Nanoscale Imaging of Ultrafast Light Coupling to Self-Organized

Subscriber access provided by Nottingham Trent University

Article

Nanoscale Imaging of Ultrafast Light Coupling to Self-Organized Nanostructures Anthony Abou Saleh, Anton Rudenko, Ludovic Douillard, Florent Pigeon, Florence Garrelie, and Jean-Philippe Colombier ACS Photonics, Just Accepted Manuscript • DOI: 10.1021/acsphotonics.9b00702 • Publication Date (Web): 13 Aug 2019 Downloaded from pubs.acs.org on August 16, 2019

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 15 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

ACS Photonics

PEEM

1 pulse e

e 10 pulses

e

e

e

ACS Paragon Plus Environment

20 pulses

ACS Photonics 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 15

Nanoscale Imaging of Ultrafast Light Coupling to Self-Organized Nanostructures Anthony Abou Saleh,† Anton Rudenko,† Ludovic Douillard,‡ Florent Pigeon,† Florence Garrelie,† and Jean-Philippe Colombier∗,† †Univ Lyon, UJM-Saint-Etienne, CNRS, Institute of Optics Graduate School, Laboratoire Hubert Curien UMR CNRS 5516, F-42023 St-Etienne, France ‡CEA IRAMIS SPEC Service de Physique de l’Etat Condensé, UMR CNRS 3680 Université Paris Saclay F-91191 Gif sur Yvette, France E-mail: [email protected]

Abstract

Keywords

In the course of laser-induced surface selforganization, optical feedback mechanism is generally evoked as the main process driving the final rippled topography. To unravel the role of light and transient nanostructures in a multipulse irradiation regime, we use a high-resolved imaging technique, the photoemission electron microscopy. The pulse-to-pulse evolution of the inhomogeneous laser field distribution on a titanium surface nanostructured by a femtosecond laser is investigated at the nanoscale. Photoelectron imaging allows to separate the contributions of radiative and non-radiative scattered fields and to unveil the correlation between the nanocavity density and quasi-periodic light absorption. The multipulse experimental observations are supported by electromagnetic calculations showing that the absorption of light on evolving surface relief drives the selection of the final periodicity. The surface distribution of light absorption is influenced by surface topography that transiently adapts following the collective optical response of nanocavities, namely the reinforcement of the dipoledipole coupling as the nanocavity density increases. Mapping the evolution of the photoelectron emission sheds a new light on the intricate mechanisms controlling the laser-induced surface self-organization features.

Ultrafast laser nanostructuring, LIPSS, PEEM, local-field enhancement, nanocavities, coupleddipole model Illumination of surface nanostructures by laser light has recently received considerable interest as the advances of nanostructuring process pave the way towards optics and energy control on the nanoscale. Light-driven material arrangement and extreme field enhancement on structures rely on the interplay between material resonance and optical modes as shown in the case of coupled periodic apertures or laser-induced periodic surface structures (LIPSS). 1–5 Optical local field enhancement and resonant excitations enable electromagnetic field design, driving light-induced ordering from random surface conformations to anisotropically periodic self-arrangement of matter. 6–8 For non-metallic materials, the combination of nonlinear laser interaction and laserinduced material instabilities triggers the formation of self-assembled structures through optical feedback. 9 In addition, a system incorporating a strong disorder can block wavelike propagation and localize light, causing photons to be confined in the medium, a mechanism referred to as the Anderson localization of light. 10,11 Designing metallic nanostructures,


1

Page 3 of 15 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 Photonics

especially periodic ordering of random surface nanostructures by ultrafast laser, aims to concentrate local light field energy into specific locations emerging in the limit of closely coupled metallic structures manipulation. Random and periodically aligned patterns with precisely controlled shapes, spacings and sizes exhibit significant sensitivity to surface plasmon (SP) excitation. 12,13 However, the singular propagation of SP along one metallic surface initiated by a nanohole matrix remains questionable. Actually, a diversity of processes can be operative depending on the corresponding nanostructure features. 14–16 Consequently, a careful investigation is required to identify the way collective and interactive coupling of irradiated nanometric objects defines the final optical response. A focused laser beam exhibits non-trivial field distribution in the focal region depending on the beam and nanostructure features. 17–21 Surface irradiation in the sub-ablative regime can trigger self-organized arrangement of nanostructures, referred to as LIPSS, with periods ranging from λ down to λ/5n, where λ is the incident laser wavelength and n is the refractive index of the medium. 6–8,22 The role of optical feedback is central in LIPSS formation but difficult to access by optical means. 23 The periodic behavior of the topography is assigned to different mechanisms such as the coherent superposition of incident with scattered waves, the surface plasmon wave excitation, the near-field enhancement, and the corrugation-mediated feedback effects. 6,8,24 In the multi-pulse regime, it is generally accepted that LIPSS periodicity exhibits a decrease when the number of laser pulses impinging the target increases. 22,24 The reasons were attributed to surface plasmon resonance affected by surface roughness, supported by macroscopic approaches such as effective medium theory 25,26 and analytical expressions, taking account for roughness-induced modifications of surface plasmon vector. 27,28 Recently, the question of roughness influence on the absorbed energy distribution was revisited from a microscopic point of view, considering singular and collective response of nanocavitiesdipoles on metal surface. 13 However, the influence of surface roughness and nanocavity con-

centration on LIPSS periods has never been experimentally demonstrated, therefore, the proposed scenarios still require careful examination. In the present work, a better understanding of the feedback mechanism is pursued by uncloaking the role of the concentration and distribution of the local-field enhancing centers in reducing the LIPSS periodicity. We report on the pulse-to-pulse evolution of the local nearfield distribution. The surface material is patterned with self-organized patterns on length scales comparable to the wavelength of light allowing a very precise light coupling observation. This way, a set of periodic surface arrangements of matter of different periodicities induced by ultrashort laser pulses on titanium are investigated by photoemission electron microscopy (PEEM). PEEM is a full-field imaging technique relying on the collection of the photoemitted electrons excited by a laser light illumination. In the past few years, n-photon absorption (nP-) PEEM has been employed to conduct plasmonic investigations. 29–37 The experimental photoemission data are interpreted by comparison with electromagnetic simulations. This unveils the prominent role of the local surface curvature of random nanocavities triggering local field-enhancement superimposed to the expected periodic light absorption. The photoemission signal follows the same modulation as that as of the LIPSS periodicity. By correlating the LIPSS periodicity with that of the photoelectrons, we demonstrate that the concentration of the collective local-field enhancement centers plays a prominent role in reducing the LIPSS periodicity.

Results and discussion Experimental methods and LIPSS characterization Multiple laser pulses are necessary for the development of a stable LIPSS pattern within an ultrashort laser spot. By raising the double question of why and how this regulation applies, a set of ultrashort laser impacts on a Ti polycrystalline sample is performed as a function of the number of applied pulses by


2

ACS Photonics 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 15 Sc

3 µm E M

1 pulse

M

PL 10 pulses

λ/2 P

20 pulses

IR-fs

D

M

OL

(b)

e

UV Lamp

L

M Ti

(a)

(c)

(d)

Figure 1: (a) LIPSS formation experimental setup. P represents a polarizer, M a mirror, D optical densities, L a lens and Ti the titanium sample. (b) SEM images of LIPSS prior to the PEEM analysis structured with consecutively larger number of laser pulses starting with 1 pulse to 20 pulses. The electric field polarization is given by the double red arrow. (c) Plot of the LIPSS periodicity evolution as a function of the number of laser pulses. The periodicity dispersion is indicated by bars (d) LIPSS-PEEM characterization setup. OL represents the objective lens, PL the projective lenses and SC the screen. the experimental set-up displayed in Figure 1a. This set presents ripple patterns of different periodicities obtained at an exciting central wavelength of λ = 800 nm, under normal incidence and a fixed peak fluence of F = 0.24 J/cm2 below the single-pulse ablation threshold. Figure 1b presents scanning electron microscopy (SEM) images of 3 different sites corresponding to low (1 pulse), medium (10 pulses) and high (20 pulses) dose of laser irradiation. After one laser shot, SEM image shows a laser-induced surface nanoroughness distributed sparsely and randomly. At 10 pulses, a microscale periodic pattern of Λ = 706 nm is marked over the initially induced random nanoroughness. At this step, we observe only small patches of periodic ripples in dispersed regions within the irradiated spot. Increasing the number of laser pulses brings LIPSS to grow and coalesce locally, leading to a clear extended ripple pattern with a period of about Λ = 640 nm for 20 laser shots. The evolution of the LIPSS periodicity as a function of the number of laser pulses is extracted by a 2D Fast Fourier Transform 2D-FFT analysis and plotted in Figure 1c. As

already reported, LIPSS periodicity decreases when increasing the number of laser pulses and the final periodicity saturates for applied number of laser pulses larger than 20. To obtain a relative measurement of the field strength while the number of applied pulses increases, the sample is analyzed by PEEM to map the local photoemission yield. The schematic representation of the PEEM-LIPSS characterization is provided in Figure 1d. In a PEEM image, the brightness of one imaged zone is directly proportional to the number of the electrons photoemitted from that zone. Thus acquiring the physical mechanisms behind electron photoemission is of high importance regarding data interpretation. Upon illumination of the Ti target with a UV lamp, electrons are linearly photoemitted since one photon (1P) absorption events produce electrons of energy exceeding the work function of Ti (4.33 eV 38 ). Under IR laser illumination at a central wavelength of λ = 800 nm (1.55 eV) and polarization perpendicular to the LIPSS orientation, overcoming the Ti work function requires a multiphoton absorption. In our case,


3

Page 5 of 15 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 Photonics

a three photons (3P) absorption process as the lowest order is expected to dominate. Please note that the photoemission signal gives a signature of the surface electric field E. Thus in a 3P-PEEM experiment, the intensity of the photoemitted electron signal is proportional to the sixth power of the local electric field E 6 , whereas linear photoelectric events scale as E 2 . Mapping the light coupling at the nanoscale In order to localize the emission centers of photoelectrons, i.e. hot spots, a surface area irradiated by 20 ultrafast laser pulses where low spatial frequency LIPSS (LSFL) were preformed is probed. We illustrate our investigation on this irradiated zone as it provides a well defined LSFL topography with different roughness centers. Figure 2a shows a 3P-PEEM image probing such LSFL pre-structured with 20 laser shots, where the local field enhancement is manifested by hot spots. Due to the high nonlinear character of the photoemission response, the near field hot spots of large intensities mask those of lower intensities. To prevent dazzling effects due to a few number of bright spots and to extract the location of a hot spot whatever its relative intensity, a generic hot spot site is cross-correlated (CC) with a 3PPEEM image. The cross-correlation technique is an effective analysis tool, notably when noise is present in the image, to identify the occurrences of the mask (the user-defined hot spot) on the reference image (the 3P-PEEM here). This procedure yields a position map of all the emitting sites whatever their relative raw intensities. One cross correlation image revealing the map of the photoelectron emitting sites is displayed in Figure 2b.

Figure 2: High-resolution near-field mapping of LIPSS (a) 3P-PEEM of a site irradiated by 20 laser pulses. The red arrow shows the fs laser polarization. Near-field enhancement sites are imaged as hot spots colored in blue. (b) 3PPEEM autocorrelation map revealing all the emitting sites dazzled in (a). Map is obtained by correlation of the raw PEEM image with a generic hot spot profile. A 2D-FFT amplitude is performed for (b), which shows two identical symmetrical zones indicating a periodical distribution of the hot spot with a moderate dispersion of about 50 nm. (c) Combination of the SEM image with the cross-correlated 3P-PEEM image of the same site. A magnification of the yellow circle clearly shows the location of the high near field hot spots mostly located in the LIPSS valleys. the SEM image of the corresponding area is sum up with the cross correlated PEEM image and shown in Figure 2c. On this superposition, the hot spots are seen to be organized along lines corresponding to the LIPSS valleys. A yellow circle area is chosen and magnified in order to highlight this effect. Note that outside the laser impact, no photoemission signal is detected ensuring that the emitted photoelectrons come solely from the nanostructures. In addition, no IR laser-induced damage of LIPSS is observed during the PEEM experiment.

The amplitude of the Fast Fourier Transform (FFT) of the CC 3P-PEEM image reveals the periodicity of the photoelectron signal. Direct visual examination of the FFT map shows a quasi-periodic distribution of the photoelectron emission sites with a large dispersion. For a better visualization of the photoelectron periodic character as well as their emission sites,

Imaging the field enhancement on LSFL generated by 20 laser shots raises the question on the influence of the local structure and roughness on the energy light distribution on the surface. To elucidate the kind of nanorelief fostering the local absorption of the laser induced


4

3 µm

1P-PEEM

10 µm

(e)

(a) 1P+3P-PEEM

5 µm

5 µm (c)

10 µm

(d)

CC 1P+3P-PEEM

SEM

(f)

(b)

Figure 3: Mapping of the near-field and the surface features of a site where LIPSS are formed with 20 laser pulses at 0.24 J/cm2 . (a) 1P-PEEM image of the irradiated zone obtained with the UV lamp illumination. (b) Combination of the 1P-PEEM and 3P-PEEM images, where the local field enhancement localization arising from 3P-PEEM is manifested by the blue hot spots. (c) Correlation of the raw PEEM image (b) with a generic hot spot mask. Regions with black color represent ridges. (d) SEM image of LIPSS of the same region as in (c) where some nanocavities in the valley structures are visible in black. (e-f) Superposition of the magnifications images of (c) and (d) corresponding to the white square zone. fields at a nanometer scale, we have concurrently superposed the single photon emission 1P-PEEM (Figure 3 a reflecting unevenly absorbed intensity driven by the surface features 39 and the local near-field obtained by 3P-PEEM in Figure 3b. After applying the described cross-correlation method, both imaging processes reveal the spatially inhomogeneous energy coupling on the surface corresponding to two degrees of intensity shown in Figure 3c. A 2D characterization of the same patterned site is shown on the SEM micrograph displayed in Figure 3d. One-photon absorption is particularly sensitive to the main periodic pattern barring some fine details as can be seen in the higher resolution SEM image. By contrast, three-photon absorption image exhibits a less expected but more informative structure punctuated by bright hot-spots. It appears that while 1P-PEEM spatially resolves the valleys of the ripples, 3P-PEEM unveils the location of nanocavities represented by the light blue color and clearly visible by superposing the SEM and the PEEM figures as depicted in Figures 3e-f. The combination of the two images ev-

idences that photoelectrons are more emitted from the valleys than the ridges and, in particular, in the deeper recesses. Observed for a specific irradiation dose condition, the modulated photoelectron emission signal questions on the origin, consistency and stability of the periodicity observed in a multipulse regime. Based on the accuracy of this method revealing two kinds of topographic patterns over the surface impacted by the laser, the optical feedback concept can be investigated by imaging the field enhancement on pre-formed LIPSS. Dynamics of LIPSS formation Mapping of the near field distribution for different topographies constitutes a unique perspective to investigate the dynamics of low spatial frequency LIPSS (LSFL) formation. Figure 4 displays the superpositions of SEM and crosscorrelated 3P-PEEM images for 4 different impact sites corresponding to LIPSS formation with 1, 5, 10 and 20 pulses. The hot spots designating the photoemission sites are colored in blue. The images reveal a marked difference regarding the evolution of the concentration of 5

Page 7 of 15 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 Photonics

the hot spots as well as their periodic character.

due to a higher absorption. As a consequence, the contrast of the ripple pattern is enhanced via positive feedback. The underlying question that must be addressed is: does the local field enhancement follow the LSFL topography or does it govern the reduction of the LSFL period ? A systematic investigation has been performed to their relative periodic character for each laser pulse irradiation conditions. Figure 5a shows a plot of LIPSS and photoelectron periodicities as a function of the number of laser pulses. LIPSS periodicity is extracted from the SEM-FFT amplitude analysis. In order to obtain a complementary visualization of the field distribution on the LSFL topography, we plot the radially averaged autocorrelation (RAA) of each CC 3P-PEEM image corresponding to impacts of consecutively larger number of laser pulses. The RAA function enables to highlight possible periodicities of a given image, which are sometimes difficult to apprehend from the original image, as in the case of circular hot spots. To discern patterns with a radial symmetry, RAA function consists of a Fourier-Bessel transform of the image and is more suited than FFT to estimate how spot density varies as a function of distance from a reference spot. Figure 5b shows the normalized autocorrelation of photoemission images from three different LIPSS sites preformed with 5, 10 and 20 laser shots. The RAA plot confirms a periodic character of the photoemission signal noticeable in Figure 5c whereas the 2D FFT spectrum shown in Figure 5d reveals a non-exploitable scattered pattern. This periodic character starts at 10 pulses where the first peak of the curve is located at 695 nm. It is well defined for LIPSS formed with 20 pulses, where the corresponding peak is shifted down by 50 nm to a periodicity of about 645 nm. The red and dashed blue curves in Figure 5a represent the evolutions of LIPSS (SEM) and PEEM-periodicities as a function of the pulse number, respectively. For LIPSS prestructured with a moderate number of laser pulses, i.e. 10 pulses, a shift between the two curves is present. This shift decreases as the pulse number increases and disappears when reaching saturation for a high number of laser pulses.

1P 5P

10P 2 µm

20P 2 µm 2 µm 22 µm µm SEM+3P-PEEM

Figure 4: Combinations of SEM and PEEM near-field mapping of LIPSS formed with different number of laser pulses. The red arrow shows the laser polarization. Hot spots are randomly localized in the case of 1 pulse topography and start to follow the LIPSS structures as the number of laser pulses increases. Hot spots reflects the stronger light absorption attributed to the presence of nanocavities on the metallic surface.

In fact, photoemission sites of the 1 pulse impact are randomly distributed whereas they begin to evolve more orderly with increasing number of laser pulses. For LSFL formed with 10 and 20 pulses, hot spots can be seen decorating the sub-wavelength hollowed structures, i.e. emitted from nanocavities. This result indicates that the energy absorption of the laser pulse is enhanced at nanocavity locations, which may drive the topography following subsequent pulse irradiation. Then, probing the local field distribution for LSFL preformed with consecutively larger number of laser pulses reveals high yield photoemission sites showing that the absorbed energy is concentrated in particular nano-depression at the surface. Periodic local field enhancement Careful examination of the photoelectron emission center locations suggests that the energy is deposited in the LIPSS valleys, where the optical intensity is always larger than that of the adjacent ridges, which deepens thus the valleys


6

20P 5

signal, evidencing that the energy is deposited with a periodicity lower than that of the LSFL.

0 -5

-5

0

5

10P 0 -5 0

5

5

1P

Ky / K0

5

-5

3D electromagnetic calculations. SEM image in Figure 2c gives us 2D information about the projection of the surface topography. Combined with an atomic force microscopy measurement of the topography evaluating a typical ripples contrast of 100 nm, a low resolution 3D surface layer has been numerically reconstructed. The energy deposition on the surface irradiated by 20 pulses is then calculated by Finite-Difference Time-Domain method (FDTD), solving full-vector Maxwell equations. 40 The cube of the intensity below the surface is plotted in Figure 6a and correlates well with 3P-PEEM image in Figure 2b. The hot spots correspond here to the positions of the deepest nanocavities and their local nonradiative near fields, whereas the periodic patterns, which are the consequence of the interference with radiative fields, are less intense. This suggests that at the bottom of a surface cavity, the material experiences a local field enhancement which produce a strong photo-emission of electrons observed by the 3P-PEEM. The calculations reveal the inhomogeneous nature of the topography below the LIPSS and a strong correlation between the near-field response, revealed by 3P-PEEM, and collective radiative field response, recognized by 1P-PEEM. This way, the nanoscale topography of a metal surface defines the non-radiative optical response of the system, which further manifests in rippled structure nanoimprinting. 13 To elucidate the role of the nanocavities in the establishment of final interference patterns and LIPSS pulse-by-pulse evolution, we performed electromagnetic simulations with different concentrations of randomly distributed nanocavities of radius R = 20 nm below the metal surface. The absorbed energies corresponding to the nanocavities with average spacing of d = 1µm and d = 0.5µm are shown in Figures 6c-d. The collective effects of coupled nanocavities influence the periodicity of the resulting interference patterns. For large nanocavity spacings, the period is found to be close to the surface plasmon wavelength ≈ 750 nm, whereas for

0

-5

0

5

0

(c)

0

5

Kx / K0

K0

-5

Distance (µm)

-5

Ky /

20P

20P 10P 5P

(b)

Kx / K0 5

RAA

3P-PEEM

-5

(a)

(d)

Figure 5: Evolution of LIPSS and photoelectron site periodicities as a function of the number of laser pulses. (a) Plot of the LIPSS periodicity obtained by FFT analysis of SEM images for surfaces irradiated with laser pulse numbers between 10 and 30, and photoelectron site periodicity obtained by a radially averaged autocorrelation RAA performed on the corresponding cross correlated 3P-PEEM images for the same sites. Red and blue diagrams represent LIPSS and photoelectron site periodicity, respectively. (b) FFT of LIPSS (c) Radially averaged autocorrelation of 3-photons photoelectron signal as a function of the distance and (d) FFT of a 3P-PEEM image for a 20 pulses LIPSS site.

Surface topography dynamics as well as LIPSS periodicity evolution with the applied number of laser pulses is a question congenitally linked to the feedback nature. This involves a strong dependence of the light coupling features on the nonlinearly formed periodic nanostructures. An early modulated pattern develops on the surface after the first laser pulses, where valleys and ridges are formed on the surface morphology. Subsequent laser pulses interact with an amended surface, where the local field is enhanced in particular regions corresponding to surface depression. Our results show that this decrease in LSFL periodicity is directly linked to the periodicity of the photoemission 7

Page 9 of 15 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 Photonics

subwavelength nanocavity spacings the period is significantly smaller ≈ 600 nm due to the surface wave retardation associated to the presence of the nanocavities. The relevant model which takes account for the phase retardation is the coupled-dipole model, based on discrete dipole approximation. 41,42 a)

(𝐼 𝐼0 )3 3

𝐸

b)

5

100

No shift

50

c)

2 0

0

0 -5

0 X[µm]

5 Redshift ~750 nm

𝐸

Redshift

-50

1 -5

c)

dominant. Such a hybrid wave has a characteristic intensity maximum close to the surface plasmon wavelength. Upon decrease of the distance between the nanocavities, the role of the evanescent QC wave becomes more pronounced and the resonant wavelength shifts to smaller values, as confirmed by FDTD simulations. The characteristic blue shift was previously evidenced by optical spectroscopy of randomly distributed nanocavities on thin gold films 43 and the increasing role of evanescent QC waves was detected with the rise of the nanocavity density. 44 According to our simulations, the situation with an average distance between the dipoles of Λdd ≈ 3λ/4 ≈ 625 nm corresponds to a local equilibrium and coincides with the final saturated LIPSS periodicity. We propose that the periodicity of LIPSS follows the average spacing between the nanocavities pulse-by-pulse, approaching the saturated value, where no further resonant wavelength shift is expected. The evolution of ripples from rough surface with nanocavities, taking account for pulse to pulse evolution of the irradiated has been recently demonstrated combining electromagnetics and ablation model. 13,45 The reduction of the calculated periods due to a surface wave retardation via coupling with neighboring nanocavities progressively formed by hydrodynamic mechanisms is fully consistent with the reported experimental observation.

d)

5

5

0

0

d)

-100 Blueshift 0 0.2 0.4 0.6 0.8 1 1.2 Distance between holes [µm] (𝐼 − 𝐼0 ) 𝐼0 3 𝐸 2

1 -5

Blueshift ~600 nm -5

0 X[µm]

5

-5 0 -5

0 X[µm]

5

Figure 6: (a) Cubic intensity distribution I 3 calculated by 3D-FDTD simulations taken 20 nm below Ti rough surface with experimentally defined topography shown in Figure 2c in the transverse plane perpendicular to laser propagation. (b) Predictions of resonant wavelength shift as a function of the average distance between the nanocavities within the coupleddipole approach. (c, d) Intensity distributions below the surface with randomly distributed nanocavities of radii R = 20 nm and average spacings (c) d = 1 µm and (d) d = 0.5 µm.

Conclusion In summary, the interpulse feedback mechanism driving the final topography of LIPSS is outlined by employing the photoemission electron microscopy. Simultaneous one-photon and three-photon emission microscopy allows to access both quasi-periodic radiative and nonradiative patterns produced by the incident light coupling with the topography nanofeatures. A periodic character of the electrons photoemitted from nanocavities in the LIPSS region is unveiled, where a periodic behavior of the laser energy deposition is verified. These results indicate that energy light absorption is well modulated but also highly localized and en-

Such an approximation is valid if the nanocavities are significantly smaller than the laser wavelength R λ and the distances between them are few times larger than their sizes d R. Following this approach, the collective response of the system as well as the resonant frequency shift are defined as a function of the positions of the dipoles. For instance, the dependency between the resonant shift and the average distance between the nanocavities is plotted in Figure 6(b). For nanocavity spacing larger than λ, each nanocavity acts as an individual source launching both evanescent quasicylindrical (QC) and SP waves, the latter is


8

ACS Photonics 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

hanced in the vicinity of nanocavities, evidencing a complex energy deposition. As revealed by the hot spot distribution, the periodicity of the three-photon photoemission signal is lower than that of the LIPSS for a low number of applied pulses, and then saturates when increasing the number of laser pulses. To elucidate the role of nanocavities and surface curvature in the constitution of final interference patterns and possible LIPSS interpulse evolution, the experimental results are coupled to electromagnetic calculations. The latter are performed on experimental roughness profiles and also on artificial roughness layers with different concentrations of randomly distributed nanocavities. It is found that the LIPSS period is shifted to lower values as a consequence of the reinforced dipole-dipole coupling for smaller spacings between nanocavities. In a multipulse feedback, LIPSS periodicity converges to the saturation value of Λ ≈ 3λ/4. This work elucidates the origin of the periodicity shifts of self-organized surface nanopattern providing an original experimental evidence that the periodicity evolution is driven by collective near fields effects via the reinforcement of dipole-dipole coupling as the nanorelief density increases. Photoemission imaging the coupling between light and nanostructures opens the way for a precise control of the electromagnetic feedback in laser-induced nanostructuring, as well as optimal light manipulation on the nanoscale.

Page 10 of 15

waveplate and a polarizer, see Figure 1a. Laser pulses are attenuated through a pair of neutral density filters and focused on the sample surface under normal incidence. Laser-induced surface structuring is analyzed using a scanning electron microscope NovananoSEM200 from the FEI company. The sample consists of a set of laser impacts where ripples are formed with different numbers of laser pulses ranging from 1 to 20 pulses at a fixed peak fluence of 0.24J/cm2 under single-pulse ablation threshold and normal incidence. PEEM measurements (CEA IRAMIS SPEC). The light source is a femtosecond TiSapphire laser system (Chameleon Ultra II, Coherent Inc.) providing IR photons in a wavelength window ranging from 680 to 1080 nm and a pulse duration equals to 140 fs. During PEEM experiment, the wavelength is fixed to 800 nm (1.55 eV) for coherency with the LIPSS formation conditions. Laser pulses are focused on the sample under normal incidence with a 400 mm focal length lens. 29,34 In parallel, a UV Hg Lamp (OSRAM HBO 103 W) is used to conduct linear one-photon imaging. The photoemission electron instrument is a LEEM/PEEM III by Elmitec GmbH working in ultrahigh vacuum in the low-pressure range of 10−10 mbar, and with no energy filtering (total electron yield). The lateral resolution achieved by nPPEEM is routinely 20 nm. A charge-coupled device CCD camera (PCO Sensicam) records digitized real space images for ulterior investigation. For further experimental details, see ref. 34–36 Solving Maxwell equations. Electromagnetic simulations are performed by using 3D Finite-Difference Time-Domain method 46 with ∆x = ∆y = ∆z = 20 nm resolution and 1000 × 1000 × 150 calculation cells. The XY profile of the Ti surface for FDTD calculations in Figure 6a has been extracted from SEM measurements shown in Figure 2c. The depth along Z axis has been artificially reconstructed from the image contrast. More precisely, 5 levels of topography have been defined based on the grey level of the SEM micrograph. The maximal difference between crest and valley has been set to 100nm as a standard ripple contrast mea-

Methods Sample preparation. Ti samples are first mechanically polished then finished with colloidal silica, providing an excellent surface mirrorlike with RMS roughness of ≈ 2 nm. For laser-induced surface structuring, a commercial femtosecond Ti-Sapphire laser system (Legend, Coherent Inc.) delivers IR linearly-polarized pulses at a central wavelength of λ = 800 nm with a pulse duration of 150 fs, and a repetition rate of 1 kHz under atmospheric pressure. An electromechanical shutter enables the selection of the required laser pulse number with a pulse energy controlled by the combination of a half


9

Page 11 of 15 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 Photonics

tive plasmonics. Nature Photonics 2009, 3, 55.

sured by AFM for titanium for these 2 irradia ~ tion conditions. Intensity I = E is calculated by averaging the sum of squared electric fields over 30 optical periods corresponding to θ = 40 fs (FWHM) pulse duration of temporally Gaussian pulse at λ = 800 nm irradiation. The optical properties of Ti with n = 3.14 and k = 4.0 for laser wavelength λ are used in simulations, whereas cavities are the half-spheres with n = 1 and k = 0. Convolutional perfect matched layers (CPML) of 15 cell thickness are applied on the borders of the grid to avoid non-physical reflections. 47 The spatially focused Gaussian beam is chosen as the source with w0 = 20µm beam waist, linear polariza~ along Y axis and propagation ~k along Z tion E axis.

(4) Prikulis, J.; Hanarp, P.; Olofsson, L.; Sutherland, D.; Käll, M. Optical spectroscopy of nanometric holes in thin gold films. Nano Letters 2004, 4, 1003–1007. (5) Young, J. F.; Sipe, J.; Van Driel, H. Laser-induced periodic surface structure. III. Fluence regimes, the role of feedback, and details of the induced topography in germanium. Physical Review B 1984, 30, 2001. (6) Zhang, H.; Colombier, J.-P.; Li, C.; Faure, N.; Cheng, G.; Stoian, R. Coherence in ultrafast laser-induced periodic surface structures. Physical Review B 2015, 92, 174109.

Acknowledgement This work was supported by the LABEX MANUTECH-SISE (ANR-10-LABX-0075) of Université de Lyon, within the program "Investissements d’Avenir" (ANR-11-IDEX-0007) operated by the French National Research Agency (ANR). Theoretical part was supported by the IMOTEP project operated by ADEME. The CEA author acknowledges financial support by the French National Agency (ANR) in the frame of its program in Nanosciences and Nanotechnologies (PEEMPlasmon Project ANR-08-NANO-034), Nanosciences Ile-de-France (PEEMPlasmonics project) and the ’Triangle de la Physique’ (PEPS Project 2012-035T).

(7) Abou-Saleh, A.; Karim, E. T.; Maurice, C.; Reynaud, S.; Pigeon, F.; Garrelie, F.; Zhigilei, L. V.; Colombier, J. P. Spallation-induced roughness promoting high spatial frequency nanostructure formation on Cr. Applied Physics A 2018, 124, 308. (8) Sedao, X.; Abou Saleh, A.; Rudenko, A.; Douillard, T.; Esnouf, C.; Reynaud, S.; Maurice, C.; Pigeon, F.; Garrelie, F.; Colombier, J.-P. Self-Arranged Periodic Nanovoids by Ultrafast Laser-Induced Near-Field Enhancement. ACS photonics 2018, 5, 1418–1426.

References

(9) Rudenko, A.; Colombier, J.-P.; Höhm, S.; Rosenfeld, A.; Krüger, J.; Bonse, J.; Itina, T. Spontaneous periodic ordering on the surface and in the bulk of dielectrics irradiated by ultrafast laser: a shared electromagnetic origin. Scientific Reports 2017, 7, 12306–1 – 12306–14.

(1) Ebbesen, T. W.; Lezec, H. J.; Ghaemi, H.; Thio, T.; Wolff, P. Extraordinary optical transmission through sub-wavelength hole arrays. Nature 1998, 391, 667. (2) Lalanne, P.; Hugonin, J. P. Interaction between optical nano-objects at metallodielectric interfaces. Nature Physics 2006, 2, 551.

(10) Segev, M.; Silberberg, Y.; Christodoulides, D. N. Anderson localization of light. Nature Photonics 2013, 7, 197.

(3) MacDonald, K. F.; Sámson, Z. L.; Stockman, M. I.; Zheludev, N. I. Ultrafast ac-


10

ACS Photonics 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 15

(11) Sánchez-Gil, J. A. Localized surfaceplasmon polaritons in disordered nanostructured metal surfaces: Shape versus Anderson-localized resonances. Physical Review B 2003, 68, 113410.

(20) Zurita-Sánchez, J. R.; Novotny, L. Multipolar interband absorption in a semiconductor quantum dot. I. Electric quadrupole enhancement. JOSA B 2002, 19, 1355–1362.

(12) Garrelie, F.; Colombier, J.-P.; Pigeon, F.; Tonchev, S.; Faure, N.; Bounhalli, M.; Reynaud, S.; Parriaux, O. Evidence of surface plasmon resonance in ultrafast laserinduced ripples. Optics express 2011, 19, 9035–9043.

(21) Dorn, R.; Quabis, S.; Leuchs, G. Sharper focus for a radially polarized light beam. Physical review letters 2003, 91, 233901. (22) Höhm, S.; Rosenfeld, A.; Krüger, J.; Bonse, J. Laser-induced periodic surface structures on titanium upon single-and two-color femtosecond double-pulse irradiation. Optics express 2015, 23, 25959– 25971.

(13) Rudenko, A.; Mauclair, C.; Garrelie, F.; Stoian, R.; Colombier, J.-P. Self-organization of surfaces on the nanoscale by topography-mediated selection of quasi-cylindrical and plasmonic waves. Nanophotonics 2019, 8, 459–465.

(23) Aguilar, A.; Mauclair, C.; Faure, N.; Colombier, J.-P.; Stoian, R. In-situ highresolution visualization of laser-induced periodic nanostructures driven by optical feedback. Scientific reports 2017, 7, 16509.

(14) Lezec, H. J.; Thio, T. Diffracted evanescent wave model for enhanced and suppressed optical transmission through subwavelength hole arrays. Optics express 2004, 12, 3629–3651.

(24) Huang, M.; Zhao, F.; Cheng, Y.; Xu, N.; Xu, Z. Origin of laser-induced nearsubwavelength ripples: interference between surface plasmons and incident laser. ACS nano 2009, 3, 4062–4070.

(15) Sarrazin, M.; Vigneron, J.-P.; Vigoureux, J.-M. Role of Wood anomalies in optical properties of thin metallic films with a bidimensional array of subwavelength holes. Physical Review B 2003, 67, 085415.

(25) Lim, H. U.; Kang, J.; Guo, C.; Hwang, T. Y. Manipulation of multiple periodic surface structures on metals induced by femtosecond lasers. Applied Surface Science 2018, 454, 327–333.

(16) Cao, Q.; Lalanne, P. Negative role of surface plasmons in the transmission of metallic gratings with very narrow slits. Physical Review Letters 2002, 88, 057403.

(26) Hwang, T. Y.; Guo, C. Angular effects of nanostructure-covered femtosecond laser induced periodic surface structures on metals. Journal of Applied Physics 2010, 108, 073523.

(17) Scully, M. O.; Zubairy, M. S. Simple laser accelerator: Optics and particle dynamics. Physical Review A 1991, 44, 2656. (18) Quabis, S.; Dorn, R.; Eberler, M.; Glöckl, O.; Leuchs, G. Focusing light to a tighter spot1. Optics communications 2000, 179, 1–7.

(27) Ionin, A. A.; Kudryashov, S. I.; Makarov, S. V.; Saltuganov, P. N.; Seleznev, L. V.; Sinitsyn, D. V.; Lednev, V.; Pershin, S. M. Electron emission and ultrafast low-fluence plasma formation during single-shot femtosecond laser surface ablation of various materials. JETP Letters 2015, 101, 308–312.

(19) Youngworth, K. S.; Brown, T. G. Focusing of high numerical aperture cylindricalvector beams. Optics Express 2000, 7, 77– 87.


11

Page 13 of 15 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 Photonics

(28) Fuentes-Edfuf, Y.; Sánchez-Gil, J. A.; Florian, C.; Giannini, V.; Solis, J.; Siegel, J. Surface Plasmon Polaritons on Rough Metal Surfaces: Role in the Formation of Laser-Induced Periodic Surface Structures. ACS Omega 2019, 4, 6939–6946.

Adam, P.-M.; Royer, P. Short range plasmon resonators probed by photoemission electron microscopy. Nano letters 2008, 8, 935–940. (36) Peppernick, S. J.; Joly, A. G.; Beck, K. M.; Hess, W. P. Plasmon-induced optical field enhancement studied by correlated scanning and photoemission electron microscopy. The Journal of chemical physics 2013, 138, 154701.

(29) Kahl, P.; Wall, S.; Witt, C.; Schneider, C.; Bayer, D.; Fischer, A.; Melchior, P.; Horn-von Hoegen, M.; Aeschlimann, M.; zu Heringdorf, F.-J. M. Normal-incidence photoemission electron microscopy (NIPEEM) for imaging surface plasmon polaritons. Plasmonics 2014, 9, 1401–1407.

(37) Hwang, T. Y.; Vorobyev, A.; Guo, C. Surface-plasmon-enhanced photoelectron emission from nanostructure-covered periodic grooves on metals. Physical Review B 2009, 79, 085425.

(30) Word, R. C.; Dornan, T.; Könenkamp, R. Photoemission from localized surface plasmons in fractal metal nanostructures. Applied Physics Letters 2010, 96, 251110.

(38) Derry, G. N.; Kern, M. E.; Worth, E. H. Recommended values of clean metal surface work functions. Journal of Vacuum Science & Technology A: Vacuum, Surfaces, and Films 2015, 33, 060801.

(31) Fitzgerald, J.; Word, R. C.; Saliba, S.; Könenkamp, R. Photonic near-field imaging in multiphoton photoemission electron microscopy. Physical Review B 2013, 87, 205419.

(39) Kubo, A.; Jung, Y. S.; Kim, H. K.; Petek, H. Femtosecond microscopy of localized and propagating surface plasmons in silver gratings. Journal of Physics B: Atomic, Molecular and Optical Physics 2007, 40, S259.

(32) Lemke, C.; Leißner, T.; Jauernik, S.; Klick, A.; Fiutowski, J.; KjelstrupHansen, J.; Rubahn, H.-G.; Bauer, M. Mapping surface plasmon polariton propagation via counter-propagating light pulses. Optics Express 2012, 20, 12877– 12884.

(40) Rudenko, A.; Mauclair, C.; Garrelie, F.; Stoian, R.; Colombier, J. Light absorption by surface nanoholes and nanobumps. Applied Surface Science 2019, 470, 228–233.

(33) Douillard, L.; Charra, F. High-resolution mapping of plasmonic modes: photoemission and scanning tunnelling luminescence microscopies. Journal of Physics D: Applied Physics 2011, 44, 464002.

(41) Zhao, L.; Kelly, K. L.; Schatz, G. C. The extinction spectra of silver nanoparticle arrays: influence of array structure on plasmon resonance wavelength and width. The Journal of Physical Chemistry B 2003, 107, 7343–7350.

(34) Douillard, L.; Charra, F.; Fiorini, C.; Adam, P. M.; Bachelot, R.; Kostcheev, S.; Lerondel, G.; Lamy de la Chapelle, M.; Royer, P. Optical properties of metal nanoparticles as probed by photoemission electron microscopy. Journal of applied physics 2007, 101, 083518.

(42) Jain, P. K.; Huang, W.; El-Sayed, M. A. On the universal scaling behavior of the distance decay of plasmon coupling in metal nanoparticle pairs: a plasmon ruler equation. Nano Letters 2007, 7, 2080– 2088.

(35) Douillard, L.; Charra, F.; Korczak, Z.; Bachelot, R.; Kostcheev, S.; Lerondel, G.;


12

ACS Photonics 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

(43) Prikulis, J.; Hanarp, P.; Olofsson, L.; Sutherland, D.; Käll, M. Optical spectroscopy of nanometric holes in thin gold films. Nano Letters 2004, 4, 1003–1007. (44) van Beijnum, F.; Rétif, C.; Smiet, C. B.; Liu, H.; Lalanne, P.; van Exter, M. P. Quasi-cylindrical wave contribution in experiments on extraordinary optical transmission. Nature 2012, 492, 411. (45) Rudenko, A.; Mauclair, C.; Garrelie, F.; Stoian, R.; Colombier, J.-P. Amplification and regulation of periodic nanostructures in multipulse ultrashort laser-induced surface evolution by electromagnetichydrodynamic simulations. Physical Review B 2019, 99, 235412. (46) others„ et al. Computational electrodynamics: the finite-difference time-domain method. Norwood, 2nd Edition, MA: Artech House, l995 1995, (47) Roden, J. A.; Gedney, S. D. Convolution PML (CPML): An efficient FDTD implementation of the CFS-PML for arbitrary media. Microwave and Optical Technology Letters 2000, 27, 334–339.


13

Page 14 of 15

Page 15 of 15 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 Photonics

PEEM

1 pulse

e

e

10 pulses

e

e

e

Figure 7: Laser-induced periodic nanostructures probed by photoemission electron microscope


14

20 pulses

Nanoscale Imaging of Ultrafast Light Coupling to Self-Organized

Recommend Documents