 21 of 33
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
323
The collected Vis-NIR extinction spectra over a period of 24 h are shown in Figures S8
324
and S9. The decline in the transverse and longitudinal absorption bands is shown in
325
Figures S10 and S11.
326
No change in the extinction of the transverse band, a
