Electro-Orientation of Silver Nanowires in Alternating Fields

Dec 17, 2018 - Electro-Orientation of Silver Nanowires in Alternating Fields ... In the electro-orientation spectrum, we observe frequency dispersion ...
0 downloads 0 Views 827KB Size
Subscriber access provided by University of Winnipeg Library

Interface Components: Nanoparticles, Colloids, Emulsions, Surfactants, Proteins, Polymers

Electro-orientation of silver nanowires in alternating fields Paloma Arenas-Guerrero, Angel V. Delgado, Antonio Ramos, and Maria L. Jiménez Langmuir, Just Accepted Manuscript • DOI: 10.1021/acs.langmuir.8b03122 • Publication Date (Web): 17 Dec 2018 Downloaded from http://pubs.acs.org on December 23, 2018

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 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 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.

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 23 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

Electro-orientation of silver nanowires in alternating elds Paloma Arenas-Guerrero,



Ángel V. Delgado,

Jiménez

†Department ‡Departament





Antonio Ramos,

and María L.

∗,†

of Applied Physics, Granada University, Granada

of Electronics and Electromagnetism, University of Sevilla, Spain

E-mail: [email protected]

Abstract In this work we analyse the orientation of silver nanowires immersed in aqueous solutions, under the eect of alternating electric elds in a broad frequency range, covering from a few Hz to several MHz. The degree of orientation is experimentally determined by electro-optical techniques, which present the advantage of measuring multiple particles at the same time. In the electro-orientation spectrum, we observe a frequency dispersion in the kHz range, and provide a theoretical explanation for this behaviour: at high frequencies, charge separation in the nanoparticles leads to a large induced dipole, responsible for strong orientation. On the other hand, at low frequencies, redistribution of the ions in solution gives rise to an induced double layer that screens the dipolar elds and, as a consequence, the degree of orientation decreases. Moreover, we measure the transient response when the electric eld is switched o, from which the size distribution of the polydisperse sample is obtained. The results match those given by electron microscopy determinations.

1

ACS Paragon Plus Environment

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

Introduction In the last decade, interest in the investigation of noble metal nanoparticles has rapidly increased due to their unique conductive, optical and mechanical properties. 1 For example, thanks to the existence of the surface plasmon resonance, applications of these particles have been found in enhanced Raman spectroscopy, photothermal therapy, or as optical sensors. In these techniques, the particle shape often plays an important role, and the performance can be improved in many cases by the use of elongated geometries. 2 For instance, it has been shown that high-aspect-ratio metal nanostructures, such as Au and Ag nanorods/nanowires, can be used to transport optical signals over distances of several micrometers. 3,4 Moreover, the antibacterial activity of silver nanoparticles, which has become an issue of intense research, 5 is also highly inuenced by particle shape. 6 In many applications of metallic non-spherical particles, promoting and/or controlling their orientation is an important goal. For instance, thin lms of silver nanowires can be used as conductive transparent electrodes in dierent optical devices. 7,8 The structure of the wires inside these lms can provide nanoscale control of the transmission, manipulation and switching of optical signals. Other examples are the reported particle assemblies with negative refractive index 9,10 or the nding that electromagnetic energy can be coherently guided in devices with size below the diraction limit, due to near-eld coupling of neighbour particles. 11,12 However, despite their potential, controlling the orientation and understanding the properties of metallic nanowires are still open issues. In this context, a system to monitor such orientation is no less important. While direct observation of the particles is an intuitive option, it presents some limitations. For instance, only large particles can be measured, and for these, viscous interactions with the cell cannot be ruled out. 13 Furthermore, the orientation cannot be accurately determined, and measuring a statistically signicant number of particles is normally time-consuming. In contrast, electro-optical methods are suitable for the study of nanosized particles, and allow the observation of the degree of orientation 2

ACS Paragon Plus Environment

Page 2 of 23

Page 3 of 23 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

of a large number at once. For these reasons, in this work we use electric birefringence, an electro-optical method based on the determination of the refractive index anisotropy induced in a suspension of non-spherical particles by an external electric eld. During the last years, this technique has been successfully applied to a number of nonspherical particles, 1421 although only a few works are devoted to a physical understanding of the observed phenomena. 2224 It is well known that the electric torque that causes the orientation is due to the dipole induced on the nanoparticle by the external electric eld. However, while this eect has been extensively studied in the case of non-conducting charged particles, only a few works deal with the electric response of conducting but weakly charged non-spherical particles, as those here treated. In these contributions, for instance, it has been shown that induced electro-osmosis 25,26 can play an important role in the orientation of metallic particles under slowly oscillating electric elds. 2729 In this work we provide a study of the electro-orientation of silver nanowires in aqueous suspension, in the presence of alternating electric elds. From the electric birefringence phenomenology, the orientation mechanisms are analysed and optical, electrical and geometrical properties of the particles are obtained.

Experimental Section Silver nanowires (AgNWs) were purchased from PlasmaChem, Germany. Figure 1 shows a transmission electron microscope (TEM) picture of these particles, where it can be observed that they are highly elongated and polydisperse, with a mean length of 3.0±2.3 µm and an average diameter of 121±17 nm (the errors indicate the standard deviations of the length and diameter distributions by number). The electric birefringence (EB) was determined with a home-made experimental device described elsewhere. 15,30 Briey, a linearly polarised He-Ne laser beam passes through the suspension. After crossing the sample, it traverses a quarter wave plate and another polariser,

3

ACS Paragon Plus Environment

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 23

Figure 1: TEM picture of the AgNWs. which is nearly crossed with the former (at 90◦ − α, being α = −3◦ ). All the optical plates were purchased from Edmund Optics, UK. After sonicating, the sample is placed in a quartz Kerr cell and thermostated at 15◦ C. A pair of electrodes with 1 mm separation and oriented at 45◦ with respect to the incident beam polarisation direction are immersed in the sample. The sinusoidal electric elds are obtained with a signal generator (Tektronix AFG 3101, USA) in a frequency range of 100 Hz-10 MHz. The light intensity transmitted by this setup is collected by a photodiode (Edmund Optics, UK) connected to a digital oscilloscope (Tektronix TDS 2012C, USA). The transmitted light intensity I at time t is directly related to the birefringence of the sample, ∆n(t), as 15

 ∆n(t) =

s

λ  −1  sin πl





I(t) sin α − α I0

(1)

where λ = 632 nm is the wavelength of the incident beam, l = 1 cm the path length inside the Kerr cell, and I0 the transmitted intensity in the absence of electric eld. The value of I0 does not change signicantly during the time of the experiment, indicating that sedimentation is negligible. Considering the relatively large size of the particles, the possibility that orientationinduced turbidity changes can aect the transmitted intensity I(t) to an extent comparable to electric birefringence, should be analyzed. With that aim, we measured I(t) for a polarized 4

ACS Paragon Plus Environment

Page 5 of 23 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

laser beam and from that we estimated the turbidity changes. This was used to evaluate the correction that must be introduced in the raw data. For this, we follow the procedure detailed in Ref. 31 The eect of the turbidity changes on the EB measurements was found to be below 0.2% in the Kerr regime and below 3% at saturating elds and hence, this correction was not considered necessary anymore.

Results and Discussion Characterisation of the wires The electrophoretic mobility ue of the AgNWs was determined by dynamic light scattering (Malvern Zetasizer NanoZS, Malvern Instruments, UK). From these measurements, the zeta potential ζ of the wires was calculated modelling them as innitely long cylinders and making use of the Ohshima model. 32 The results for dierent values of the electrolyte concentration are shown in Figure 2, where it can be observed that the particles present some negative electric charge, which can be attributed to the caping agents used to terminate the reaction during the synthesis, necessary for the stability of the suspension. The decrease of the zeta potential with the ionic strength is a consequence of the contraction of the Debye layer.

Figure 2: Electrophoretic mobility (left axis) and zeta potential (right axis) of the AgNWs as a function of the ionic strength. Figure 3a displays a zoom-in TEM picture of the wires, where it can be observed that 5

ACS Paragon Plus Environment

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

they seem to present a coating with a thickness of approximately 10 nm (and up to 40 nm on the wire tip). In order to analyse the composition of this layer, X-ray photoelectron spectroscopy (XPS) mearurements were carried out. This technique indicates the presence of oxygen, in a ratio of approximately 3:2 with respect to silver. The XPS peaks corresponding to these two elements are shown in Figure 3b. These results suggest that the particle surface is at least partially oxidised. (a)

(b)

Figure 3: a) Zoom-in TEM picture of the AgNWs. b) XPS peaks corresponding to silver and oxygen.

General features of the EB of silver nanowires When an electric eld is applied to a suspension of non-spherical nanoparticles, they tend to orient along the eld direction, and the suspension becomes optically anisotropic. For 6

ACS Paragon Plus Environment

Page 6 of 23

Page 7 of 23 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

axially symmetric particles, the electric birefringence ∆n = nk −n⊥ is dened as the dierence between the refractive index of the suspension along the directions parallel and perpendicular to the electric eld. In general, this quantity is given by

∆n = ∆nmax S

(2)

where S is the orientational order parameter of the suspension, which for elongated geometries ranges from 0 (random orientation) to 1 (total alignment). The saturation birefringence,

∆nmax , reads:

∆nmax =

N ∆αo 2ns ε0

(3)

being N the particle concentration by number, ns the refractive index of the solvent and

ε0 the vacuum permittivity. ∆αo is the optical polarisability anisotropy of the particles, dened as ∆αo = αao − αbo , where a and b indicate the directions parallel and perpendicular to the major axis, respectively. It must be noted that Equation 3 is strictly valid for a dilute suspension of small (as compared to the wavelength) anisotropic particles. For larger dimensions, a rigorous evaluation of the forward scattering matrix would be required. 33 Nevertheless, for our purposes it is enough to consider the existence of saturation and use

∆nmax as a tting parameter without attempting its theoretical calculation. Figure 4a shows the EB of a silver nanowire suspension as a function of time, for dierent frequencies of the applied electric eld. Here, it can be observed that when the eld is turned on the EB signal grows and reaches a stationary value. When the eld is turned o, the birefringence decays to zero. AgNWs exhibit negative birefringence in the whole explored frequency range, a feature that has been already observed for other conducting materials. 34 In those cases, the eect was due to a negative value of ∆αo , caused by the strong optical absorbance of the metallic particles. 34 However, the negative sign can also be due to an anomalous orientation of the AgNWs, with their major axis perpendicular

7

ACS Paragon Plus Environment

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 23

to the eld direction, resulting in a negative value of S , and hence of the birefringence (Equation 2). This has been observed for polydisperse suspensions and bidisperse systems at low eld frequencies. 15,35 In order to discriminate between both possibilities, we performed ow birefringence experiments. When rigid rods ow through a thin capillary, the velocity gradient produces an orientation of the particles with their major axis parallel to the channel, independently of their nature. This gives rise to an induced birefringence, whose sign is determined solely by

∆αo . In our experiments, we substituted the cell and vertical electrodes of the EB setup by an horizontal channel with a cross section 1×1 mm2 , and we pumped the suspension at a ow rate of 0.2 mL/s, which means a shear rate of approximately 300 s−1 . The results for AgNWs are shown in Figure 4b, together with measurements of sepiolite needles (Sigma Aldrich, USA), for the sake of comparison. The data indicate that, unlike the case of sepiolite particles, silver nanowires present a negative optical polarisability anisotropy, responsible for the negative sign of the electric birefringence. (b)

(a)

Figure 4: a) Electric birefringence of a 100 mg/L AgNW suspension in 0.03 mM KCl as a function of time. The electric eld pulses have an amplitude of 2 V/mm, a duration of 12.5 s and the indicated frequencies. b) Flow birefringence of suspensions of AgNWs at 200 mg/L and sepiolite needles at 2 g/L, as a function of time. For very elongated particles, a theoretical value of the polarisability anisotropy at optical frequencies can be obtained from the refractive index of the medium, nm , and the particle, 8

ACS Paragon Plus Environment

Page 9 of 23 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

np , as: (n2p − n2m )2 ∆α = Vp ε0 Re n2p + n2m o





(4)

where Vp is the particle volume. Using the bibliography value of np 36 and the particle dimensions obtained by microscopy, a theoretical value of ∆αo = −4.9 · 10−30 Fm2 was found for the AgNWs. Thus, the negative sign of the optical polarisability anisotropy is a consequence of particle absorbance. It is worth mentioning that, given the length of the AgNWs, one may expect some interaction among them. This would be reected in a slowing down of the transient decay, or equivalently a larger hydrodynamic size, for more concentrated samples, which is not observed. Furthermore, Figure 5a shows that birefringence is proportional to particle concentration, which further rules out any particle interactions. Figure 5b shows that the EB is also proportional to the square of the eld amplitude, E0 , a quadratic dependence known as Kerr's law, which is fullled for suciently weak elds. (b)

(a)

Figure 5: Electric birefringence of a silver nanowire suspension at 0.03 mM KCl a) as a function of particle concentration and b) as a function of the square of the eld amplitude, for the indicated eld frequencies. In a) E0 is 2 V/mm and in b) the particle concentration is 100 mg/L.

9

ACS Paragon Plus Environment

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 23

Electric birefringence dynamics In Figure 6a, the electric birefringence signal as a function of time is presented normalised by the stationary value, ∆nst , to make evident that the dynamics of the process is independent of the eld frequency. The decay of the birefringence after the eld is turned o is due to thermal agitation, and hence it is related only to the rotational diusion coecient Θ of the wires. Due to the strong dependence of Θ with the particle dimensions, the analysis of the birefringence decay has been long used to determine the size of non-spherical particles in suspension. 15,17,20,3740 (b)

(a)

Figure 6: a) Normalised birefringence of the AgNWs in 0.03 mM KCl solution, as a function of time. b) Length distribution by volume of the AgNW sample as obtained from microscopy pictures (bars) and from the multi-exponential method (line). For polydisperse samples, the birefringence decay can be obtained as a superposition of independent single-exponential processes, as 40,41

∆n(t) = ∆nst

N X

Ci e−t/τi

(5)

i=1

where τi = 1/6Θi the characteristic rotation time of particles of a well-dened size, and

Ci are coecients which measure the contribution of each population to the overall signal. Since electric birefringence is proportional to the particle volume, the Ci 's provide a size distribution in volume of the polydisperse suspension. The rotational diusion coecient of 10

ACS Paragon Plus Environment

Page 11 of 23 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

each population can be calculated from the wire dimensions, with the expressions: 42

Θi =

3kB T F (ri ) πη(2ai )3

F (ri ) = log ri − 0.2/ log 2ri − 16/ (log 2ri )

(6) 2

being kB the Boltzmann constant, T the temperature of the suspension, ai the major semiaxis of the particles in population i, ri = ai /b their aspect ratio and η the solvent viscosity. The particle radius b is considered constant for all populations. In this manner, the length distribution of the AgNW sample, shown in Figure 6b, was obtained from the analysis of the birefringence decay. 40,41 The results are compared to electron microscopy determinations, with a very good agreement, indicating that particles are well dispersed in the solution. The volume-averaged wire length obtained from the birefringence decay is 4.3±1.7 µm, within a 10% dierence of the microscopy value, 4.5±2.9 µm. Nevertheless, it can be observed that the EB-obtained distribution is slightly narrower, which can be associated to the diculties of size determination from the microscope pictures in such a long particle system.

Electric birefringence spectra In Figure 7a we show the electric birefringence spectra of AgNWs, measured inside the Kerr regime, at three dierent values of the salt concentration. With the aid of Equation 2, the orientational order parameter S can be experimentally determined. Two contributions are expected to the electro-orientation. On one hand, the external electric eld induces an electric dipole that produces an orientation S1 given by:

S1 =

∆αe (ν)E02 30kB T

(7)

where ∆αe = αae − αbe the anisotropy of the electrical polarisability of the particles at a given eld frequency ν .

11

ACS Paragon Plus Environment

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 12 of 23

(a)

(b)

Figure 7: a) Electric birefringence of AgNW samples with the indicated salt concentrations, as a function of the eld frequency. The eld strength is 2 V/mm and the particle concentration 100 mg/L. The points are the experimental data and the dotted lines ttings to the model for conducting wires. b) Same as a) but only for 0.03 mM KCl (points), with the electric (dotted grey line) and viscous (dashed red line) contributions given by the model, presented separately. The solid black line is the sum of the two eects. The polarisation of low-charged conducting particles has been studied elsewhere, 27,43 modelling them as ideally polarisable, a reasonable assumption for moderate potentials. Following this approach, it was shown that two mechanisms contribute to polarisation: The fastest one is the large dipole generated by electron migration inside the particles, opposite to the eld. The slowest mechanism is related to the subsequent redistribution of ions in the electrolyte solution, which accumulate near the region of the particle with opposite charge. In the case of very elongated particles, it can be demonstrated that αbe ∼

1 r2

log(r)αae