Nano-Localized Heating Source for Photonics and Plasmonics
Article first published online: 25 AUG 2013
© 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Advanced Optical Materials
Volume 1, Issue 12, pages 899–904, December 2013
How to Cite
De Sio, L., Placido, T., Serak, S., Comparelli, R., Tamborra, M., Tabiryan, N., Curri, M. L., Bartolino, R., Umeton, C. and Bunning, T. (2013), Nano-Localized Heating Source for Photonics and Plasmonics. Advanced Optical Materials, 1: 899–904. doi: 10.1002/adom.201300275
- Issue published online: 11 DEC 2013
- Article first published online: 25 AUG 2013
- Manuscript Revised: 1 AUG 2013
- Manuscript Received: 2 JUL 2013
- U.S. Air Force Office of Scientific Research (USAFOSR), Air Force Material Command (AFMC), U.S. Air Force. Grant Number: FA8655–12–1–003
- METACHEM. Grant Number: CP-FP 228762–2
- PRIN Anno. Grant Number: 2010–2011 – prot. 2010C4R8M8
- soft matter;
- liquid crystals;
Nanomaterials offer unique properties that bridge classic to quantum behavior of matter. Plasmonic metallic nanoparticles (NPs) are a particular class of nanomaterials which possess the capability to localize light down to the nanoscale: visible electromagnetic radiation induces an oscillation of the free electrons localized at the metal (NPs)/dielectric (surrounding medium) interface. This phenomenon, called localized plasmon resonance (LPR), can be controlled in frequency by varying both the size and the shape of the nanoparticles and the dielectric constant of the surrounding medium. Due to the electron-electron scattering and the electron-phonon coupling associated with the LPR effect, the strong electric field generated around the NPs is converted into heat and the NPs behave as nano-sources of heat.[2, 3] In the last ten years, this resonant effect has enabled a remarkable breakthrough in fighting cancer through the exploitation of nanoshells, particles consisting of a dielectric core covered with a thin metallic shell. These can be delivered to tumor cells, and the efficient conversion of NIR light to heat opens up a new “drug-free” cancer therapy, called “plasmonic photo-thermal therapy” (PPTT).[5, 6] In cancer therapy applications, NPs are very useful agents for photo-thermal therapy due to an enhanced absorption cross section (four to five orders of magnitude larger) compared to conventional photoabsorbing dyes;[7, 8] this strong absorption ensures light effectiveness at lower radiation energy levels, thus rendering the therapy method minimally invasive. Moreover, NPs have higher photostability and do not suffer from photobleaching. In building a bridge between biosystems and nanomaterials, it is important to keep the LPR of NPs close to the water transparency window (700–900 nm), where tissue absorption is low and, thus, the penetration depth of radiation is high. In this framework, gold nanorods (GNRs) are a particularly interesting class of plasmonic nanomaterials; they exhibit two (transverse and longitudinal) LPRs, which are tunable from the visible to NIR, depending on the nanorod shape.[10, 11] In addition, the longitudinal LPR exhibits a very high sensitivity to variations in the refractive index of the surrounding medium. Thanks to these properties, GNRs have been widely exploited in bio-applications, sensing and monitoring. Investigation of the heat transport mechanism, from the heated GNRs to their surrounding medium, is a necessary step in realizing nanoscale heat sources for applications in nanotechnology and thermal-based therapies. In particular, monitoring temperature variations under optical illumination is an important issue not only for nanomedicine but also for photonics and plasmonics. To this end, many techniques have been developed so far, such as embedding the NPs in ice, employing suspensions of NPs in organic solvents, or using a complex made of semiconductors and NPs. Very recently, a breakthrough idea has been reported by Kim et al. which is based on silk inverse opal photonic crystals doped with NPs. In this framework, it has been demonstrated an enhancement of NPs absorption at the band-edge frequency of the photonic crystal, by means of a direct measurement of the NPs local heating, performed by using a thermal-camera. Despite obtained results appear quite exciting, the possibility of employing “smart materials” to study the heating process at the nanoscale represents a new horizon for nanomaterial based applications. Thermotropic cholesteric liquid crystals (CLCs) are short pitch (P) chiral materials which organize in layers with no positional ordering of the molecules within each layer, but exhibiting a mesoscopic order, described by a helicoidal orientation of the director axis from layer to layer. Due to the helix periodicity, CLCs behave as one-dimensional photonic bandgap systems where propagation of light of given wavelengths and polarization states is forbidden. It is worth noting that the spectral characteristics (width and position) of the photonic bandgap associated with the CLC helix can be controlled by applying an electric field, or changing the system temperature, or exposing the sample to a suitable optical radiation. In this paper, we report on the possibility of combining GNRs properties and CLC self-organization for simultaneously controlling the selective reflection of CLCs and the spectral position of the LPR of GNRs; both effects are obtained by exploiting the local heating induced by an external NIR light source. Compared to previously exploited techniques[15-17] the reported method represents an innovative non-invasive tool, wherein the properties of well known materials like CLCs are used for continuously monitoring photoinduced temperature variations around GNRs with high sensitivity. Cetyltrimethylammonium bromide (CTAB) capped, water dispersible, GNRs have been synthesized and subsequently transferred in chloroform, in order to obtain a dispersing medium with liquid crystals (LCs). The general protocol for seed-mediated synthesis of GNRs and their transfer from water to chloroform is described in details in the Supporting Information 1. Transmission electron microscopy (TEM, by Jeol JEM-1011 microscope, operating at 100 kV) analyses have been performed by depositing one droplet of different aqueous GNRs dispersions onto a carbon-coated copper grid, and then allowing the aqueous solvent to evaporate. For a statistical determination of the average GNRs size, shape and aspect ratio, at least 200 objects have been counted for each investigated sample.
The TEM image of Figure 1a indicates that the particle population consists mainly of GNRs with a 2.3 ± 0.3 (see Supporting Information 1) aspect ratio (AR). All GNRs have been successfully extracted into the chloroform by exploiting decanoic acid. This molecule is characterized by a carboxylic terminal group, able to electrostatically bind, once deprotonated, to the CTAB bilayer, which confers a positive charge to the GNR surface. At the same time, the alkyl chain provides dispersibility in organic solvent. The TEM image of GNRs deposited from chloroform, shown in Figure 1b, indicates that no change in shape and size of particles (2.4 ± 0.3 AR) occurred upon the transfer in organic solvent. In Figure 1c, the picture shows the result of the transfer of GNRs from water (in the upper part of the biphasic solution) to chloroform, where the presence of GNRs in the lower, blue coloured, phase is evident. Normalized UV-Vis absorption spectra of GNRs before and after the water to chloroform transfer are shown in Figure 1d. The GNRs dispersion exhibits two typical plasmon bands: a transverse one at 520 nm and a longitudinal one at 674 nm (Figure 1d, blue line). The spectral features are retained after the transfer of GNRs into the chloroform phase, although the peak wavelength of the longitudinal band is red-shifted. This effect can be explained by taking into account that optical properties of ellipsoidal particles are predicted in the framework of the Gans theory, through the expression of the extinction cross section:
where V is the volume of the particle, λ is the wavelength of light, εm is the dielectric constant of the surrounding medium, ε1 and ε2 are the real and imaginary parts of gold dielectric constant, respectively, while the depolarization factors Pj are defined by:(2)
where a, b, and c refer to the dimensions of GNRs along their three axes (typically, a > b = c), while e = (1-(1/AR)2)1/2, and AR is the aspect ratio of the GNRs. Based on this theory, for small and isolated GNRs, the spectral position of LPRs peaks depend on the refractive index of the surrounding medium, according to the condition that minimizes the denominator of Equation (1):
Based on the above expression (Equation 3), a modification in the dielectric behavior of the host material can induce a resonance along each coordinate axis; in any case, however, if εm increases, the resonance condition is fulfilled for higher absolute values of ε1. In fact, it is well known that, in the visible range, the absolute value of the real part of the electric permittivity of Gold NPs increases with λ; therefore, fulfillment of Equation(3) takes place for higher values of λ. In our case, due to the variation of the refractive index of the surrounding medium from 1.33 (water) to 1.44 (chloroform) the longitudinal plasmon band of the chloroform GNRs solution (Figure 1d, red curve) is red-shifted of about 15 nm with respect to the water-solution of the same GNRs (Figure 1d, blue line). It is worth noting that, due to its low sensitivity to refractive index variations, the transverse band (520 nm) does not exhibit any shift; thus, this behaviour underlines a high sensitivity of the longitudinal plasmon band to variation of the refractive index of the surrounding medium. We stress out that the general protocols for GNRs synthesis typically lead to 80–90% of GNRs and a certain percentage of NPs with different shape (10–20%); despite all efforts reported in literature, residuals of particles with different shapes (spheres, cubes, prism and unshaped) are always present. Also in our case, a small percentage of cubic-shaped NPs is present; in this case, electrons can oscillate both along the cube side and the cube diagonal, thus leading to a plasmon band split. However, since amplitudes of such oscillations are very similar also in the spectrum of the water solution of NPs, such two plasmon bands are convoluted in the band centered at 522 nm (Figure 1d, blue line) which is indeed asymmetrical, with a tail in the red region. The change of the surrounding refractive index, when moving from water to chloroform dispersion, can induce a slight shift of the diagonal plasmon band of cubic NPs, along with that of the longitudinal plasmon band of GNRs, thus leading to the shoulder (at 540 nm) observed in the spectrum of the chloroform dispersed sample (Figure 1d, red line)
The CLC has been prepared by twisting a nematic LC (MDA-00–1444, by Licristal) by inclusion of 20% of a chiral agent (ZLI-811, by Licristal), obtaining a helix pitch of about 400 nm. Then, we added the highest concentration of GNRs (10 wt%) which allows to obtain a homogeneous mixture, and we observed that the presence of GNRs red-shifted the typical reflection band of the CLC into the NIR (a detailed characterization of this effect is reported in the Supporting Information 2). Therefore, additional chiral agent (7% wt) has been added to adjust the reflection band back into the visible range (peak of the reflection band ≈ 526 nm). This is necessary to avoid the overlap between the CLC reflection and the GNRs absorption bands; in this way, in our system, the photonic and plasmonic properties are very well separated. Furthermore, in these conditions, the structural effect of the CLC cannot affect the GNRs by means of photonic-band edge.
100 μm thick glass cells have been treated with a thin polyimide layer, where induction of a preferred planar LC direction was obtained via rubbing, and have been then filled at room temperature by capillary action.
Figure 2a is a polarized optical microscope (POM) view of a sample, where the presence of structural defects, called oily streaks, is evident. In a flat cell, with layers parallel to the glass slabs, oily streaks appear as long bands that separate ideal domains of the flat layers, the inner structure of these domains being quite complicated and depending on many parameters, such as elastic constants, surface anchoring, cell thickness and GNRs concentration (see Supporting Information 2). Oily streaks are markers of a lamellar phase and confirm the existence of a Grandjean-Cano texture of the CLC phase (CLC helix axis oriented perpendicularly to the glass surfaces). In order to study the effects produced by the presence of GNRs, their distribution within the CLC was analyzed by performing an electron back-scattering diffraction (EBSD) characterization of the sample by means of a scanning electron microscope (SEM). This analysis is valuable to distinguish gold from other materials since the yield of backscattered electrons is proportional to specimen atomic number (Z); in our case, gold (Z = 79) produces a high contrast with the CLC surrounding (a carbon based system with Z∼15). As shown in the EBSD view of Figure 2c, the bright stripes indicate a presence of gold in the oily streaks only, as well as a preferred orientation of the GNRs axes along the streaks, as evident in the high magnification image (Figure 2d). It is also worth noting that, despite the circumstance that AR of observed particles remains almost unchanged, the average size (length and diameter) is increased of about 2–3 times, as a consequence of a self-assembly process induced by the confinement effect which is due to the presence of the CLC structural lines. In fact, the presence of a confined geometry can affect the electrostatic repulsion between two close GNRs (ligand layer avoid aggregation by means of electrostatic interaction) leading to a GNRs self-assembly.
It is well known that a circularly polarized light of the same handedness as the CLC helix and having wavelength between noP and neP (no and ne being the ordinary and extraordinary refractive indices of the material, respectively, and P being the CLC pitch), impinging at normal incidence on a CLC system in a planar configuration (helical axis perpendicular to the plane of the cell), is reflected by the CLC layer, while light of the opposite handedness propagates, unaffected, through the CLC. For unpolarized light in the wavelength range noP < λ < neP, an ideal sample reflects therefore 50% of the impinging intensity and transmits the remaining 50%, whereas the sample remains transparent to light whose wavelength is outside that range. The center of the reflection band occurs at:
where 〈n〉 = (ne + no)/2 is the average refractive index, while the bandwidth is given by:
It is worth noting that the exact value of the CLC pitch P is sensitive to all those factors (e.g. temperature) that can affect the balance of molecular interactions and the orientation of the CLC director.
Optical experiments have been performed to understand the influence, on the CLC configuration, of local heating induced by a suitable optical radiation, trough the GNR resonance. The all-optical setup, reported in Figure 3, utilizes a low power density (Pprobe = 0.1 W/cm2) CW probe laser, emitting at λ = 532 nm (in the center of the reflection band, λ0 ≈ 526 nm), a collimated and co-launched white source (450 nm < λ < 850 nm) for monitoring the spectral reflection properties of the CLC configuration, and a CW NIR pump laser emitting at λ = 680 nm (Ppump = 0.2 W/cm2) in the high absorption range of the GNRs (longitudinal band). The sample has been probed by the white light source, monitoring its back reflected components by means of a reflection fiber (F2).
Figure 4a reports the behavior of the CLC reflection band under illumination with the pump beam, for different exposure times. The CLC (inset Figure 3, resonant pump off) acts as a mirror for all the wavelengths within the reflection band of the impinging white light, which are back reflected. When optically pumping the same sample area, the photoexcitation of GNRs induces the formation of a heated electron gas that subsequently cools rapidly (ps process) by exchanging energy with the GNRs lattice; this process is followed by a phonon-photon interaction, where the GNRs lattice cools rapidly by exchanging energy with the surrounding medium. This local-heating induces an elongation of the CLC pitch (inset Figure 3, resonant pump on) with a consequent linear red-shift of the reflection band. In Figure 4a, it can be observed that, by keeping constant the pump power and increasing the exposure time, a linear red shift (more than 130 nm, Figure 4b) and a partial suppression of the reflection band take place, due to a gradual enhancement of the local temperature. Indeed, according to Equation (4), the central wavelength of the reflection band is directly proportional to the pitch P which increases with temperature; in addition, the elongation of P reduces the number of periods in the bulk of the cell, an effect which yields an attenuation and a broadening (Equation 5) of the reflection band amplitude with temperature.
To validate the effect of the GNRs-induced local heating on the CLC optical response, we have performed a control experiment by varying the sample temperature from 25 °C up to 95 °C and monitoring the reflection band behavior (Figure 4c): Once again, a linear red-shift is observed, which clearly confirms that the behavior reported in Figure 4a is due to a photo-thermal mechanism. It is worth noting that the two calibration functions reported in Figures 4b and d exhibit, within the experimental error, the same linear behavior; this result shows that it is possible to measure the temperature around GNRs at a given illumination time with a sensitivity of about 0.03 °C. We point out that, in order to test the temperature stability of the ligand layer (CTAB, in our case), several temperature cycles (from 25 °C up to 100 °C) have been performed, both on GNRs (water and chloroform dispersed) solutions and GNRs mixed with CLC. A spectral analysis of our samples, performed after each single cycle, shows that no significant variation in the spectral features can be observed, thus confirming the quite good thermo-stability of the ligand layer in the used temperature range.
An interesting aspect is represented by the possibility of using the heating induced variations in the local refractive index of the CLC for controlling the plasmonic response of GNRs.
Figure 5a shows the absorption spectra of the sample for two orthogonal polarizations of the impinging probe white light: Both polarizations (red and blue curve) exhibit a photonic absorption peak at λ = 520 nm, which is due to the selective reflection of the CLC configuration, while only the spectrum of light polarized along the y direction (see Figure 2d) exhibits a secondary peak at λ = 680 nm (blue curve), due to the longitudinal LPR of GNRs. This is a clear indication that GNRs have a preferential organization along the y direction, thus validating the SEM analysis reported in Figure 2d.
In order to check the influence of local heating on the plasmonic proprieties of the sample, we have observed its spectral response for different values of the pump light illumination time; results are reported in Figure 5b. A red-shift of the reflection band from 512 nm to 578 nm is well evident, which can be attributed to a local-heating induced variation of the CLC pitch. Furthermore, a red-shift of the LPR, from 685 nm to 731 nm is also observed. This behavior can be explained in terms of a local heating that leads to a broadening of the oily streaks and to a dissolution of the GNR self-assembly process. Indeed, single, small-sized, NPs, when excited by an incoming radiation, experience the same phase of the electromagnetic wave on their whole area, acting as simple dipoles; the width of the plasmonic peak is quite narrow and, depending on the particle size, it is centered in the green–red part of the electromagnetic spectrum. On the contrary, different areas of NPs with sizes comparable with the impinging radiation wavelength, experience different phase of the wave; thus, quadrupolar, octupolar and even higher oscillation modes are induced, whose excitation frequencies are higher than the dipolar one. As a consequence, a macroscopic broadening of the plasmon peak (with its center in the blue region of the electromagnetic spectrum) occurs. The temperature dependent shift of the plasmonic resonance observed in our sample can therefore be attributed to a temperature induced transition from densely packed to mono-dispersed NPs. It is worth noting that in a sample under intense optical illumination, due to the induced transition of the CLC to the isotropic phase, the refractive index is decreased; according to Equation (3), this phenomenon induces a competitive effect on the red-shift of the LPR, and reduces the tuning range.[33, 34] We have reported the CLC refractive index behaviour versus temperature in the Supporting Information 2.
Finally, to check the reversibility and repeatability of the observed effects, we have performed reflection dynamic experiments by using the green probe light of the setup reported in Figure 3 (λ = 532 nm, in the middle of the reflection band). By means of the photodiode PD2 we have monitored the reflection component while turning the pump laser on and off. Due to the induced local-heating, the CLC helix configuration is gradually destroyed and the intensity of the reflected component is reduced to zero; when the pump laser is turned off, a CLC self-organization takes place and the intensity of the reflected component is restored to its initial value (Figure 5c,d). We point out that a strong correlation has been observed between the turn off times and the polarization of the pump laser. Indeed, for light polarized along the y direction (Figure 5c) the local heating is enhanced (due to the high extinction cross section of the longitudinal LPR), with a corresponding turn-off time of ≈5 s, while, light polarized along x exhibits a quite long turn-off time (≈30 s), due to the low extinction cross section of the transversal LPR. In both cases, however, the turn-on time is about 2 s.
In conclusion, we have reported on the realization and characterization of an innovative method for simultaneously achieving optical control of both the selective reflection of a CLC and the plasmonic resonance of GNRs. In particular, if surface chemistry functionalization is used for inducing planar alignment of the LC director in a glass cell, the obtained self-organization process of CLCs is able to effectively induce confinement and topological organization of GNRs along the sample structural defects. A photo-thermal effect, induced by the presence of a NIR selective plasmonic resonance, can be exploited for controlling the position of the selective reflection exhibited by the CLC configuration. Furthermore, the local heating induces a variation of the surrounding medium refractive index, with a consequent shift of the plasmonic resonance. This synergy between plasmonics and photonics can be profitably used for building up a method for detecting the temperature around NPs under optical illumination. In our opinion, the developed method opens up prospects for realizing, in the future, a nanomedicine tool, useful to avoid invasive treatments (e.g. hypothermic effects) on patients during therapy treatments.
Authors are grateful to: Dr. Giovanni Desiderio for performing the SEM analysis and Dr. Vincenzo Caligiuri for drafting the experimental setup reported in Figure 3. The research leading to reported results has received funding from: The U.S. Air Force Office of Scientific Research (USAFOSR), Air Force Material Command (AFMC), U.S. Air Force, under grant FA8655–12–1–003 (P.I. L. De Sio), by the EC-funded project METACHEM (Grant CP-FP 228762–2) and by PRIN Anno 2010–2011 – prot. 2010C4R8M8.
- 19Electro-optical and Magneto-optical Properties of Liquid Crystals, Wiley, Chichester 1983, p.212.,
- 20The Physics of Liquid Crystals, Oxford Science Publications, Oxford 1993., ,
- 31Optical Properties of Metal Clusters, Springer-Verlag, Berlin 1996., ,
As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.
Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.