Surface nanostructuring via femtosecond lasers

Mu-Tian Lib, Monan Liu*a and Hong-Bo Sunb
aDepartment of Condensed Matter Physics, College of Physics, Jilin University, Changchun, Jilin 130012, China. E-mail:
bState Key Laboratory on Integrated Optoelectronics, College of Electronic Science and Engineering, Jilin University, Changchun, Jilin 130012, China

Received 30th September 2019 , Accepted 23rd October 2019

First published on 30th October 2019

Periodical structures induced by pulsed lasers are a unique phenomenon when pulsed lasers irradiate on some material surfaces. These periodical structures with a subwavelength-scale period hold potential in integrated-optics and biomimetic micro-nanodevices for their direct shaping by laser pulses. However, the blurred nature of the laser-induced structuring hinders its further exploration in these application scopes. In this review, the plasmon-mediated structuring targeted on various materials, both organic and inorganic, will be discussed profoundly.


Laser-induced periodic surface structures (LIPSS) are a series of subwavelength structures identified in the light–matter interaction when lasers irradiate materials. They were first found in the irradiation of pulsed ruby lasers on semiconductors by Birnbaum in 1965,1 which are generally ripple-like periodic nanostructures either perpendicular or parallel to the polarization direction of the incident lasers.2–5 In the following decades, a lot of work has been carried out on LIPSS, not only to reveal the formation mechanism, but also to explore their potential applications. So far, the phenomenon of LIPSS has been proved universal as long as a material is ablated with pulse energy close to its threshold. A variety of materials have been tried for obtaining LIPSS, including semiconductors,6–10 metals,6,11,12 dielectrics6,13 and even 2D materials.14 The natural nanoscale periodicity and structural uniformity of LIPSS made them advantageous in structuring. In the pursuit of highly functional micro-nanostructures, such as biomimetic structures,15 LIPSS can be well introduced to almost any surfaces and realize complex structure design (e.g. hierarchical structures). Today, they have been explored in a wide range of application scopes, including nanofabrication,9 optical property tailoring,16 biomimetic surfaces,17–19 oil–water separation,20,21 water-splitting,22 antibacterial surfaces,23 etc.

With the further exploration of LIPSS, their origin and the rich physics behind keep attracting people. According to the scale of their periodicity Λ, LIPSS can be classified into subwavelength structures (SWS) and deep-subwavelength length structures (DSWS), also known as low spatial frequency LIPSS (LSFL) and high spatial frequency LIPSS (HSFL).2 The formation of the former type has been widely regarded as a result of the interference between the incident light and the scattered one. Sometimes the interference can also happen between the incident light and the excited surface plasmon polaritons (SPPs).2 Therefore the resulting subwavelength structures usually show periods comparable to the wavelength λ of the incident lasers. The latter type, DSWS, show even smaller periods (<λ/2).2–5 This type is of great interest for their deep sub-wavelength scale as well as their controversial origin. Besides the “light–SPP interference” model mentioned above, a series of possible mechanisms have also been proposed to explain the formation of DSWS, such as defect-induced nanoplasmas,24 nanoplanes pinned to standing waves,25 second harmonic generation,26,27 etc. Among them, the incident light–SPP interference model is the most acceptable one since it has already provided certain positive results in simulations.28 More experimental data are needed to testify the model and enrich the theory. For example, most current LIPSS have been formed on inorganic material surfaces, while a few attempts have been made on organic materials. With the fast development of stretchable devices and wearable devices, it is meaningful to realize deep-nanoscale structuring on various organic materials, such as resins and conductive polymers. Moreover, the fabrication control of LIPSS is not perfect and needs further optimization. In spite of the structural uniformity and deep subwavelength scale of LIPSS, the structuring of LIPSS is challenged by the throughput problem. The line-to-line scanning of nanoripples or nanogratings over wafer scale usually costs hours. Solving these problems is primitive to make ultimate use of LIPSS in nanodevice design and fabrication.

Based on the above motivations, we reviewed a series of work for obtaining LIPSS on both inorganic and organic materials. The target is to find out both the transverse and longitudinal structure evolution of LIPSS with fine-tuned pulse number and laser fluence. Moreover, solid experimental results could be provided to verify the proposed structure formation mechanism and hence enable better understanding of the light–matter interaction during the formation of LIPSS. By combining both the experimental measurements and theoretical simulations, it can be shown that the LIPSS formed on the target material surfaces are due to direct surface plasmon ablation. On the whole, this review is intended to provide deep insight into the formation mechanism of LIPSS for achieving better structure control and encouraging more innovative utilizations.

Femtosecond laser direct-writing (FsLDW) system

Pulsed lasers were generated via a Ti:sapphire regenerative amplifier system (Spectra Physcis) with a wavelength of 800 nm. The pulse duration and the repetition rate were 100 fs and 1 kHz, respectively. The outcoming lasers were expanded and then focused by a cylindrical lens to obtain an elliptical focal spot. The scanning direction for forming LIPSS was set perpendicular to the laser polarization. The focal spot was elongated such that the scanning efficiency could be significantly increased.

Plasmonic ablation on silica under low laser fluence

As for LIPSS, the period and the orientation of the nanoripples are of main concern for structure control. They are found to be closely dependent on laser power, polarization and pulse number.31–37 Plenty of research has been focused on laser-polarization-dependent nanoripple orientation within the focal plane, which is either parallel or perpendicular to the polarization direction. However, few insights have been provided into the effect of longitudinal energy distribution on structure formation. In order to reveal the structure characteristics along its vertical direction, a ZnS layer with a thickness of 100 nm was sputtered onto a silica substrate. A series of such double-layer samples were scanned under different laser fluences to find out the ablation thresholds for both layers. They were evaluated to be 0.45 J cm−2 (ZnS) and 1.98 J cm−2 (silica), respectively. Careful observations have been made on these ablated samples to identify the LIPSS evolution of each layer, as shown in Fig. 1(a1)–(a4). At a laser fluence of 0.5 J cm−2, nanogroove-like structures have been found in the ZnS layer with a period ∼190 nm, as in Fig. 1(a1). They are similar to those observed on a polycrystalline ZnS film only with lower uniformity.30 Further increasing the laser fluence would lead to wildly scattered particles and break the periodicity to some degree [Fig. 1(a2)]. After immersing the as-ablated samples in H2SO4 for 30 s, the upper layer of ZnS can be totally removed from the silica substrate. A further observation on the remaining silica substrate reveals a series of nanoripples imprinted on the substrate surface at a laser fluence much lower than the threshold (∼a fourth). The structure period is almost the same as that of the ZnS nanogrooves. It indicates that the upper nanogroove structures have been printed through the whole ZnS-layer thickness. The formation of the LIPSS on silica is dependent on the incident laser fluence, the scanning speed and the top-coated ZnS thickness. Neither a laser fluence higher than 0.7 J cm−2 nor a top layer thicker than 200 nm would result in any LIPSS on silica. A scanning speed higher than 20 μm s−1 would destroy the periodicity of the structure formed on silica as well.29
image file: c9cp05351d-f1.tif
Fig. 1 Plasmonic nanoprinting on ZnS, silica and silicon. (a1) and (a2) Nanogrooves formed on the top ZnS layer at laser fluence 0.5 and 0.6 J cm−2; (a3) and (a4) nanoripples imprinted on the silica substrate surface at a scanning speed of 20 and 10 μm s−1.(b) Plasma density and the corresponding reflection evolution of both top ZnS and Silica layers, with respective multiple-photon absorptions.29 (c) Wafer-size nanoprinting on silicon enabled by cylindrical lens at a repetition rate of 0.5 MHz and a scanning speed of 50 mm s−1. (d) Uniform nanoripples formed in the wafer-size nanoprinting.30 Copyright 2017, OSA. Copyright 2017, Nature.

The formation mechanism of nanoripples on the silica surface under low laser fluence has been discussed profoundly. The discussion is based on an assumed multi-photon excitation39 of a “plasma layer”40 at the interface of the top-coated ZnS-layer and the underlying silica substrate. According to the band gaps of the two layers, i.e. 3.35 eV and 8.00 eV, the incoming fs laser of 800 nm will be absorbed in the form of three photons and six photons, respectively. In this case, the nonlinearly ionized electrons are generated by two distinct processes, i.e. the multi-photon ionization (MPI) and the impact ionization.29,39 The corresponding plasma density evolution can be simulated based on the assumption via the iterative method on MATLAB, as shown by the plots in Fig. 1(b). The calculation has undergone 300 iterations and considered a Gaussian distribution of the pulse energy. It should be pointed out that the in situ reflectivity actually changes with the plasma accumulation. Therefore, the Drude model has been adopted as well to calculate the reflectivity that covers the contribution of the excited plasma layer. In Fig. 1(b), the line of the critical plasma density Nc provides the calculated ablation thresholds for the ZnS layer and the silica substrate, which are 0.41 J cm−2 and 1.80 J cm−2, respectively. The two values basically agree with the evaluated ones, indicating that the assumption of the MPI-excited plasma layer is acceptable in explaining the origin of the LIPSS formed on the ZnS–silica samples.29

Meanwhile, a fast-scanning strategy aiming at efficient wafer-size nanoprinting has also been proposed. A cylindrical lens was introduced to elongate the focal spot to an elliptical one, which was set in front of the objective lens (NA = 0.26). The cylindrical lens of 80 mm focal length finally formed a 1250 μm × 5 μm linear focal spot on the target surface. Given the laser fluence of 0.08 J cm−2 used, the actual ablation length reached 0.4 mm. Together with a high repetition frequency in the range of 0.1–0.5 MHz, the fast scan can realize a wafer-size nanoprinting in 1 h or so, as shown in Fig. 1(c) and (d). A 4 inch Si wafer was printed with uniform nanoripples with a period close to the wavelength of the incident lasers (1030 nm/230 fs). This can be well explained by Sipe's model, that ripples formed on an absorbing material surface are due to the interference between the incident light and the scattered light waves.30

SWS and DSWS: a competitive mechanism

While continuous efforts have been made to uncover the distinct nature of the two kinds of LIPSS, little attention has been paid to the situation of their coexistence. Can they coexist or even transform to each other? Questions like these are important for the ultimate control over LIPSS related structuring. Consider silicon as an example. Both SWS and DSWS have been observed on the material under quite different excitation conditions. SWS can be formed on a silicon substrate when exposed to amplified fs lasers, with periods (500–800 nm) comparable to the incident laser wavelengths.41 For the formation of DSWS on silicon, it needs excitations from a fs-laser oscillator with a repetition frequency higher than 10 MHz or from a fs-laser amplifier under a solution environment. These DSWS show similar nanoripple structures only with much smaller periods (70–130 nm). Now the same excitation parameters were chosen (local laser fluence and pulse number) to trigger both kinds of structure formation simultaneously. By studying the structure evolution of coexisting SWS and DSWS under a unified excitation condition, more knowledge on their formation mechanisms can be obtained for seeking a better structure control.38

In this strategy, 〈100〉 silicon was chosen as the processing substrate. A Ti:sapphire fs laser was used as the excitation light source, with 100 fs pulse duration and 1 kHz repetition rate. There are also reports on two-wave irradiation for generating LIPSS, in which a two-wave interaction was adopted in the laser ablation on silica with gold nanoparticles.42 The induced surface structures are surface arrays similar to those induced by single beam ablation, only with larger period and reduced depth. The wavelength of the laser beam was 800 nm. The laser fluence has been tuned in a relatively wide range to obtain coexisting SWS and DSWS, i.e. from two times to one fourth of the ablation threshold. The ablation threshold for SWS and DSWS on silicon has been identified as 170 mJ cm−2 and 76 mJ cm−2, respectively. The upper limit of the laser fluence for DSWS is 170 mJ cm−2. Moreover, all the as-ablated samples need to be rinsed in HF (5 mol L−1) for at least 15 min to remove the surrounding oxidized region and ablation debris. Four representative laser fluence values (436, 344, 287 and 258 mJ cm−2) have been selected for further observations on the surface structure evolution with accumulating laser pulses, as shown by the SEM images in Fig. 2(a)–(l). The period of the obtained SWS has also been plotted with the pulsed numbers, as shown in Fig. 2(m). An overall trend can be told from the three plots that the period of SWS always decreases with accumulating pulses and finally approaches a stable value. Only an exceptional small jump can be identified in the plot corresponding to the laser fluence of 344 mJ cm−2. The three groups of insets show the corresponding SWS evolution at the laser fluence of 436, 344 and 287 mJ cm−2. At the first few pulses, a circle with a serrated rim is formed with scattered particles inside and around. With increasing pulses, SWS show themselves inside the circle region and the ablation gets enhanced simultaneously due to the light scattered back by the periodic structures around the rim. A further accumulation of the laser pulses up to 100 would damage the SWS and lead to holes. Next, the structure evolution of coexisting SWS and DSWS has been revealed by slowing down the pulse accumulation. As shown in Fig. 2(a)–(l), twelve pulse numbers have been chosen for observation in the extended range of 1–2000. The laser fluence has been lowered to 258 mJ cm−2 as well. The first few pulses (1–5) only lead to scattered particles and then a clear rim appears. With the pulse number slowly increased to 200, one can find that the rim gradually turns to a serrated one. These “serrations” are DSWS formed both parallel and perpendicular to the polarization, with periods of ∼190 nm and ∼210 nm, respectively. When the pulse number is up to 250, SWS appear in the central region within the DSWS rim with a period of ∼580 nm. Further increasing the pulse number would lead to irregular structures in the central region and meanwhile destroy the DSWS along the polarization. This is due to a strong light scattering.38 The orientation of SWS has been found to be always perpendicular to the polarization direction. As for the orientation of DSWS, it has undergone an evolution process. At the early stage of pulse accumulation [Fig. 2(d)–(h)], the orientation of DSWS is along the polarization direction. However, with the further increase of pulse number, they are reconstructed and finally become perpendicular to the polarization direction [Fig. 2(l)]. There are also reports showing the opposite case, i.e. the orientation of DSWS parallel to the polarization direction, which need specific installations such as a slit.43 It has been observed that at the early stage of laser ablation, phase transitions such as melting and amorphization can happen in the surface region of silicon.38 This is known as the “seed layer” formation preparing for the final formation of LIPSS. As for the composition of the ablated substrates, superficial oxidization has been reported for TiN at high laser power.44 Therefore, in the case of silicon substrates, there might be superficial oxidation as well.

image file: c9cp05351d-f2.tif
Fig. 2 Coexisting SWS and DSWS during pulse accumulation. (a)–(l) Structure morphology at pulse numbers of 1, 5, 10, 20, 50, 100, 200, 250, 300, 500, 1000 and 2000, respectively. (m) Period evolution with pulse accumulation. Three representative laser fluences have been characterized, which are 436, 344 and 287 mJ cm−2. The insets are the corresponding structure evolution with the numbers in the images referring to the pulse number applied. The double arrows stand for the direction of laser polarization. (n) Schematic of the local-fluence-induced structure competition.38 Copyright 2017, OSA.

In order to explain the coexistence and competition between SWS and DSWS, theoretical simulations have been carried out. According to the former experimental results, a basic picture that light-induced changing plasma density leads to a period-varied nanoprinting has been proposed. Unlike the case in the previous section, the ionized carriers here are generated via single-photon ionization (SPI) and two-photon ionization (TPI), which are decided by the indirect and direct band-gap of silicon (1.12 and 3.40 eV). Both mechanisms contribute to the total carrier concentration. According to Sipe's model, these excited charge carriers form a layer of plasma, or “plasma plain”. The effective permittivity of this plasma plain is strongly dependent on the total carrier concentration,45 given by

image file: c9cp05351d-t1.tif(1)
εd = (nSi + kSi)2 (2)
image file: c9cp05351d-t2.tif(3)
where Neh is the total carrier concentration, ε* is the effective permittivity, ω and ωp are the cyclic and plasma frequency, and τD is the damping time constant (1 fs). For the calculation of εd, the real part nSi and the imaginary part kSi in the refractive index of Si at the incident wavelength of 800 nm are given as 3.681 and 0.005. For ωp, the values of electron charge e, the vacuum permittivity and the effective electron mass are given as 1.6 × 10−19 C, 8.854 × 10−12 F m−1 and 0.18.

The period of SWS Λ is decided by

Λ = λ/ne (4)
where λ is the wavelength of the incident light and ne is the real part of the complex refractive index.

The period of the DSWS ΛDSWS is decided by the wavelength of the plasmonic waves λsp

ΛDSWS = λsp/2 (5)
image file: c9cp05351d-t3.tif(6)
It is well known that the plasmons are the electrons collectively oscillating in response to the light on metal structure surfaces, in which the size and shape determine their nature. However, the case is a little bit different for the femtosecond-laser-excited plasmon nanoimprinting. On incidence of femtosecond lasers, the high pulse energy leads to single or multi-photon ionization, in which the carrier concentration is increased to a very high level (above 6 × 1020 cm−3) instantaneously and stable plasmonic waves are formed under the extremely high electric field (up to 109 V m−1). In this case, the resulting plasmons are mainly dependent on the ionized charge carrier concentration rather than surface morphology. Based on the above calculation, the structure period dependence on laser fluence can be well simulated. According to Sipe's model, the fluence threshold for forming the plasma layer should be 147 mJ cm−2. As mentioned above, this plasma layer results from the large number of ionized charge carriers and supports the interference between the incident light and the scattered surface waves. The theoretical fluence thresholds for forming stable SWS and DSWS have also been calculated to be 184 mJ cm−2 and 65 mJ cm−2, which are close to their experimental values (170 mJ cm−2 and 76 mJ cm−2). The structure competition picture has been further shown by the schematic in Fig. 2(n). It should be noted that, in most cases, the light field distribution of the laser beam focus is Gaussian-like. Therefore, the structure formation is actually decided by the local laser fluence rather than the mean laser fluence over the focal area, which explains the coexistence and competition between SWS and DSWS. That is, during pulse accumulation, as long as the local fluence reaches either threshold, SWS or DSWS begin to form. Therefore, as the mean laser fluence gradually increases, SWS first appear in the central region of the ablation spot. When the local laser fluence of the rim region reaches the threshold of DSWS, they are formed locally and result in the coexistence of central SWS and outer DSWS.38

Defining DSWS on organic materials

While plenty of explorations have been made on inorganic materials for fs-laser structuring, limited work has been carried out on organic materials. With the fast development of biomedicine and wearable devices, organic materials are drawing more and more attention due to their bio-compatibility.47 Laser engineering and structuring are also powerful tools for functionalizing bio-surfaces in a series of situations, such as tissue engineering22 and cell adhesion.17 It has been shown that the cell behaviors are susceptible to the engineered surface properties.47 In this strategy, uncovering the laser-structuring mechanism on biomaterials and organic materials is in urgent need. As discussed above, the formation of DSWS is closely dependent on the excitation of a large amount of free electrons in the nonlinear process. For the case of organic materials, the point is whether enough free electrons can be generated by breaking bonds such as π-bonds. With this aim, we tried to structure four most-commonly used organic materials in optoelectronics, which are poly-3,4-ethylenedioxythiophene:poly-(styrenesulfonate) (PEDOT:PSS), SU-8 photoresist, PMMA and PVA. They are chosen for they act differently as metals (PEDOT:PSS), semiconductors (SU-8 photoresist) and insulators (PMMA and PVA) during their charge carrier excitations by femtosecond lasers.46

The four organic materials were all pre-coated on a glass substrate with 500 nm film thickness. The wavelength of the incident pulsed lasers was 800 nm, with a repetition rate of 1 kHz and a pulse duration of 100 fs. In order to increase the laser-scanning throughput, the incident laser beam was first expanded and then focused by a cylindrical lens, which acquired a line-shaped focal spot (8 mm × 8 μm). Hence the scanning was directed along the short axis of the linear-shaped focus. Similarly to the exploration on inorganic materials, the LIPSS on the organic materials was acquired and observed at varied pulse numbers. The pulse number was controlled by the scanning speed, i.e. (W 1 kHz)/(20 μm s−1), where W is the width of the linear-shaped focus. The structure morphologies of the as-formed LIPSS on PEDOT:PSS, SU-8 photoresist, PMMA and PVA are shown in Fig. 3(a)–(d). Identical DSWS can be induced on both PEDOT:PSS and SU-8 photoresist substrates. The optimal pulse dosages for the two substrates are quite different, which are 500 and 1333 pulses per spot. This can be well understood by their different electron excitation levels. Since PEDOT:PSS is conductive and possesses large numbers of free electrons, DSWS start to take shape as nanoslits at only 50 pulses with a period of ∼280 nm. When the pulse number is increased to 500, the formed DSWS turn to nanogratings with an even smaller period of ∼120 nm, as shown in Fig. 3(a). In the case of SU-8 photoresist, although it is nonconductive, ionized electrons can still be generated by accumulating considerable pulses. The induced DSWS start to take shape at 800 pulses with a period of ∼240 nm. The DSWS on SU-8 photoresist becomes more uniform at 1333 pulses. When the pulse number is increased to 4000, the resulting DSWS is partially ablated and damaged. The induced structures found on PEDOT:PSS and SU-8 photoresist are both perpendicular to the laser polarization, which is similar to that reported on other inorganic materials.48 However, no continuous DSWS can be found on PMMA or PVA but only some fragments of induced structures, as shown in Fig. 3(a) and (ad).46

image file: c9cp05351d-f3.tif
Fig. 3 Representative DSWS formed on a series of organic materials: (a) PEDOT:PSS, (b) SU-8, (c) PMMA and (d) PVA. The insets are fast Fourier transform (FFT) analysis, showing that the period is 240 ± 70 nm. The solid red arrow and the dashed yellow arrow point to the polarization and scanning directions, respectively.46 Copyright 2017, IEEE.

Considering the above-mentioned similarities in both structure period and orientation relative to laser polarization, it is not difficult to unify the formation mechanism of the DSWS on both organic and inorganic materials. Just like the case of inorganic materials, the DSWS formation on organic materials probably results from a similar electron “excitation-ionization” process. Further evidence can be provided by Fourier-transformed infrared (FTIR) spectroscopy. For PEDOT:PSS, it can be found that the transmission of the structured sample has been greatly increased in the infrared range compared to the pristine one, indicating bond breakings during the laser irradiation. The strongly suppressed absorption corresponds to C–C, C[double bond, length as m-dash]C, C–O, C–S and C–H bonds, which are all broken by the high-energy laser pulses. Even part of benzene rings have been broken in this process and hence a large number of free electrons have been generated and ionized. For the case of SU-8, its benzene skeleton and C–O bonds keep intact after laser structuring. However, the epoxy group, C–H (within the ring) and C–S bonds show noticeable changes. These changes indicate possible electron transfer and hence are mostly likely responsible for the DSWS formation on the SU-8 substrate.46


In summary, a series of work for obtaining LIPSS on both inorganic and organic materials has been conducted for a better understanding of the structure formation mechanism. By fine-tuning of pulse number in the laser ablation, both the transverse and longitudinal structure evolutions have been successfully observed. Supported by the self-consistent simulations, the longitudinal structure evolution has further confirmed that the LIPSS formation under current parameter control is due to a plasmonic nanoimprinting. The transverse structure evolution has revealed the local energy dose effect on the formed LIPSS, that the Gaussian light-field distribution is responsible for the coexistence of SWS and DSWS. These results will no doubt provide people with more inspirations for LIPSS structuring. By combing with other approaches such as light-field modulation, the LIPSS structuring may reach its new stage and enable more advantageous nano-opitcs design.

Conflicts of interest

There are no conflicts to declare.


The authors sincerely acknowledge the support from National Natural Science Foundation of China (Grant No. 51501070).

Notes and references

  1. M. Birnbaum, J. Appl. Phys., 1965, 36, 3688–3689 CrossRef CAS.
  2. J. Bonse, J. Krueger, S. Hoehm and A. Rosenfeld, J. Laser Appl., 2012, 24, 042006 CrossRef.
  3. Q. H. Wu, Y. R. Ma, R. C. Fang, Y. Liao, Q. X. Yu, X. L. Chen and K. Wang, Appl. Phys. Lett., 2003, 82, 1703–1705 CrossRef CAS.
  4. M. Rohloff, S. K. Das, S. Hoehm, R. Grunwald, A. Rosenfeld, J. Krueger and J. Bonse, J. Appl. Phys., 2011, 110, 014910 CrossRef.
  5. T. Q. Jia, H. X. Chen, M. Huang, F. L. Zhao, J. R. Qiu, R. X. Li, Z. Z. Xu, X. K. He, J. Zhang and H. Kuroda, Phys. Rev. B: Condens. Matter Mater. Phys., 2005, 72, 125429 CrossRef.
  6. J. Bonse, S. Hoehm, S. V. Kirner, A. Rosenfeld and J. Krueger, IEEE J. Sel. Top. Quantum Electron., 2017, 23, 9000615 Search PubMed.
  7. M. Straub, M. Afshar, D. Feili, H. Seidel and K. Koenig, Opt. Lett., 2012, 37, 190–192 CrossRef CAS PubMed.
  8. S. Hoehm, A. Rosenfeld, J. Krueger and J. Bonse, Appl. Phys. Lett., 2013, 102, 054102 CrossRef.
  9. L. Wang, Q.-D. Chen, R. Yang, B.-B. Xu, H.-Y. Wang, H. Yang, C.-S. Huo, H.-B. Sun and H.-L. Tu, Appl. Phys. Lett., 2014, 104, 031904 CrossRef.
  10. C. Wang, H. Huo, M. Johnson, M. Shen and E. Mazur, Nanotechnology, 2010, 21, 075304 CrossRef PubMed.
  11. L. Gemini, M. Hashida, Y. Miyasaka, S. Inoue, J. Limpouch, T. Mocek and S. Sakabe, Appl. Surf. Sci., 2015, 336, 349–353 CrossRef CAS.
  12. Y. Miyasaka, M. Hashida, T. Nishii, S. Inoue and S. Sakabe, Appl. Phys. Lett., 2015, 106, 013101 CrossRef.
  13. R. Buividas, S. Rekstyte, M. Malinauskas and S. Juodkazis, Opt. Mater. Express, 2013, 3, 1674–1686 CrossRef.
  14. H.-B. Jiang, Y.-L. Zhang, Y. Liu, X.-Y. Fu, Y.-F. Li, Y.-Q. Liu, C.-H. Li and H.-B. Sun, Laser Photonics Rev., 2016, 10, 441–450 CrossRef CAS.
  15. M.-N. Liu, L. Wang, Y.-H. Yu and A.-W. Li, Opt. Mater. Express, 2017, 7, 2208–2217 CrossRef CAS.
  16. M. Martinez-Calderon, J. J. Azkona, N. Casquero, A. Rodriguez, M. Domke, M. Gomez-Aranzadi, S. M. Olaizola and E. Granados, Sci. Rep., 2018, 8, 14262 CrossRef CAS PubMed.
  17. Y. Sato, M. Tsukamoto, S. Masuno, Y. Yamashita, K. Yamashita, D. Tanigawa and N. Abe, Appl. Phys. A: Mater. Sci. Process., 2016, 122, 439 CrossRef.
  18. S. Yada and M. Terakawa, Opt. Express, 2015, 23, 5694–5703 CrossRef CAS PubMed.
  19. M. I. Abid, L. Wang, Q.-D. Chen, X.-W. Wang, S. Juodkazis and H.-B. Sun, Laser Photonics Rev., 2017, 11, 1600187 CrossRef.
  20. K. Yin, D. Chu, X. Dong, C. Wang, J.-A. Duan and J. He, Nanoscale, 2017, 9, 14229–14235 RSC.
  21. Y.-Q. Liu, D.-D. Han, Z.-Z. Jiao, Y. Liu, H.-B. Jiang, X.-H. Wu, H. Ding, Y.-L. Zhang and H.-B. Sun, Nanoscale, 2017, 9, 17933–17938 RSC.
  22. A. Bandyopadhyay, V. K. Balla, M. Roy and S. Bose, JOM, 2011, 63, 94–99 CrossRef CAS.
  23. A. H. A. Lutey, L. Gemini, L. Romoli, G. Lazzini, F. Fuso, M. Faucon and R. Kling, Sci. Rep., 2018, 8, 10112 CrossRef PubMed.
  24. R. Taylor, C. Hnatovsky and E. Simova, Laser Photonics Rev., 2008, 2, 26–46 CrossRef CAS.
  25. R. Buividas, L. Rosa, R. Sliupas, T. Kudrius, G. Slekys, V. Datsyuk and S. Juodkazis, Nanotechnology, 2011, 22, 055304 CrossRef PubMed.
  26. A. Borowiec and H. K. Haugen, Appl. Phys. Lett., 2003, 82, 4462–4464 CrossRef CAS.
  27. D. Dufft, A. Rosenfeld, S. K. Das, R. Grunwald and J. Bonse, J. Appl. Phys., 2009, 105, 034908 CrossRef.
  28. R. Buividas, M. Mikutis and S. Juodkazis, Prog. Quantum Electron., 2014, 38, 119–156 CrossRef.
  29. L. Wang, X.-W. Cao, M. I. Abid, Q.-K. Li, W.-J. Tian, Q.-D. Chen, S. Juodkazis and H.-B. Sun, Opt. Lett., 2017, 42, 4446–4449 CrossRef CAS PubMed.
  30. L. Wang, Q.-D. Chen, X.-W. Cao, R. Buividas, X. Wang, S. Juodkazis and H.-B. Sun, Light: Sci. Appl., 2017, 6, e17112 CrossRef CAS PubMed.
  31. Y. Shimotsuma, M. Sakakura, P. G. Kazansky, M. Beresna, J. Qiu, K. Miura and K. Hirao, Adv. Mater., 2010, 22, 4039–4043 CrossRef CAS PubMed.
  32. A. Rudenko, J.-P. Colombier and T. E. Itina, Phys. Rev. B, 2016, 93, 075427 CrossRef.
  33. P. S. Lemos, G. S. Silva, R. A. Roca, M. Assis, R. Torres-Mendieta, H. Beltran-Mir, G. Minguez-Vega, E. Cordoncillo, J. Andres and E. Longo, Phys. Chem. Chem. Phys., 2019, 21, 6101–6111 RSC.
  34. C. J. Rodrigues, J. A. Bobb, M. G. John, S. P. Fisenko, M. S. El-Shall and K. M. Tibbetts, Phys. Chem. Chem. Phys., 2018, 20, 28465–28475 RSC.
  35. H. Belmouaddine, M. Shi, L. Sanche and D. Houde, Phys. Chem. Chem. Phys., 2018, 20, 23403–23413 RSC.
  36. M. Assis, E. Cordoncillo, R. Torres-Mendieta, H. Beltran-Mir, G. Minguez-Vega, A. F. Gouveia, E. Leite, J. Andres and E. Longo, Phys. Chem. Chem. Phys., 2018, 20, 13693–13696 RSC.
  37. M. Oujja, J. G. Izquierdo, L. Banares, R. de Nalda and M. Castillejo, Phys. Chem. Chem. Phys., 2018, 20, 16956–16965 RSC.
  38. L. Wang, B.-B. Xu, X.-W. Cao, Q.-K. Li, W.-J. Tian, Q.-D. Chen, S. Juodkazis and H.-B. Sun, Optica, 2017, 4, 637–642 CrossRef.
  39. H. M. V. Driel, J. F. Young and J. E. Sipe, MRS Online Proc. Libr., 1982, 13, 302–318 Search PubMed.
  40. T. J. Y. Derrien, R. Koter, J. Krueger, S. Hoehm, A. Rosenfeld and J. Bonse, J. Appl. Phys., 2014, 116, 074902 CrossRef.
  41. J. Bonse, A. Rosenfeld and J. Krueger, J. Appl. Phys., 2009, 106, 104910 CrossRef.
  42. D. Torres-Torres, C. Torres-Torres, O. Vega-Becerra, J. C. Cheang-Wong, L. Rodríguez-Fernández, A. Crespo-Sosa and A. Oliver, Opt. Laser Technol., 2015, 75, 115–122 CrossRef CAS.
  43. T. Q. Jia, F. L. Zhao, M. Huang, H. X. Chen, J. R. Qiu, R. X. Li, Z. Z. Xu and H. Kuroda, Appl. Phys. Lett., 2006, 88, 111117 CrossRef.
  44. J. Bonse, S. V. Kirner, R. Koter, S. Pentzien, D. Spaltmann and J. Krüger, Appl. Surf. Sci., 2016, 418, 572–579 CrossRef.
  45. K. Sokolowski-Tinten and D. von der Linde, Phys. Rev. B: Condens. Matter Mater. Phys., 2000, 61, 2643–2650 CrossRef CAS.
  46. L. Wang, X.-W. Cao, C. Lv, H. Xia, W.-J. Tian, Q.-D. Chen, S. Juodkazis and H.-B. Sun, IEEE J. Quantum Electron., 2018, 54, 9200207 Search PubMed.
  47. Z. Wang, R. Zhou, F. Wen, R. Zhang, L. Ren, S. H. Teoh and M. Hong, J. Mater. Chem. B, 2018, 6, 3612–3631 RSC.
  48. K. Miyazaki and G. Miyaji, J. Appl. Phys., 2013, 114, 153108 CrossRef.

This journal is © the Owner Societies 2019