Quantifying Remote Heating from Propagating ... - ACS Publications

Subscriber access provided by University of Pennsylvania Libraries

Communication

Quantifying Remote Heating from Propagating Surface Plasmon Polaritons Charlotte I. Evans, Pavlo Zolotavin, Alessandro Alabastri, Jian Yang, Peter Nordlander, and Douglas Natelson Nano Lett., Just Accepted Manuscript • DOI: 10.1021/acs.nanolett.7b02524 • Publication Date (Web): 10 Aug 2017 Downloaded from http://pubs.acs.org on August 11, 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.

Nano Letters 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 20

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

Nano Letters

Quantifying Remote Heating from Propagating Surface Plasmon Polaritons Charlotte I. Evans,† Pavlo Zolotavin,† Alessandro Alabastri,† Jian Yang,† Peter Nordlander,†,‡,¶ and Douglas Natelson∗,†,‡,¶ †Department of Physics and Astronomy, Rice University, 6100 Main St. Houston, Texas 77005, United States ‡Department of Electrical and Computer Engineering, Rice University, 6100 Main St. Houston, Texas 77005, United States ¶Department of Materials Science and NanoEngineering, Rice University, 6100 Main St. Houston, Texas 77005, United States E-mail: [email protected]

Abstract We report a method to electrically detect heating from excitation of propagating surface plasmon polaritons (SPP). The coupling between SPP and a continuous wave laser beam is realized through lithographically defined gratings in the electrodes of thin film gold "bow tie" nanodevices. The propagating SPPs allow remote coupling of optical energy into a nanowire constriction. Heating of the constriction is detectable through changes in the device conductance, and contains contributions from both thermal diffusion of heat generated at the grating and heat generated locally at the constriction by plasmon dissipation. We quantify these contributions through computational modeling and demonstrate that the propagation of SPPs provides the dominant contribution. Coupling optical energy into the constriction via propagating SPPs in this geometry

1

ACS Paragon Plus Environment

Nano Letters

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

produces an inferred temperature rise of the constriction a factor of 60 smaller than would take place if optical energy were introduced via directly illuminating the constriction. The grating approach provides a path for remote excitation of nanoconstrictions using SPPs for measurements that usually require direct laser illumination, such as surface-enhanced Raman spectroscopy.

Keywords plasmonics, surface plasmon polariton (SPP), nanowire, grating, thermoplasmonics, bolometric detection, heating Metallic nanojunctions are unique nanostructured probes that allow for simultaneous surface enhanced Raman spectroscopy (SERS) and electronic transport measurements with single-molecule sensitivity. 1–3 Optical excitation of plasmonic resonances induces local heating. 4 The heating effects are particularly pronounced for isolated plasmonic nanoparticles used in SERS. 5 A local temperature increase as large as 100 K was reported for plasmonic nanostructures at low substrate temperatures due to the reduced thermal conductance and increased thermal boundary resistance. 6 Such localized laser heating in these nanostructures is detrimental if one wants to combine optical spectroscopy with measurements that require cryogenic temperatures, such as inelastic electron tunneling spectroscopy (IETS). 7–11 One potential solution to this heating problem is to limit direct heating by remotely exciting plasmonic modes in the nanojunctions via propagating surface plasmon polaritons (SPP). Depending on the geometry of the nanostructure, the wavelength of the incident light, and the dielectric properties of the metal and surrounding materials, an incident photon can couple with a grating to achieve momentum-matching conditions to excite a propagating SPP. 12 Remote plasmon excitation via grating was used to increase the laser light focusing and improve spatial separation in the tip-enhanced Raman spectroscopy of surface molecules. 13 Remote plasmon excitation has also been induced to perform SERS of molecules adsorbed at distal positions along metallic nanowires. 14,15 Remote heating and coupling of 2


Page 2 of 20

Page 3 of 20

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

Nano Letters

optical energy into metallic point contacts has been achieved via SPP excitation by illuminating devices with specially designed gratings or wave guides. 16,17 The SPP propagation length can exceed tens of microns, 17–22 and is usually measured by designing a second antenna-type structure at a specified distance to convert SPP to a free photon for optical detection. 21,23–27 SPP excitation can be also detected via electronic transport measurements due to its effects on device conductance. 28–30 This electrical SPP detection technique is beneficial in eliminating the additional burden of optical detection but do not provide a quantitative assessment of the steady state heating contribution from the SPP dissipation. We report a study of the SPP propagation in thin gold films with a nanowire constriction, in which we quantify the local temperature increase in the constriction through bolometric changes in the device conductance. The SPPs are launched using gratings fabricated at various distances from the constriction. The intact nanoscale constrictions examined here are not in the atomic-scale or tunneling regimes, allowing for less variation between samples, 17 and do not result in photon emission. This bolometric detection approach presents an opportunity to infer the contributions to the steady state heating of the constriction from both local SPP dissipation and thermal diffusion of heat generated at the grating without the complications of tunneling physics, such as photo-assisted tunneling. Previous works 17,29,31 study in detail the laser-driven local heating effects and thermal expansion effects of SPPs in atomic-scale limit. By studying the polarization, grating distance, and laser position dependence of the constriction’s conductance in the steady state regime and comparing with numerical simulations, we find that this approach transfers significant optical energy to the constriction via SPPs while lessening the constriction’s temperature increase by up to a factor of 60 compared to direct illumination. These measurements show that remote excitation of SPPs with optimized thermal design of the electrodes is a promising path toward combining excitation of plasmonic modes in nanojunctions and measurements requiring cryogenic temperatures, such as IETS. The devices studied in this work are films of 30 nm Au/1 nm Ti with a plasmonically-

3


Nano Letters

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 ACS Paragon Plus Environment

Page 4 of 20

Page 5 of 20

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

Nano Letters

taken at timescales much longer than thermalization. 17 A change in total device conductance, ∆G, is measured using a two-terminal measurement under 5 mV bias and detected via lock-in amplifier, Fig. 1b, under chopped illumination of a 785 nm free space continuous wave laser, see Methods. As the constriction heats up, its conductance decreases. Therefore a measured negative value of total device ∆G implies a net temperature increase occurs at the constriction. All data presented here are acquired from devices on a Si with 2 µm SiO2 substrate at 40 K and with laser intensity of 117 kW/cm2 . Additional measurements involving different laser intensities, substrates, and grating parameters are discussed in the SI. Spatial maps of ∆G when the laser is raster scanned around the constriction in the transverse and longitudinal polarizations are demonstrated in Fig. 2a and 2b, respectively. The largest response in ∆G occurs when the constriction is directly illuminated by the transversely polarized light. The polarization dependence of ∆G, Fig. 2c, is consistent with heating of the constriction from the resonant excitation of the transverse plasmon mode in the constriction 6,32–34 in addition to the direct absorption of the metal. The data in Fig. 2c fits well with cos2 (θ) dependence, displayed by the red line, which confirms the dominant role of transverse plasmon contribution, in agreement with the previous reports. Spatial maps of total device ∆G when the laser is scanned around a grating located 8.8 µm from the constriction indicate that detectable heating of the constriction only occurs when the laser is in the longitudinal polarization and incident on the grating, Fig. 2d,e. The polarization dependence of total device ∆G when the laser is incident on the center of the grating, Fig. 2f, is consistent with the excitation of the longitudinal plasmon mode, propagating from the grating to the constriction. 17,27,35 This polarization dependence is rotated by ∼90◦ compared to that of Fig. 2c, confirming the dominant contribution from SPP excitation. The remote excitation of the constriction via the grating located 8.8 µm away results in a 4.5 fold decrease of in ∆G compared to that of direct illumination of the center of the nanowire.

5


Nano Letters


Page 6 of 20

Page 7 of 20

Nano Letters


Nano Letters

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

is illuminated, the largest temperature increase occurs at the grating, as the heat generated at the focal spot from the light absorption dissipates through the substrate and metal layer. This means that the temperature rise of the constriction measured when illuminating the grating has contributions from thermal diffusion from the illuminated grating and from SPP dissipation in the constriction. The matter is further complicated by the fact that resonance SPP excitation in the grating will increase the heat dissipation in the grating compared to direct absorption in a plain metal film of similar thickness. While experimental data indirectly suggests that the local temperature increase in the constriction due to thermal diffusion is negligible even for the closest of gratings, determining whether the main contribution of the temperature rise is from SPP excitation and not from enhanced absorption at the grating location is imperative if these devices are to be used for remote excitation applications. For this reason, we performed COMSOL Multiphysics simulations to determine the temperature rise of the constriction and the relative importance of its dominant contributions. A model was developed to determine ∆G with illumination incident on gratings at various distances from the constriction. The device geometry consisted of the constriction with one full electrode with grating, truncating the other after the fan-out electrode. A bias was applied across the device and the total current through the device was calculated with and without the laser present on the grating. The calculated ∆G values, seen in red in Fig. 3a, are in reasonable agreement with the experimentally measured ∆G. This same geometry was then used to calculate the temperature rise of the constriction due to SPP as described in Methods. 38,39 The total temperature rise of the constriction decreases as the distance to the grating is increased, Fig. 3b. By comparison with the simulation, which has very good agreement with the experimentally measured conductance changes, we can convert the experimentally measured conductance to the calculated constriction temperature rise. The remote excitation of the constriction using the closest grating located at 5.4 µm from the constriction decreased ∆T by a factor of 11 compared to that of direct illumination of the constriction. Remotely exciting the grating located at 13.4 µm from the constriction had

8


Page 8 of 20

Page 9 of 20

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

Nano Letters

a clearly detectable electrical response and resulted in factor of 60 decrease of ∆T at the constriction compared to direct illumination. The grating in the electrodes can increase the absorption of the laser illumination, thus increasing the heating at the grating. To quantify the SPP propagation contribution to the constriction temperature rise, we used a 3D model to calculate the device’s electric field spatial dependence with and without a grating at a distance of 7.2 µm from the constriction, Fig. S7. Without the presence of the grating, the majority of the field intensity occurs in the connecting electrode lead, decaying quickly before reaching the constriction. As expected, the addition of the grating changes the electric field distribution considerably. The field now propagates from the grating, through the constriction, and to the other side of the fan-out electrode as expected for SPP propagation. Using a closely related model described in the SI, the heat source density is shown to oscillate as a function of position, Fig. S8b, which demonstrates that SPP dissipation is the dominant contribution to the heating. The heating contribution solely due to the SPP dissipation within the constriction, excluding thermal flux from the connecting electrodes, was calculated with and without a grating, Fig. 4. This calculation neglects any thermal diffusion due to direct absorption in the focal spot of the laser. On average, the temperature rise with the grating is an order of magnitude larger than without the grating, indicating that SPP propagation drives the temperature rise of the constriction. Although the magnitude of the temperature rise in Fig. 4 is smaller than that in Fig. 3b, the former is underestimated in magnitude in part because it neglects the heating contribution of SPPs in the immediately surrounding tapered electrodes, Fig. S8b. We then calculated the thermal transport in the device by incrementally decreasing the dissipation region acting as the heating source, from the entire device with the grating, to only the constriction, Fig. S8a. Excluding the heating at the site of the beam only slighly changes the constriction temperature. The majority of the heating contribution to the constriction occurs in the tapered region around the constriction, which further confirms that SPP dissipation provides a larger role in the constriction temperature rise than

9


Nano Letters


Page 10 of 20

Page 11 of 20

Nano Letters


Nano Letters

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

A calculation was conducted for two thicknesses of SiO2 , one peak and one dip value seen in Fig. 5a, to determine the electric field amplitude as a function of position with a 785 nm free space wavelength Gaussian beam incident on the gratings. Fig. 5b shows the electric field amplitude as a function of position with a substrate with 1.53 µm SiO2 , which is a maximum point in the oscillations in Fig. 5a. The electric field at the SiO2 /Au interface indicates SPP propagation with a decay length of ∼7 µm, which closely matches the decay length we find in Fig. 3. Fig. 5c shows the electric field amplitude as a function of position with a substrate with 1.63 µm SiO2 , which is a minimum point in Fig. 5a. The SPP contributions to the electric field damp out much more quickly than in the 1.53 µm SiO2 case. This indicates that even a change of 100 nm in SiO2 thickness can have a significant effect on the effectiveness of SPP excitation and propagation due to the Fabry-Perot effect. We also used the 2D model to verify that the parameters of the grating we used in experiments, Fig. 1b, are similar to the ones that optimize the coupling between laser light and SPP, Fig. S4. Gratings with more than two slits, Fig. S3, only slightly increased the signal at the constriction. In the face of increased fabrication complexity, we chose to retain the two slit geometry. Additional discussion, figures, and references about these points can be seen in the SI. In this work, we use electronic transport measurements to examine the contributions to heating of grating-excited plasmonic constrictions. The addition of gratings in the electrode design of gold "bow tie" structures offers a way to remotely excite the nanoconstriction while reducing additional contributions of heating from direct absorption. Heating predominantly from excitation of SPPs launched from gratings up to 13 µm away can be electronically detected via the differential conductance of the device and reduces the temperature rise of the constriction by as much as a factor of 60 compared to that of direct excitation. The reduction of local device heating is imperative for measurements that rely on cryogenic temperatures, such as single-molecule IETS. Remote excitation of the nanoconstriction allows for the majority of the laser heating to dissipate in the larger electrodes which could be used

12


Page 12 of 20

Page 13 of 20

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

Nano Letters

in applications such as SERS. Adding a nanogap in the constriction would allow for the outcoupling of photons which can be inelastically scattered by a molecule located in the gap. The addition of a nanogap would require consideration of conduction contributions from tunneling mechanisms, such as photo-assisted tunneling, 17,29,31 in addition to the bolometric technique described here for a more accurate temperature estimation at the gap. Grating placement requires some consideration, as closer gratings contribute to higher constriction temperatures but allow greater coupling to the local plasmonic modes of the junction, contributing to larger SERS signal. These experiments imply that with proper electrode design and thermal considerations, remote excitation of nanogaps via SPPs, similar to guided excitation of TERS, 13 should be possible with minimal temperature rise of the nanogap.

Methods Device Fabrication. Devices were fabricated on nearly degenerately-doped p-type (boron) Si wafers with 2 µm thermally grown oxide. Au/Ti contact pads and a common ground (50 nm Au with 5 nm Ti adhesion layer) were deposited via shadow mask evaporation. The finer features of the device were fabricated using single-step electron beam lithography. After development and 10 seconds of oxygen plasma cleaning to reduce PMMA (poly(methyl(methacrylate)) residue, a final layer of Au/Ti (30/1 nm) was deposited using electron beam evaporation. The grating consists of two rectangular holes (width: 300 nm, length: 8 µm) spaced 0.5 µm apart that go through the entire gold film. Each chip had 24 devices in total. The chip was mounted to the chip carrier using Apiezon N thermal grease and the devices and common ground were wire-bonded to the carrier using 25 µm gold wires. The average device resistance was 47 Ω. The linear regime of the temperature dependence of the resistance is described by the equation R(T ) = (0.0997 ± 0.002)T + (46.79 ± 0.01)Ω. The gratings did not cause big changes to the total device conductance as the grating locations are far from the 100 nm nanowire constriction and are in the 10 µm wide strips surrounded

13


Nano Letters

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

by additional metal. The differential conductance of the nanowire was measured using lock-in amplifier under focused illumination of a 785 nm CW diode laser with maximum intensity 400 kW/cm2 and under high vacuum (

Quantifying Remote Heating from Propagating ... - ACS Publications

Recommend Documents