Microscopic origin, unbiased measurement, and predictive control of grain-boundary scattering losses of surface plasmon polaritons in polycrystalline gold films

Surface plasmon polaritons (SPPs) propagating on evaporated or sputtered gold films are damped not only by the metal's intrinsic absorption but by extrinsic scattering from surface roughness and from the grain boundaries of the polycrystalline film. A cathodoluminescence imaging study (Kuttge et al., 2008) separated these channels on gold and introduced a purely phenomenological grain-boundary loss term, a scattering rate $1/L_G = S_G/d$ with grain size $d$ and a dimensionless fitted coefficient $S_G \approx 0.2\%$ of unknown microscopic origin. The open problem is to replace this phenomenology with a quantitative, predictive, microscopically grounded account of SPP loss in polycrystalline metal films: derive $S_G$ from boundary structure (groove geometry, misorientation, impurities), establish its frequency dependence and universality, and enable the supporting measurements — treating the leaky $\lambda \lesssim 600\ \mathrm{nm}$ spectral region excluded by the original analysis, correcting the extraction of propagation lengths for angle-dependent grating outcoupling, and deconvolving grain-boundary, void, and electron mean-free-path contributions to the film dielectric functions into processing–property relations. No microscopic theory of $S_G$, and no re-analysis of the original extraction, is available to date, so the core remains open.

Background

A surface plasmon polariton (SPP) is an electromagnetic surface wave bound to the interface between a metal and a dielectric, with fields confined to a few hundred nanometers on either side; it is the elementary excitation of plasmonics, underpinning nanoscale waveguiding, sensing, and metasurface optics. A key figure of merit is the propagation length $L_{\mathrm{SPP}}$, the distance over which the SPP intensity decays by $1/e$. In noble-metal films grown by evaporation or sputtering — the standard fabrication for practical plasmonics — $L_{\mathrm{SPP}}$ is limited by intrinsic ohmic damping (fixed by the metal's dielectric function) and by extrinsic scattering from surface roughness and from grain boundaries, because such films are polycrystalline: they consist of crystallites tens of nanometers across, so a propagating SPP crosses many internal interfaces where crystal orientation changes and shallow grooves often form at the surface. Prior context: Aspnes, Kinsbron, and Bacon (1980) showed that measured dielectric functions of thin gold films vary strongly with deposition conditions and attributed this to reduction of the electron mean free path by grain boundaries, voids, and roughness; Sambles (1984) had connected grain-boundary scattering with surface-plasmon attenuation in attenuated-total-reflection (ATR, a prism-coupling technique) measurements on noble-metal films, but without a quantitative propagation-based decomposition. Cathodoluminescence (CL) imaging spectroscopy, developed by the same AMOLF group, provides a direct local probe: a focused electron beam excites SPPs at a chosen point on the film, the SPP propagates to a lithographically defined grating that converts ('outcouples') it to free-space photons, and the emitted intensity measured as a function of the excitation–grating distance yields $L_{\mathrm{SPP}}(\lambda)$ spectrally resolved. Kuttge et al. (Appl. Phys. Lett. 93, 113110 (2008)) applied this to gold films of different deposition conditions, modeled the intrinsic loss from measured dielectric functions and the roughness loss from film topography measured by atomic force microscopy (AFM) plus established roughness-scattering theory, and attributed the residual to grain boundaries via a simple phenomenological model: if $d \ll L_G$, grain-boundary scattering adds a loss channel with scattering rate $1/L_G = S_G/d$ (an effective grain-boundary length $L_G = d/S_G$), and fitting yielded a dimensionless $S_G \approx 0.2\%$ for their films. The authors stated that, to their knowledge, no quantitative studies of the effect of grain-boundary scattering on SPP propagation had been published. The analysis deliberately excluded $\lambda \lesssim 600\ \mathrm{nm}$, where the SPP normal wavevector component $k_z$ grows, the mode is no longer purely bound (radiative loss into the collected light cone), and the extraction equation does not account for this radiative channel; and it assumed an angle-independent grating outcoupling efficiency $\alpha(\lambda,\phi)=1$ when converting CL intensity to propagation length.

Problem Statement

Provide a quantitative, predictive, microscopically grounded account of surface plasmon polariton propagation loss in polycrystalline gold films, replacing the phenomenological grain-boundary term of Kuttge et al. (2008) with derived physics, and enabling its unbiased measurement. The common objective — predictive control of SPP loss at real polycrystalline metal–dielectric interfaces — centers on a core theoretical component, accompanied by supporting measurement corrections and materials context identified in the original analysis. (1) Derive the grain-boundary scattering coefficient microscopically: starting from boundary structure (grain-boundary grooves at the surface, crystallographic misorientation statistics, voids, and chemical impurity segregation), compute $S_G$ or an equivalent grain-boundary loss measure (any equivalent measure must reduce to the grain-boundary loss rate $S_G/d$ in the regime $d \ll L_G$), its frequency dependence across the visible and near-infrared, and, on the basis of that derivation, whether a universal coefficient $S_G$ of this $1/L_G = S_G/d$ form holds across films, deposition methods, and metals, or must be replaced by a structure-dependent quantity. (2) Extend the cathodoluminescence-based extraction to the previously excluded leaky regime $\lambda \lesssim 600\ \mathrm{nm}$ on gold, where the growing normal wavevector component $k_z$ makes the SPP only weakly bound and radiative loss enters the collected signal, by incorporating the radiative channel into the intensity-to-propagation-length model and treating that spectral region quantitatively. (3) Quantify how the angle-dependent grating outcoupling efficiency $\alpha(\lambda,\phi)$ and the full angular distribution of SPP propagation bias the decay lengths extracted under the assumption $\alpha(\lambda,\phi)=1$, and provide a corrected extraction procedure. (4) Deconvolve the separate contributions of grain boundaries, voids, and electron mean-free-path reduction to the sample-to-sample variation of the measured dielectric functions, toward predictive processing–property relations for SPP propagation in polycrystalline metal films. The core of the open problem is component (1): a microscopic, predictive derivation of the grain-boundary scattering coefficient, its frequency dependence, and a verdict on whether a universal coefficient of this $1/L_G = S_G/d$ form holds. A complete answer must deliver component (1) together with the measurement corrections of components (2) and (3), identified as limitations of the original analysis and needed to obtain the unbiased $L_{\mathrm{SPP}}(\lambda)$ and $S_G$ against which such a derived account can be tested; component (4) is related materials context — the sample-to-sample variation of the measured dielectric functions attributed qualitatively to grain boundaries, voids, and mean-free-path reduction — rather than an additional mandatory deliverable.

References

1. M. Kuttge, E. J. R. Vesseur, J. Verhoeven, H. J. Lezec, H. A. Atwater, A. Polman, 'Loss mechanisms of surface plasmon polaritons on gold probed by cathodoluminescence imaging spectroscopy,' Applied Physics Letters 93, 113110 (2008). DOI: 10.1063/1.2987458, https://doi.org/10.1063/1.2987458
2. J. R. Sambles, 'Grain-boundary scattering and surface plasmon attenuation in noble metal films,' Solid State Communications 49, 343-345 (1984). DOI: 10.1016/0038-1098(84)90583-0, https://doi.org/10.1016/0038-1098(84)90583-0
3. D. E. Aspnes, E. Kinsbron, D. D. Bacon, 'Optical properties of Au: Sample effects,' Physical Review B 21, 3290-3299 (1980). DOI: 10.1103/PhysRevB.21.3290, https://doi.org/10.1103/PhysRevB.21.3290
4. H. S. Lee, C. Awada, S. Boutami, F. Charra, L. Douillard, R. Espiau de Lamaestre, 'Loss mechanisms of surface plasmon polaritons propagating on a smooth polycrystalline Cu surface,' Optics Express 20, 8974 (2012). DOI: 10.1364/OE.20.008974, https://doi.org/10.1364/OE.20.008974
5. T. Coenen, E. J. R. Vesseur, A. Polman, 'Angle-resolved cathodoluminescence spectroscopy,' Applied Physics Letters 99, 143103 (2011). DOI: 10.1063/1.3644985, https://doi.org/10.1063/1.3644985
6. X. Ma, M. Grüßer, R. Schuster, 'Angular Dependence of Cathodoluminescence of Linear and Circular Au Gratings: Imaging the Coupling Angles between Surface Plasmon Polaritons and Light,' The Journal of Physical Chemistry C 118, 23247-23255 (2014). DOI: 10.1021/jp506918p, https://doi.org/10.1021/jp506918p
7. C. I. Osorio, T. Coenen, B. J. M. Brenny, A. Polman, A. F. Koenderink, 'Angle-Resolved Cathodoluminescence Imaging Polarimetry,' ACS Photonics 3, 147-154 (2016). DOI: 10.1021/acsphotonics.5b00596, https://doi.org/10.1021/acsphotonics.5b00596
8. H. Saito, T. Sannomiya, 'Angle-resolved cathodoluminescence microscopy on plasmonic crystals,' Microscopy 75, 145-156 (2026). DOI: 10.1093/jmicro/dfag002, https://doi.org/10.1093/jmicro/dfag002
9. P. Nagpal, N. C. Lindquist, S.-H. Oh, D. J. Norris, 'Ultrasmooth Patterned Metals for Plasmonics and Metamaterials,' Science 325, 594-597 (2009). DOI: 10.1126/science.1174655, https://doi.org/10.1126/science.1174655
10. J. H. Park, P. Ambwani, M. Manno, N. C. Lindquist, P. Nagpal, S.-H. Oh, C. Leighton, D. J. Norris, 'Single-Crystalline Silver Films for Plasmonics,' Advanced Materials 24, 3988-3992 (2012). DOI: 10.1002/adma.201200812, https://doi.org/10.1002/adma.201200812
11. K. M. McPeak, S. V. Jayanti, S. J. P. Kress, S. Meyer, S. Iotti, A. Rossinelli, D. J. Norris, 'Plasmonic Films Can Easily Be Better: Rules and Recipes,' ACS Photonics 2, 326-333 (2015). DOI: 10.1021/ph5004237, https://doi.org/10.1021/ph5004237
12. S. H. Chew, A. Gliserin, S. Choi, X. T. Geng, S. Kim, W. Hwang, K. Baek, N. D. Anh, Y.-J. Kim, Y. M. Song, D. E. Kim, S.-Y. Jeong, S. Kim, 'Large-area grain-boundary-free copper films for plasmonics,' Applied Surface Science 521, 146377 (2020). DOI: 10.1016/j.apsusc.2020.146377, https://doi.org/10.1016/j.apsusc.2020.146377
13. N. Hiramatsu, F. Kusa, K. Imasaka, I. Morichika, A. Takegami, S. Ashihara, 'Propagation length of mid-infrared surface plasmon polaritons on gold: Impact of morphology change by thermal annealing,' Journal of Applied Physics 120, 173103 (2016). DOI: 10.1063/1.4966934, https://doi.org/10.1063/1.4966934
14. M. Prämassing, M. Liebtrau, H. J. Schill, S. Irsen, S. Linden, 'Interferometric near-field characterization of plasmonic slot waveguides in single- and poly-crystalline gold films,' Optics Express 28, 12998 (2020). DOI: 10.1364/OE.384629, https://doi.org/10.1364/OE.384629
15. P. Chhantyal, T. Birr, D. Hinrichs, U. Zywietz, D. Dorfs, B. Glasmacher, A. B. Evlyukhin, C. Reinhardt, 'Influence of Fabrication Methods of Gold and Silver Layers on Surface Plasmon Polaritons Propagation Length,' Plasmonics 13, 1359-1366 (2018). DOI: 10.1007/s11468-017-0639-5, https://doi.org/10.1007/s11468-017-0639-5