Periodic nanoripples formation on the semiconductors possessing different bandgaps

CHAPTER

Periodic nanoripples formation on the semiconductors possessing different bandgaps

1

1.1  NANORIPPLE FORMATION ON DIFFERENT BANDGAP SEMICONDUCTOR SURFACES USING FEMTOSECOND PULSES Surface structuring with lasers is a highly competitive method due to its capability to implement changes in the structure design. Compared to the picosecond and the nanosecond laser pulses, the energy in a femtosecond pulse can be precisely and rapidly deposited in a solid material with fewer thermal effects. Therefore the femtosecond lasers are widely used for microfabrication in transparent materials, metals, and semiconductors for applications such as modifying the optical properties of the surface [1], improvement of photovoltaic devices [2] by increasing their effective surface area, and in turn increasing the response and conversion efficiency. The other applications include modification inside refractive index of bulk materials for fabricating photonic devices, optical data storage, and biophotonic components [3–5]. Mid-infrared femtosecond laser radiation is generally used for the generation of surface ripples, also known as laser-induced periodic surface structures (LIPSS). Various mechanisms have been proposed to explain the nanoripple formation. Initially, it was thought to be the result of the interference between the incident laser light and the scattered light from the surface roughness [6]. However, if this mechanism was to be true, nanoripple width should always be of the order of the incident laser light wavelength, whereas, in many experiments, very narrow ripple formation has been observed. For example, Ozkan et al. [7] found that ripples resulting from 248 nm femtosecond laser irradiation of thin diamond films had a period varying between 50 and 100 nm. Yasumaru et al. [8] reported formation of ripple patterns with mean periods of 100–125 and 30–40 nm on TiN and diamond-like carbon after irradiation with 800 and 267 nm femtosecond pulses, respectively. To explain the nanoripple formation in the case of metals, the surface plasmon created during initial random surface heterogeneities was considered [9]. In dielectrics, the coupling of the electron plasma wave and incident laser explains the observed periodic structure in glass [10]. Other proposed mechanisms for ripple formation are self-organization [11], Coulomb explosion [12], influence of second harmonic generation [13], anisotropic local field enhancement invoking nanoplasmonics [14], etc. Nanostructured Nonlinear Optical Materials. http://dx.doi.org/10.1016/B978-0-12-814303-2.00001-5 Copyright © 2018 Elsevier Inc. All rights reserved.

1

2

CHAPTER 1  Periodic nanoripples formation on the semiconductors

Below, we analyze surface nanoripple formation on different bandgap semiconductors [15]. A study of the ripple formation with different laser parameters and ambient media is presented with the objective of identifying conditions of forming narrow period ripples. Nanoripples were formed using a Ti–sapphire laser with 8 mJ energy, 45 fs pulse duration, and 800 nm wavelength (1.56 eV) at a fluence in the range of 100 mJ cm−2–1 J cm−2. The effects of the number of laser shots, the angle of incidence, polarization of the laser, fluence, incident laser wavelength (λ), bandgap, and ambient medium were studied. Depending upon the experimental parameters, the nanoripple sizes varied in the range of λ/9–λ. Narrow nanoripples were formed on the wide bandgap (WBG) semiconductors. The width of the nanoripples decreased with the laser wavelength and the laser fluence. The observation of high and low spatial frequency ripples in different conditions is explained considering the transient metallic nature of the semiconductor surface on irradiation with intense femtosecond pulses. The surface eventually supports the surface plasmon excitation, which interferes with incident laser light for ripple formation. The very delicate dependence of ripple period with incident laser parameters is attributed to the critical role of electron density. The finding helps identifying suitable bandgap materials and laser parameters for obtaining nanoripple period considerably small compared to the incident laser wavelength.

1.1.1  EXPERIMENTAL ARRANGEMENTS AND RESULTS For studying the nanoripple formation from ultrashort laser pulse irradiation of semiconductor materials of different bandgaps, the following semiconductor materials of narrow (<1.5 eV, which is the energy of the incident 800 nm photon) bandgap materials such as InAs (0.36 eV), GaAs (1.42 eV), InP (1.35 eV), and wide (>1.5 eV) bandgap materials such as GaP (2.3 eV), GaN (3.4 eV), SiC (3.37 eV), and ZnSe (2.82 eV) were used. Multiple laser shots from a Ti–sapphire laser with 45 fs pulse duration, and 800 nm wavelength were focused in air on the semiconductor wafers at a fluence in the range of 100 mJ cm−2–1 J cm−2, that is, around the plasma formation threshold. Using a beta barium borate (BBO) crystal, the second harmonic (400 nm) of the 800 nm laser beam was also used to study the influence of the laser wavelength on the ripple formation while keeping the other parameters the same. Fig. 1.1 shows the surface morphologies resulting from the irradiation of a GaP wafer by femtosecond pulses with two different polarizations and in two different spatial regions. Fig. 1.1A shows the formation of spherical nanoparticles (∼100 nm) with circularly polarized laser light also reported by several groups [16]. Fig. 1.1B shows narrow nanoripples with ∼200 nm spacing at the peripheral regions of the laser irradiated spot with linearly polarized laser pulses. Fig. 1.1C shows wider nanoripples with about 600 nm spacing near the central hot region of the laser-irradiated spot. As expected, for linearly polarized light, the nanoripple orientation was always orthogonal to the laser polarization [17,18]. Nanoripple formation was observed with 10–100 shots fired on the semiconductors. It was also observed that number of shots fired on the semiconductor did not have any effect on the ripple period.

1.1 Nanoripple formation

FIGURE 1.1  GaP Irradiated by 800 nm Ultrashort Laser Pulses, (A) with circularly polarized beam, and (B) and (C) with linearly polarized beam in two different regions; (B) is around the periphery of the irradiated spot (low fluence); and (C) is in the central region (high fluence). [The length of the horizontal bar is 1 µm in (A), (B), and 2 µm in (C).]. Reproduced with permission from U. Chakravarty, R.A. Ganeev, P.A. Naik, J.A. Chakera, M. Babu, P.D. Gupta, Nanoripple formation on different band-gap semiconductor surfaces using femtosecond pulses, J. Appl. Phys. 109 (8) (2011) 084347, AIP Publishing.

Fig. 1.2 shows the nanoripple formation using linearly polarized laser pulses. In the narrow bandgap semiconductors such as GaAs and InP, the scanning electron microscope (SEM) pictures of the irradiated spot show the spacing to be of the order of 500–600 nm. On the contrary, in WBG materials such as GaN and SiC, the spacing is of the order of 170–270 nm. Thus it was observed that, in general, the narrow nanoripples are formed from material with a WBG. This observation of narrow ripple formation in WBG semiconductor is consistent with previous reported results by Borowiec et al. [19], but no explanation of the physical processes involved was given in those experiments. Fig. 1.3 shows different microstructure formations using different fluences and ambient conditions for narrow bandgap semiconductors such as GaAs. The SEM picture of the irradiated spot, as seen in Fig. 1.3A, shows that, at a high fluence in air, no nanoripple formation takes place, and only random heterogeneities of a few microns order appear. Fig. 1.3B shows formation of nanoripples of a typical size of 600 nm at low fluence (in air). Interestingly, Fig. 1.3C shows formation of nanoholes at high fluence in water. Fig. 1.3D shows narrow nanoripples of 150 nm size in GaAs at low fluence in water. This shows the crucial role of the ambient medium in formation of the nanoripples. Very narrow nanoripples were observed in narrow bandgap materials such as GaAs in water, whereas in air, the corresponding nanoripple period was of the order of the laser wavelength. Even for a WBG material such as GaP, which

3

4

CHAPTER 1  Periodic nanoripples formation on the semiconductors

FIGURE 1.2  Nanoripple Formation Using 800 nm Pulses in Narrow Bandgap Semiconductors: (A) GaAs and (B) InP; and in wide bandgap semiconductors: (C) GaN and (D) SiC [The length of the horizontal bar is 1 µm in (A), (B), (C), and (D).]. Reproduced with permission from U. Chakravarty, R.A. Ganeev, P.A. Naik, J.A. Chakera, M. Babu, P.D. Gupta, Nanoripple formation on different band-gap semiconductor surfaces using femtosecond pulses, J. Appl. Phys. 109 (8) (2011) 084347, AIP Publishing.

FIGURE 1.3  Nanoripple Formation Using 800 nm Pulses in GaAs: (A) at high fluence in air, (B) at low fluence in air, (C) at high fluence in water, and (D) at low fluence in water. [The lengths of the horizontal bars correspond to 5 µm in (A), (B), and (C), and 1 µm in (D).]. Reproduced with permission from U. Chakravarty, R.A. Ganeev, P.A. Naik, J.A. Chakera, M. Babu, P.D. Gupta, Nanoripple formation on different band-gap semiconductor surfaces using femtosecond pulses, J. Appl. Phys. 109 (8) (2011) 084347, AIP Publishing.

1.1 Nanoripple formation

shows narrow nanoripples of period ∼200 nm in air, a narrower period of 150 nm was observed during irradiation in water. Fig. 1.4 shows the angular dependence of nanostructure size obtained from narrow (GaAs) and wide (GaP) bandgap materials irradiated by 800 nm pulses in water. It was observed that, for the narrow bandgap (GaAs) semiconductor, the ripple size increased with increasing angle of incidence, as shown in Fig. 1.4A. In the case of the WBG (GaP) material with increasing angle of incidence, the ripple period first increases and becomes constant at higher angles, as shown in Fig. 1.4B. It was mentioned earlier in the context of Fig. 1.3C that, at high fluence, nanoholes are formed in GaAs. Fig. 1.4C shows that the size of the holes increases with increasing the angle of incidence of the incident p-polarized light. In many photonic applications, for example, two-dimensional photonic crystals and micro-diffraction elements, the nano/micro holes on the surface of bulk materials are required as the building blocks. Although the mechanism of formation of such structures is still not clear, this process was achieved during irradiation in water medium, thus emphasizing the role of dense surrounding medium in formation of such novel hole structures. It is commonly accepted that the nanoripple width is dependent on the wavelength of the incident light. Therefore to obtain narrow nanoripples, second harmonic of the fundamental laser radiation is a better choice. Fig. 1.5 shows the influence of laser wavelength on nanoripple formation. Fig. 1.5A and B show that, in the case of SiC, the use of 400 nm wavelength reduces the ripple period by more than a factor of two as compared to the nanoripples produced by the 800 nm pulses (from 190 nm to about 90 nm). Similarly, for a narrow bandgap material such as InP, there is a reduction of ripple period from 620 nm to about 280 nm, as shown in Fig. 1.5C and D. It may be noted that, for GaP, the ripple period using 400 nm pulse becomes larger than that formed using 800 nm pulses (from 180 nm to about 300 nm).

1.1.2  DISCUSSION OF EXPERIMENTAL RESULTS AND DESCRIPTION OF THE MODEL The consistent observations regarding the nanoripple formation are as follows: (1) the ripple period is larger at higher fluence, (2) the ripple period is large for the materials having bandgap narrower than the incident photon energy compared to that for materials having a wider bandgap, and (3) a denser ambient medium results in the formation of narrower ripple period. The plausible explanation for such observations is consistent with the theory that ripple formation is due to excitation of the surface plasmon at the semiconductor surface by the incident laser light with the rough target surface. The surface plasmon is a well-known phenomenon in the context of the metals, which have free electrons for excitation of this longitudinal charge density oscillation. However, in the case of semiconductors, the origin of free electrons is because of the excitation of free electrons due to multiphoton ionization by the laser. The typical intensity of 2 × 1012 W cm−2 is just sufficient to start plasma formation for an ultrashort (45 fs) laser pulse, wherein multiphoton ionization is the dominant process. These laser-generated free electrons give a transient metallic character to

5

6

CHAPTER 1  Periodic nanoripples formation on the semiconductors

FIGURE 1.4  Angular Dependences of Formation of Various Nanostructures Using 800 nm Pulses in Water. (A) Nanoripple formation in GaAs, (B) Nanoripple formation in GaP, and (C) Nanohole formation in GaAs. Reproduced with permission from U. Chakravarty, R.A. Ganeev, P.A. Naik, J.A. Chakera, M. Babu, P.D. Gupta, Nanoripple formation on different band-gap semiconductor surfaces using femtosecond pulses, J. Appl. Phys. 109 (8) (2011) 084347, AIP Publishing.

1.1 Nanoripple formation

FIGURE 1.5  Nanoripple Formation Using 800 nm Pulses and 400 nm Pulses Respectively in Wide Bandgap Semiconductors: SiC (A) and (B), GaP (E) and (F), and in narrow bandgap semiconductor InP (C) and (D). [The length of the horizontal bar is 2 µ m in (A), 500 nm in (B), 5 µ m in (C), 1 µm in (D), 1 µm in (E), and 2 µm in (F).]. Reproduced with permission from U. Chakravarty, R.A. Ganeev, P.A. Naik, J.A. Chakera, M. Babu, P.D. Gupta, Nanoripple formation on different band-gap semiconductor surfaces using femtosecond pulses, J. Appl. Phys. 109 (8) (2011) 084347, AIP Publishing.

the molten surface, which can now support the surface plasmon. During the initial few shots, the surface roughness is enhanced due to irradiation by the high-intensity pulses and formation of random nanostructures. The subsequent shots fired on the roughened surface lead to more efficient excitation of surface plasmon. The molten material assumes the shape of a grating, which satisfies the relation between wave vectors of the incident laser (which depends on its wavelength, the angle of incidence, and the refractive index of the ambient medium) and the surface plasmon (which depends on the electron density of the molten surface). This grating gets frozen once the material cools down as soon as the laser pulse ends giving rise to the observed nanoripples. The wave vector of this grating satisfies the following relation, as per the momentum conservation [18] G = k i − ks , (1.1)

7

8

CHAPTER 1  Periodic nanoripples formation on the semiconductors

where ki is the wave vector of incident laser and ks is the wave vector of surface plasmon. By substituting the expression for k's in terms of corresponding wavelengths, and |G| as 2π/d (where d is the ripple period), one gets the expression for d as [18] λL (1.2) d= λL ± µ sin θ λS

Here λL, λS, θ, and µ are the incident laser wavelength (in vacuum), the surface plasmon wavelength, the incident angle, and the refractive index of the ambient medium, respectively. It is clear from Eq. (1.2) that for normal incidence (i.e., θ = 0) d = λS. Thus at normal incidence the ripple period d is equal to the surface plasmon wavelength, which is obtained from its dispersion relation as λL λS = (1.3) εε m ε + εm

Here ε is dielectric constant of the surface plasma created by the intense femtosecond pulse, and εm is the dielectric constant of the ambient medium (i.e., εm = µ2; 1 for air and 1.76 for water). Because the multiphoton ionized surface has free electrons, its dielectric constant can be taken similar to a plasma, that is, (neglecting the collision term frequency) n ε = 1− e (1.4) nc

where ne is the surface plasma free electron density, and nc is the critical density corresponding to the laser wavelength. Putting the expression of ε in Eq. (1.3) and then using Eq. (1.2), one gets for the case of normal incidence λL (1.5) d= n ε m 1− e n c ne ε m + 1− n c

(

(

)

)

From Eq. (1.5), it is clearly seen that the ripple period is related to the incident laser wavelength, electron density of the surface plasma, and the dielectric constant of the ambient medium. It is also evident from the expression that real values of d will be obtained only if ne < nc or ne > ( ε m + 1) nc (1.6) When this condition given by Eq. (1.6) is not satisfied, the nanoripples are not formed. Fig. 1.6A shows the variation of nanoripple period in air and water with electron density plotted using Eq. (1.5) for 800 nm wavelength. Fig. 1.6B shows the same curve for 400 nm wavelength.

1.1 Nanoripple formation

FIGURE 1.6  Nanoripple Period as a Function of the Electron Density for Air (Dotted Line) and Water (Continuous Line), using (A) 800 nm pulses and (B) 400 nm pulses. Reproduced with permission from U. Chakravarty, R.A. Ganeev, P.A. Naik, J.A. Chakera, M. Babu, P.D. Gupta, Nanoripple formation on different band-gap semiconductor surfaces using femtosecond pulses, J. Appl. Phys. 109 (8) (2011) 084347, AIP Publishing.

A few conclusions can be made from these figures. Two branches of solutions exist for the nanoripples; one is the super-wavelength nanoripple, and the other is the subwavelength nanoripple. The super wavelength can be formed if the electron density is below the critical density, but such structures are not observed primarily due to two reasons; the first one is that being a low electron density process the plasmon excitation is very weak, and the second one is that the long period structure will have lower groove depth, so that any short period structure may overshadow the long period one. The third possible reason could be that at such low fluence, the material simply does not melt to reorganize, thereby precluding any possibility of ripple formation. The subwavelength nanoripples are formed if the electron density is greater than (εm + 1)nc. The nanoripple period is equal to laser wavelength if the electron density is high, and a very sharp decrease in the

9

10

CHAPTER 1  Periodic nanoripples formation on the semiconductors

ripple period is observed when the electron density starts approaching (εm + 1)nc. If water is the ambient medium, then the nanoripple period is smaller for the same electron density. Further, the graph also brings out the critical role of electron density in formation of large and narrow ripple formation. Because the electron density generating surface during femtosecond laser pulse irradiation depends on the laser parameters such as fluence and wavelength, a drastic change in the nanoripple period is expected on a small variation of these parameters. Now let us explain the experimental observations in terms of the aforementioned formulation. The observation from Fig. 1.1 regarding narrow ripple of 200 nm width at the periphery of the focal spot and wider 600 nm ripple width in the central hot spot region can be explained as follows. Because the electron density of the laser irradiated surface is proportional to the incident laser fluence, higher intensity of irradiation in the central region of the focal spot will generate a higher electron density leading to formation of wider nanoripples of 600 nm spacing, whereas, at the region of the edges of focal spot, a lower fluence generates a lower electron density, which leads to a narrower (∼200 nm) ripple formation. It is also seen from Fig. 1.6, that a slight change in the electron density can cause a drastic change of the ripple period, and this explains the observation of the large difference in the ripple period in two spatial regions. The abrupt change in the ripple period is because, as explained earlier, it is very critically dependent on the electron density (see Fig. 1.6). At high intensity, ripples slightly smaller than the laser wavelength are formed, and at mid-range intensity or lower intensity, the ripple period can become very narrow depending on the free electron density generated on the surface. At very low intensities, no rippling takes place, as explained earlier. Of course, the semiconductor material chosen, its bandgap and material breakdown property, and irradiation intensity decide the electron density generated. The main observation that narrow LIPSS are formed on WBG material can also be explained in a similar way. When a femtosecond laser irradiates a narrow bandgap material semiconductor surface, more free electrons are likely to be available in comparison to the case if the WBG material is irradiated by the same laser. This is because in addition to the multiphoton excitation of electrons, the single photon absorption process will also allow the excitation of free electrons, provided the incident photon energy is larger than the bandgap. Therefore, the absorption of the ultrashort pulses in narrow bandgap semiconductor material is through a combination of linear and nonlinear absorption processes leading to a larger availability of free electrons in the conduction band. Therefore as observed from Fig. 1.2, the ripple period is large (∼600 nm) in the case of narrow bandgap materials, as a higher electron density [ne >> (εm + 1)nc] leads to formation of larger width nanoripples. On the other hand, as the generated electron density for WBG material is low as single photon absorption is not possible, narrow ripples are formed in this case (provided the incident laser intensity is low).

1.1 Nanoripple formation

Next, we discuss the nanoripple formation in GaAs in air and water. The ambient medium has substantial effect on the nanoripple formation as expected from Eq. (1.2). This was experimentally observed in Ref. [20] for methanol as the surrounding medium. The discussed experiments in air and water were carried out at two fluences. At higher fluence, no nanoripple formation takes place. As seen in Fig. 1.3A, only random heterogeneities of few microns order appear. At low fluence, subwavelength nanoripples of 600 nm period are formed (Fig. 1.3B. Such micron-order microstructures have also been observed in Ref. [21]. They observed that, as the laser fluence increases, the ripples become disordered due to the enhancement of the thermal effects. Also, as the laser fluence becomes high, the surface morphology evolves and the degree of ripple irregularity becomes large, resulting in the dominant shift of spatial scales of the laser-induced structures from a few hundred nanometers to a few micrometers. Next, in water, a drastic reduction of the nanoripple period is recorded from 600 nm in air to about 150 nm in water. Such a reduction in nanoripple period is expected from Fig. 1.6 as already discussed. An interesting observation is that at high fluence, irradiation of GaAs in water leads to nanohole formation. The reason of nanohole formation is still not understood, although it seems to be like a self-organization process due to surface tension relaxation of strained molten GaAs against another liquid (in this case, water). The experiment was repeated at low fluence for GaP (wide bandwidth material), which shows 200 nm nanoripple period in air and 150 nm in water. Many groups have done LIPSS width study with different incident angles of the laser [22,23]. There are contradictory observations regarding this, as some have reported increase of period with increase in angle of incidence [22], whereas, some have shown decrease in the ripple period with angle of incidence [23]. In the discussed experiment, an overall increase in ripple period with the angle of incidence for GaAs in water was recorded as shown in Fig. 1.4A. GaP in water also shows an increase of ripple period from 150 nm to 300 nm, as the angle is increased from normal incidence to an angle of 30 degree. However, the ripple period saturates at larger angles as seen from Fig. 1.4B. These trends can be partially explained from Eq. (1.2) if one takes the “minus” sign. In that case, the denominator decreases in magnitude as the angle of incidence in increased, leading to increase in the ripple period. Although this explanation holds true for GaAs which show a continuous increase of ripple period with increasing angle, the same is not correct for GaP where ripple period shows saturation at larger angles. This shows that it is important to consider other factors that contribute to the variation of ripple period with angle of incidence. The most crucial among them is perhaps the electron density of the surface plasma created after irradiation of the semiconductor. The electron density generated, in turn, depends on the laser irradiation conditions (fluence, laser wavelength, etc.), material properties (bandgap, melting point, etc.), and the absorption of laser energy in the material through linear (single photon absorption) and nonlinear processes (multiphoton absorption). As the angle of

11

12

CHAPTER 1  Periodic nanoripples formation on the semiconductors

incidence is increased, the circular focal spot gets elongated and becomes elliptical, leading to a decrease in the laser fluence. For example, at an incident angle of 45 degree, the fluence becomes about 70% of the fluence at normal incidence. Therefore as the angle is increased, the fluence decreases, resulting in the decrease of electron density of the surface plasma. This should lead to a decrease in the ripple period. However, as the angle of incidence is increased, the laser energy absorption also increases (up to the Brewster angle). The increase in the absorption is because the reflectivity of the p-polarized light keeps decreasing as the angle of incidence increases and approaches the Brewster angle. The range of angles in discussed studies was below the Brewster angle of the semiconductor (the Brewster angle of semiconductors is as high as 70–80 degree). This decreasing reflectivity leads to an enhanced absorption and, hence, a higher electron density, which should result in generation of larger period ripple. Therefore resultant ripple period is governed by the aforementioned two competing processes of electron generation, one being fluence and the other being absorption. On increasing the angle of incidence, the fluence reduces and less electron generation is expected, but on the other hand, on increasing the angle, more laser light absorption will occur resulting in more electron generation. The relative dominance of the electron generation process, namely, laser energy absorption or laser fluence, determines whether as a function of angle the electron density decreases or increases. One can now explain the observation that in the case of GaP the LIPSS period initially increases and then becomes nearly constant at larger angles (Fig. 1.4B). It seems that in the used experimental conditions, the laser energy absorption and the laser fluence compete against each other equally. The generated electron density is such that the ripple period becomes constant at higher angles for GaP. The monotonic increase of ripple period with the angle of incidence in the case of GaAs (Fig. 1.4A) implies that the electron density in this case is continuously increasing on increasing the angle. This means that electron generation is dominated by absorption process and the fluence has comparatively less influence in the case of GaAs. Had fluence been the more important parameter, the electron density would be reduced on increasing angle (due to the decrease of fluence), and the nanoripple period would have decreased. Further, it may be noted that for s-polarized light, the period should be independent of the angle of incidence, as the magnitude of the laser k vector in the direction of the surface plasmon remains unchanged in this case. However, as the angle increases, due to increasing reflectivity (due to s-polarization), the absorption of the laser light will keep decreasing with angle, leading to lower electron density and corresponding decrease in ripple period. Interestingly, the nanoholes formed on GaAs (at high laser fluence, in water) also showed increase for the hole diameter for higher angles of incidence (Fig. 1.4C). Further experiments are needed to understand generation mechanism of nanoholes. The role of the laser wavelength in changing the LIPSS period is seen from Fig. 1.5. As follows from Eq. (1.5), the ripple period is expected to decrease by a factor of two as one goes from fundamental to the second harmonic laser beam, while other conditions remain the same. This is precisely what is seen in Fig. 1.5A for SiC

1.1 Nanoripple formation

[a WBG (3.37 eV) material for both wavelengths] and Fig. 1.5B for InP [a narrow bandgap (1.35 eV) material for both wavelengths]. In both cases, there is a reduction in ripple period by a factor of two (190 nm to 90 nm and 620 nm to 280 nm). The observation in the case of GaP is just the opposite. Here the ripple period is observed to double (180–300 nm) instead of becoming half. This is because GaP has a bandgap energy of 2.3 eV, which is wide for 800 nm radiation (1.56 eV), but narrow for 400 nm radiation (3.12 eV). So, as a WBG material, GaP shows a narrow ripple period (180 nm) for the fundamental (800 nm), and as a narrow bandgap material, it shows larger ripple period (300 nm) for the second harmonic radiation (400 nm). The aforementioned correlation of the bandgap with ripple period becomes more obvious in Fig. 1.7. The vertical dotted line in Fig. 1.7A is the photon energy of the

FIGURE 1.7  Observed Nanoripple Period as a Function of Bandgap. The dotted line indicates the incident laser photon energy for (A) 800 nm pulses and (B) 400 nm pulses. Reproduced with permission from U. Chakravarty, R.A. Ganeev, P.A. Naik, J.A. Chakera, M. Babu, P.D. Gupta, Nanoripple formation on different band-gap semiconductor surfaces using femtosecond pulses, J. Appl. Phys. 109 (8) (2011) 084347, AIP Publishing.

13

14

CHAPTER 1  Periodic nanoripples formation on the semiconductors

800 nm laser, and for Fig. 1.6B, the dotted line represents the photon energy of the 400 nm laser radiation. From these two graphs, it is clear that if the bandgap of the semiconductor material is narrower than the incident photon energy, the ripple width formed is slightly less than, but of the order of the laser wavelength. For the materials having a bandgap larger than the energy of the laser photon, the ripple period is much smaller (typically of the order of 1/4th laser wavelength). The reason for this difference has been already explained earlier in terms of single photon and multiple photon ionization of the molten surface. Although most of the observations of LIPSS formation could be explained from the aforementioned formulation, there could be other parameters, which are responsible for finer variations of different nanoripple formation of different materials. We have neglected the material parameters such as melting point, conductivity of semiconductors, charge mobility, surface tension, and viscosity of the molten material of the surface, which may also be very critical in deciding the resultant structure. For example, melting is one of the primary requirements of restructuring the surface. To see if the melting point plays a deciding role in ripple period, the ripple period has been plotted in Fig. 1.8 as a function of melting point for various materials, at a constant laser fluence. There is clear trend that shows materials having a high melting point forming narrow ripples. Fig. 1.8, seen in isolation, would give an impression that melting point decided the ripple period. However, when seen with Fig. 1.7B for the ripples formed with second harmonic laser beam, it becomes clear that looking at GaP ripple period, it is the bandgap (relative to the laser photon energy) that decided the ripple period, and not the melting point, although melting is crucial in nanoripple formation.

FIGURE 1.8  Observed Nanoripple Period as a Function of Melting Point for Various Semiconductors. Reproduced with permission from U. Chakravarty, R.A. Ganeev, P.A. Naik, J.A. Chakera, M. Babu, P.D. Gupta, Nanoripple formation on different band-gap semiconductor surfaces using femtosecond pulses, J. Appl. Phys. 109 (8) (2011) 084347, AIP Publishing.

1.2 Nanosecond laser-induced periodic surface structures formation

1.2  NANOSECOND LASER-INDUCED PERIODIC SURFACE STRUCTURES FORMATION ON WIDE BANDGAP SEMICONDUCTORS USING NANOSECOND ULTRAVIOLET PULSES As the early observation of LIPSS on semiconductors [24], this kind of nanostructures has been imprinted on almost all kinds of materials and has been extensively investigated using low-power cw and pulsed laser sources of nanosecond (ns) and femtosecond (fs) duration [25–28]. In general, the ripples have a period d dependent on laser wavelength λ, on the angle of incidence of the radiation θ, and on index of refraction n and can be described by the relation d = λ/(n − sinθ) [29]. After exposure of a smooth solid to a linearly polarized radiation at normal incidence, often the lateral period of the fabricated LIPSS is very close to the wavelength of the incident radiation. It has been proposed that this type of ripples arises from optical interference effects due to the superposition of the incident radiation with a surface electromagnetic wave, which is created at the material–medium interface during irradiation together with a feedback mechanism. LIPSS resulting from femtosecond laser irradiation of solids have received considerable attention in attempts to determine their formation mechanism (see Section 1.1.1 and Refs. [30–33]). Irradiation of surfaces at normal incidence usually leads to the formation of low spatial frequency LIPSS (LSFL) with period comparable to the laser wavelength. This type of structures is explained in reference to the previously mentioned interference mechanism. In the case of metals, semiconductors, and dielectrics, the formation of ripple structures with subwavelength periods has also been observed. These high spatial frequency LIPSS (HSFL) have been obtained using femtosecond laser pulses of different duration, wavelength, fluence, and number of pulses [34–37]. Several mechanisms have been proposed as the origin of HSFL, as was mentioned in the previous section. WBG semiconductors have expanded the scope of applications beyond those of silicon. The developing list of such materials for use in device production is remarkable and continues to provide new design possibilities. The inherent properties of WBG make them ideal candidates for high-power, high-temperature electronic devices, power amplifiers, switches, and short wavelength light sources. Therefore modification of structure and properties of WBGs at the nanometer scale attracts great interest [38]. In the studies reviewed in the forthcoming paragrphs [39], LIPSS were imprinted on the surface of WBG semiconductor wafers of indium phosphide (InP), gallium arsenide (GaAs), gallium phosphide (GaP), and silicon carbide (SiC) by irradiating in air with linearly polarized, 266 nm, 6 ns laser pulses. The period and amplitude of the LIPSS were characterized by atomic force microscopy (AFM) as a function of the laser fluence and number of pulses. It was observed that, as the bandgap of the semiconductor material increases, higher fluence or number of pulses are needed for LIPSS formation, whereas the amplitude of the ripples is related with the optical and

15

16

CHAPTER 1  Periodic nanoripples formation on the semiconductors

Table 1.1  Initial Roughness (Ra), Energy Bandgap [41], Melting Temperature [41] (Tm), Specific Heat [41] (c), Thermal Conductivity [41] (k), Density [41] (ρ), and Linear Absorption Coefficient at 266 nm [40] (α) of Semiconductor Wafers. Material

Bandgap Ra (nm) (eV) Tm (°C)

c k (J kg−1 K) (W mK−1)

ρ (kg m−3) α (cm−1)

InP GaAs GaP SiC

0.30 0.28 0.62 0.65

310 330 430 690

4810 5316 4138 3210

1.35 1.42 2.30 3.37

1060 1240 1457 2730

68 55 110 370

1.47 × 106 1.72 × 106 1.21 × 106 2.44 × 106

Reprinted from M. Sanz, E. Rebollar, R.A. Ganeev, M. Castillejo, Nanosecond laser-induced periodic surface structures on wide band-gap semiconductors, Appl. Surf. Sci. 278 (2) (2013) 325–329, with permission from Elsevier.

thermal penetration depth. Estimations of surface temperature increase are discussed with reference to the WBG semiconductor's electrical, optical, and thermal properties.

1.2.1  EXPERIMENTAL SETUP For studying ripple formation, multiple pulse laser irradiation of undoped semiconductor wafers of InP, GaAs, GaP, and SiC was carried out in ambient air at normal incidence. The electrical, optical, and thermal properties of those materials [40,41] are summarized in Table 1.1. For irradiating the samples, the linearly polarized fourth harmonic output of a Q-switched Nd:YAG laser (266 nm, λ = 6 ns, 10 Hz repetition rate) was used. This irradiation wavelength corresponds to an energy of 4.67 eV, well above the bandgap energies of these semiconducting materials (Table 1.1). The ­central (4 mm diameter) most uniform part of the beam spot was selected for irradiation by using a diaphragm. The laser beam was focused on the substrate surface with a spherical lens of 15 cm focal length. The irradiation fluence was below the ablation threshold (Fth) for a single pulse of each material. The fluence was determined as the ratio of the laser pulse energy, measured in front of the sample with a joulemeter, and the area of the irradiated spot. Ablation thresholds of samples were determined by measuring the minimum single pulse energy necessary to yield a surface change as detected by optical microscopy using a 160× microscope objective equipped with a charge coupled device (CCD) camera. The obtained values were 190, 380, 470, and 950 mJ cm−2 for InP, GaAs, GaP, and SiC, respectively. The morphology of the lasertreated semiconductor surfaces was characterized using AFM. The AFM measurements were performed in 5 different positions of each sample to check the uniformity of the fabricated nanostructures. The pristine substrates, of around 300 µm thick, present a flat surface, with mean roughness (Ra) values <1 nm. Ra values are listed in Table 1.1 and indicate the arithmetic average of the deviations in height from the center plane of the sample. Each Ra value corresponds to the average of three independent measurements in different locations of the substrate surface.

1.2 Nanosecond laser-induced periodic surface structures formation

Table 1.2  Minimal Fluence (Fm) Required for LIPSS Fabrication and Experimental Conditions (Fluence and Number of Pulses) and Ripple Properties (Period and Amplitude) for Optimum LIPSS of Semiconductor Wafers. Material

Fm (mJ cm−2)

Fluence (mJ cm−2)

Number of Pulses

Ripple Period (nm)

Ripple Amplitude (nm)

InP GaAs GaP SiC

100 135 125 >300

125 150 125 –

200 200 300 –

248 ± 6 253 ± 7 263 ± 10 No LIPSS observed

5 ± 1 9 ± 2 15 ± 5

Reprinted from M. Sanz, E. Rebollar, R.A. Ganeev, M. Castillejo, Nanosecond laser-induced periodic surface structures on wide band-gap semiconductors, Appl. Surf. Sci. 278 (2) (2013) 325–329, with permission from Elsevier.

1.2.2  NANORIPPLE FORMATION BY NANOSECOND PULSES Irradiation of the semiconductor wafers was performed at different fluences and number of pulses to find the conditions for obtaining the most uniform ripples in terms of period and amplitude. The minimum fluence (Fm) needed for LIPPS fabrication is displayed in Table 1.2 together with the experimental conditions for the optimum LIPSS fabrication for each semiconductor wafer. Fig. 1.9 shows AFM height images and corresponding cross-section of the LIPSS obtained in InP, GaAs, and GaP. Fig. 1.9A displays the ripples fabricated in InP with 200 pulses of 125 mJ cm−2. Ripples were perpendicular to the laser polarization direction, and the measured period was estimated to be 248 nm with amplitude of 5 nm. The structures were observed at fluences of 100–150 mJ cm−2 with 100–300 pulses. In the case of GaAs, the optimum conditions for LIPSS formation of InP (125 mJ cm−2, 200 pulses) resulted in isolated rounded nanostructures (Fig. 1.10A). When increasing the number of pulses, the ripples started to form and align (Fig. 1.10B), whereas the increase of fluence finally results in LIPSS formation (Figs. 1.9B and 1.10C). Further fluence increase induced the destruction of the nanostructures (Fig. 1.10D). The best ripples were obtained with 200 pulses at 150 mJ cm−2. Under these conditions, the estimated period was 253 nm and the amplitude 9 nm. For GaP, a number of pulses higher than that used for GaAs are necessary to fabricate uniform LIPSS. The best ripples, with a period of 263 nm and amplitude of 15 nm, were obtained at 125 mJ cm−2 and 300 pulses (Fig. 1.9C). Increasing the fluence and/or the number of pulses caused the disappearance of the uniform ripples. In the case of SiC, no LIPSS were obtained for fluences as large as 300 mJ cm−2 and for a large number of pulses (up to 600).

17

18

CHAPTER 1  Periodic nanoripples formation on the semiconductors

FIGURE 1.9 AFM height images (2 × 2 µm2 size) (left) and corresponding cross-sections (right) of LIPSS fabricated using 266 nm radiation in (A) InP with 200 pulses at a fluence of 125 mJ cm−2, (B) GaAs with 200 pulses at 150 mJ cm−2, and (C) GaP with 300 pulses at 125 mJ cm−2. The ripples are perpendicular to the laser polarization direction. Reprinted from M. Sanz, E. Rebollar, R.A. Ganeev, M. Castillejo, Nanosecond laser-induced periodic surface structures on wide band-gap semiconductors, Appl. Surf. Sci. 278 (2) (2013) 325–329, with permission from Elsevier.

1.2 Nanosecond laser-induced periodic surface structures formation

FIGURE 1.10  AFM Height Images (3 × 3 µm2 size) of LIPSS Fabricated Using 266 nm Radiation in GaAs at the Indicated Conditions. The ripples are perpendicular to the laser polarization direction. Reprinted from M. Sanz, E. Rebollar, R.A. Ganeev, M. Castillejo, Nanosecond laser-induced periodic surface structures on wide band-gap semiconductors, Appl. Surf. Sci. 278 (2) (2013) 325–329, with permission from Elsevier.

1.2.3  DISCUSSION OF NANOSTRUCTURE FORMATION As mentioned, the mechanism of LIPSS formation can be explained as the result of the optical interference effects due to the superposition of the incident radiation with a surface electromagnetic wave, which is created and scattered along the irradiated surface [42]. This results in a modulated distribution of energy on the surface, which consequently induces a similarly modulated heating. The thermal penetration depth of the irradiated zone can be calculated by dth = (Dτ)0.5 with D = k/ρc being the thermal diffusivity, k the thermal conductivity, ρ the density, c the specific heat of the material (Table 1.1), and τ the pulse duration (6 ns). For the materials under study, dth is below 1 µm, thus much smaller than the diameter of the irradiated area. Therefore for estimating the temperature increase of the irradiated substrate, one can assume for simplicity that the temperature distribution is associated with depth x and time t. It is also assumed that, during the irradiation time interval, the material parameters and the irradiation intensity are constant. Fig. 1.11 shows for the materials studied herein the temporal evolution of the estimated surface temperature upon irradiation with a single laser pulse of 266 nm at 125 mJ cm−2. The maxima of the curves correspond to the maximal temperature attained in the surface of the different samples.

19

20

CHAPTER 1  Periodic nanoripples formation on the semiconductors

FIGURE 1.11  Time Dependence of the Temperature Reached on the Semiconductor Surface Under Irradiation With a Single Pulse of 266 nm at 125 mJ cm−2, for the Different Materials, as Indicated. Reprinted from M. Sanz, E. Rebollar, R.A. Ganeev, M. Castillejo, Nanosecond laser-induced periodic surface structures on wide band-gap semiconductors, Appl. Surf. Sci. 278 (2) (2013) 325–329, with permission from Elsevier.

The optimum conditions for LIPSS formation can be related with the surface temperature reached upon irradiation and with the semiconductor melting point. For InP and GaAs, the temperature attained at the surface is slightly below the corresponding melting point (Table 1.1). In the case of GaP, the temperature of the irradiated surface is clearly below the melting point. For SiC, and due to the high values of specific heat c and thermal conductivity k of this material, higher fluences than those explored in discussed study are expected to be required to melt the outer sample layer. Those results indicate that to obtain LIPSS a minimum fluence is necessary to assure that the surface temperature is high enough for allowing melting and material rearrangement. On the other hand, material evaporation produces emission of atoms from the semiconductor surface. Evaporation is the origin of microdefects that raise the surface roughness and enhance the absorption. This in turn increases the surface heterogeneities, which facilitate the feedback mechanism necessary for the ripple formation. The relative increase of fluence and number of pulses needed for imprinting optimum LIPSS on the semiconductors are observed to rise as the energy bandgap increases. This is related to the fact that higher melting temperatures correspond to wider bandgaps (Table 1.1) and thus higher temperatures should be reached upon irradiation to allow melting and rearrangement of material. Measured ripple periods are of the order of the irradiation wavelength for all analyzed semiconductors, whereas a moderate increase in amplitude is observed as the optical and thermal penetration depths of the material increase. The optical absorption depth for a single pulse, calculated as the inverse of the linear optical absorption coefficient (Table 1.1), is larger for GaP than for InP and GaAs. The thermal penetration depth dth is also higher for GaP (610 nm) than for InP and GaAs (around 450 nm)

1.3 Fabrication of two-dimensional periodic nanostructures

for the given irradiation conditions. The larger optical and thermal penetration depths induce a deeper effect close to the surface in GaP as compared to InP and GaAs, in good agreement with the observed ripple amplitude values: the ripple amplitude for GaP is 15 nm, whereas 5 nm amplitude ripples are obtained in the case of InP.

1.3  FABRICATION OF TWO-DIMENSIONAL PERIODIC NANOSTRUCTURES BY TWO-BEAM INTERFERENCE OF FEMTOSECOND PULSES 1.3.1  2D AND 3D STRUCTURES There has been considerable interest in the fabrication of two- and three-dimensional (2D and 3D) PCs, which consist of artificial periodic structures [43–48]. Various techniques were used to fabricate these periodic structures such as self-assembly of colloidal particles, direct laser writing, and holographic lithography (HL), etc. HL is a highly useful technique to fabricate periodic patterns in photosensitive materials by the interference of several coherent laser beams. One can fabricate various 2D and 3D periodic structures by changing the number of laser beams (usually >3) and their arrangements. The periods were larger or nearly equal to the laser wavelength; therefore it is still a challenge to improve the resolution of laser fabrication to obtain PCs with bandgap in visible spectrum. Periodic structures induced by single laser beam have been studied intensively in the last four decades [49]. As already mentioned, it was found that the periods were usually close to the laser wavelength when long pulse (nanosecond and picosecond) laser or cw laser were used. Meantime, nanoripples with periods much less than the laser wavelengths have been fabricated in semiconductors and dielectrics after irradiation by femtosecond laser pulses [14,19,35]. LIPSS with periods of 40–500 nm have been fabricated by using lasers at wavelengths of 267–2000 nm. Further studies of the formation of regular short periodic nanogratings on the surface of semiconductor crystals, particularly ZnO irradiated by two collinear laser beams, were described in Ref. [50], which suggested that the short periodic nanostructures have potential applications in optical recording and PCs fabrication. ZnO is an important compound semiconductor with a WBG of 3.37 eV and a large exciton binding energy of 60 meV at room temperature. These properties make it a very useful material in optoelectronics. Several groups have reported the fabrication and the applications of various types of ZnO nanostructures such as nanoparticles, nanowires, nanobelts, and nanofilms [51,52]. Below we discuss the fabrication of 2D periodic nanostructures formed on the surface of ZnO crystals applying a method of two-beam interference of femtosecond pulses [53]. This method is based on the two aspects discussed previously: one is the long periodic structures determined by the two-beam interference pattern, another is the short periodic nanostructures induced by single femtosecond laser beam. Furthermore, the formation mechanism is discussed.

21

22

CHAPTER 1  Periodic nanoripples formation on the semiconductors

FIGURE 1.12  Experimental Setup of Two-Beam Interference Used for Fabrication of 2D Periodic Structures. BS, 50% beam splitter; GZ, Glan polarizer along z axis; HF, half-wave plate; L, lens; M, mirror. Reproduced with permission from T. Jia, M. Baba, M. Suzuki, R.A. Ganeev, H. Kuroda, J. Qiu, X. Wang, R. Li, Z. Xu, Fabrication of two-dimensional periodic nanostructures by two-beam interference of femtosecond pulses, Opt. Exp. 16 (3) (2008) 1874–1878, Optical Society of America.

1.3.2  EXPERIMENTS AND DISCUSSION Fig. 1.12 shows the experimental setup for fabrication of 2D periodic structures. Laser pulses at the wavelength of 790 nm with pulse duration of 120 fs were delivered from a commercial Ti:sapphire regenerative amplifier operated at 10 Hz repetition rate. The laser beam was propagated through a half-wave plate and Glan polarizer, and then was split into two beams via a 50% beam splitter. Both of the two beams polarized along z direction. One beam went through a delay line and reached the sample surface at the exact same time with the other one. Zero temporal point was determined by the signal of double frequency via a BBO crystal. The angle between the two laser beams was denoted as 2θ and could be easily changed by rotation of two mirrors of M1 and M2. ZnO crystal plate was 1.0 mm thick, and its surface was normal to the x axis and optically polished. The sample was mounted on a XYZ-translation stage. After laser pulses irradiation, the sample was dipped in ethanol and pure water, and cleaned for 5 min with ultrasonic cleaner, respectively. The periodic structures formed on the sample surface were observed by using SEM. The laser beams were focused with lenses of 250 mm focal length. The sample was transmitted to the position of 0.5 mm in front of the focal plane to enlarge the focus to 110 µm in diameter. SEM images of the periodic structures are shown in Fig. 1.13. The spatial overlapping length was 36 µm. Only in the central part of overlapping area, two laser beams interfered thoroughly. Therefore, the periodic structures were observed on the ablation area of 30 × 100 µm2 (see Fig. 1.13A). Fig. 1.13B and C show the SEM images in the white square marked in Fig. 1.13A and B at higher magnifications, respectively. These photos represented 2D periodic structures. Only one long-dimensional periodic structure was obtained by two-beam

1.3 Fabrication of two-dimensional periodic nanostructures

FIGURE 1.13  SEM Images of 2D Periodic Nanostructures. The total pulse energy and irradiation time were 127 µJ and 6 s in (A–C), 154 µJ and 4 s in (D), and 250 µJ and 10 s in (E), respectively. Reproduced with permission from T. Jia, M. Baba, M. Suzuki, R.A. Ganeev, H. Kuroda, J. Qiu, X. Wang, R. Li, Z. Xu, Fabrication of two-dimensional periodic nanostructures by two-beam interference of femtosecond pulses, Opt. Exp. 16 (3) (2008) 1874–1878, Optical Society of America.

interference method, and the period was determined by the interference pattern described by relation Λ = λ/2sinθ. The laser wavelength λ was 790 nm, and the angle between the two laser beams was 2θ = 35.2 degree, thus the period was estimated to be 1.31 µm. The long period was 1.33 µm (see Fig. 1.13B and C), which was close to the theoretical value. Therefore the long periodic structures are determined by the usual interference pattern of the two laser beams. Besides the long periodic structures, there were short periodic nanostructures (nanogratings) embedding in the long ones, similar to those discussed in the two previous sections. The nanograting was of 1.14 µm wide, and its period was only 250 nm. If the total pulse energy increased to 154 µJ and the irradiation time decreased to 4 s, the nanograting width decreased to 1 µm, and its period increased to 270 nm (shown in Fig. 1.13D). Once the laser conditions adjusted, it was found that the width of the short periodic structure changed in the range of 0.4–1.15 µm while the long period was kept as ∼1.33 µm. If the pulse energy was higher than 240 µJ, the nanogratings were ablated and disappeared. However, on the interval line between each nanogratings, periodic nanospots of 0.5 µm long and 0.25 µm wide appeared (Fig. 1.13E). In one square centimeter in the sample surface, there was about 3 × 108 nanoripples or nanospots, and the short period was only 250 nm. The 2D periodic nanostructures have great potential applications in ultrahighdensity optical record and PCs in ultraviolet and visible light range. The evolution of short periodic nanostrucutures with the increase of laser intensity was studied, and the results are shown in Fig. 1.14. At the edge of ablation area denoted as strip 1, the sample surface was melted slightly. Some small ripples emerged randomly on strips 2–4 with the increase of laser intensity. In strip 5, several groups of nanoripples

23

24

CHAPTER 1  Periodic nanoripples formation on the semiconductors

FIGURE 1.14  The Evolution of Short Periodic Nanoripples With the Increase of Laser Intensity in Each Strip Denoted as 1, 2,…9. The total pulse energy is 153 µJ, and laser irradiation time is 6 s. Reproduced with permission from T. Jia, M. Baba, M. Suzuki, R.A. Ganeev, H. Kuroda, J. Qiu, X. Wang, R. Li, Z. Xu, Fabrication of two-dimensional periodic nanostructures by two-beam interference of femtosecond pulses, Opt. Exp. 16 (3) (2008) 1874–1878, Optical Society of America.

began to emerge. There were two or three parallel nanoripples in one group, and all of these nanoripples were perpendicular to the laser polarization. With further increase of laser intensity, these groups of nanoripples extended and connected each other (see strips 6 and 7). Regular short periodic nanoripples formed in strips 8 and 9. The evolution processes with laser irradiation time were also studied. The two processes were similar to each other. The experiments were conducted to study the formation of nanoripples in ZnO crystal surface induced by single beam of 790 nm laser and found the periods were in the range of 200–240 nm. The results, as shown in Fig. 1.13, were very close to this value. The laser polarization was rotated by 45 degree, and it was found that the orientation of nanogratings embedding in the long periodic structures rotated by 45 degree, too. The above-discussed formation processes were also similar to the cases of single femtosecond laser beam. These three aspects indicated that the formation mechanism of the short periodic nanostructures in the 2D periodic patterns was similar to that induced by single femtosecond laser beam. The interference method of two femtosecond laser beams is rather complicated, and it is not convenient to be used in 3D microfabrication. To overcome this problem, the laser beam was enlarged to more than 40 mm in diameter via a couple of positive and negative lenses, and then a double iris was used to select two laser beams of the same profile, same polarization, and same intensity. These two laser beams were exactly focused on the sample surface. Fig. 1.15 represented the results induced by a Ti:sapphire femtosecond laser operated at 1 kHz repetition rate. Several ablation spots were observed on the sample surface. They were distributed periodically

1.4 Extended homogeneous nanoripple formation

FIGURE 1.15  SEM Images of Periodic Nanostructures Induced by 1 kHz, 130 fs Laser. The pulse energies are 145 µJ in (A–C), and 130 µJ in (D). The laser irradiation time is 1 s in (A). The sample is transmitted at a speed of 0.2 mm s−1 in (B–D). Reproduced with permission from T. Jia, M. Baba, M. Suzuki, R.A. Ganeev, H. Kuroda, J. Qiu, X. Wang, R. Li, Z. Xu, Fabrication of two-dimensional periodic nanostructures by two-beam interference of femtosecond pulses, Opt. Exp. 16 (3) (2008) 1874–1878, Optical Society of America.

and symmetrically, and the period was equal to that expected by λ/2sinθ. Therefore these periodic ablation spots were determined by the interference between two laser beams. On each spot, there were many nanoripples with periods of 230 nm. The nanoripples were in 45 degree direction for the laser polarization rotated by 45 degree (Fig. 1.15C). The sample was transmitted at a speed of 0.2 mm s−1, which allowed obtaining regular 2D periodic structures with periods of 3.3 µm and 250 nm, respectively (Fig. 1.15D). More details on the extended LIPSS formation are presented in the following section.

1.4  EXTENDED HOMOGENEOUS NANORIPPLE FORMATION DURING INTERACTION OF HIGH-INTENSITY FEW-CYCLE PULSES WITH A MOVING SILICON WAFER Surface patterning has a variety of applications [54]. Surface roughening can be beneficial, for instance, when an increased surface area is desired. Such is the case for improving adhesion of other materials and applications where high optical absorption is required [55]. Pointed structures can be used as field emission sources [56]. Texturing can change the hydrophobic properties of surfaces and affect cell attachment in biomedical applications [57]. Femtosecond laser milling can provide various advantages in prototyping and material manufacturing as an alternative to photolithography [58]. Long-range ordering is relevant to the use of nanostructures in electronic materials applications [59]. As was discussed in previous sections, periodic structure formation through the interaction of laser pulses with semiconductors is a precise method of surface

25

26

CHAPTER 1  Periodic nanoripples formation on the semiconductors

modification and has been an important topic of laser-matter studies during the last few decades. In particular, the fabrication of two- and three-dimensional PCs, which consisted of artificial periodic structures, has been reported. Another approach leading to 3D PC structures has been described in Refs. [60,61]. The interest in extended structures has led to studies of different approaches to the modification of semiconductor surfaces (multibeam interference of femtosecond pulses, HL, multiexposure of two-beam interference technique, etc.). An alternative approach is an extension of the area of periodic structure using a moving target technique [62–64]. Among various semiconductors used as the targets for LIPSS formation, silicon has attracted a special attention due to its frequent use in the optoelectronic industry. Most femtosecond laser modification experiments to date have been performed with a photon energy greater than the bandgap energy of silicon (1.11 eV) corresponding to a wavelength of 1120 nm. Most experimental observations of nanoripples were based on the study of structures created during focusing of short laser pulses on the same spot of the semiconductor surface. By the meaning of “short pulse” we assume the commonly used multicycle femtosecond pulses (of 50–200 fs duration). Meanwhile, it would be interesting to analyze application of considerably shorter pulse duration (i.e., of a few optical cycles), which may considerably change the dynamics of ripple formation and maintenance. The area of interest of most previous studies was restricted by the sizes of the beam waist of the focused radiation. In most cases, the LIPSS appeared in the ablated area at the center of the focal spot, with limited nanorippling seen at the edges of the ablated area. Another pattern dominated at stronger ablation levels, when the ripples appeared at the edges of the focal spot, while at the center of the ablated area, the ripples were destroyed, and the chaotically distributed irregular nanostructures were observed. This heterogeneous distribution of nanoripples across the ablated area was induced by the heterogeneous distribution of laser fluence over the focal spot. In most of the experiments, the number of shots on the ablating spot, the intensity and fluence of the laser radiation, and its wavelength were the main variable parameters, which allowed the optimal conditions of LIPSS formation to be determined. It has been shown that short laser pulses are favorable for the creation of nanoripple structures, which has led to the application of laser pulses as short as 25 fs [65]. Even at this short pulse duration, all these observations showed the heterogeneity of LIPSS across the focal spot. This heterogeneity has frequently appeared in the case of longer pulses; however, even in the case of the shortest pulses used, there are no reports showing homogeneous ripple structures across the whole ablated area. Another feature of reported studies was a very small area of observed nanoripples, because almost all previous experiments were carried out using fixed (stationary) ablating samples or fixed position of the ablating spot on the semiconductors surface. However, one of the primary goals of such studies is the formation of extended nanostructures along the whole surface of the material. For this reason, one has to considerably increase the area of the ablating spot, which has typically been in the range of hundreds of micrometers in diameter.

1.4 Extended homogeneous nanoripple formation

All of the aforementioned problems in the formation of extended homogeneous LIPSS have motivated the researchers to search for new approaches in their formation. Below we review LIPSS experiments using extremely short, few-cycle laser pulses (3.5 fs), which interact with the continuously moving silicon sample along the focal plane. We analyze the results of those studies, which have shown a way to produce extended homogeneous nanoripple structures over entire length of the target surface [66].

1.4.1  EXPERIMENTAL ARRANGEMENTS The Ti:sapphire laser, which was used in those studies, provided after first compression the carrier-envelope phase-stabilized pulses of 25 fs duration and energies of up to 0.8 mJ at a repetition rate of 1 kHz. The amplified pulses were focused into a 1 m long differentially pumped hollow core fiber filled with neon (250 µm inner core diameter) [67]. The spectrally broadened pulses at the output of the fiber system were compressed in the second compressor by 10 bounces of double-angle technology chirped mirrors. A pair of fused silica wedges was used to fine tune the pulse compression and compensate for the dispersion of the air and focusing optics. It was confirmed by analyzing the high-order harmonic generation in gas jets, when highest harmonic cutoff was achieved by introducing the glass of the same thickness as the one of focusing lens for nanoripple formation and changing the thickness of fused silica wedges in the second compressor. High-intensity few-cycle pulses (E = 0.2 mJ, τ = 3.5 fs, λ = 760 nm, 1 kHz pulse repetition rate) were typically obtained at the output of this laser [68]. The compressed pulses were characterized with a spatially encoded arrangement for direct electric field reconstruction by spectral shearing interferometry [69]. Part of this radiation was focused using a 400 mm focal length lens to a spot of 200 µm onto the surface of a silicon wafer at normal incidence. The Si wafers with 〈111〉 surface orientation were used for LIPSS formation. The wafers were cleaned in acetone and ethanol before ablation. The intensity and fluence of the laser beam on the ablated surface were maintained at the level of 8 × 1013 W cm−2 and 0.3 J cm−2. The sample was moved perpendicular to the focused beam along the X axis with different velocities (0.1–5 mm s−1) using a motorized translation stage (Fig. 1.16A). The variation of the speed of sample movement allowed to vary the number of overlapping shots on the same spot. The experiments were performed in ambient air. The ablated samples were rinsed in deionized water and cleaned in an ultrasonic bath. The morphology of ablated surfaces at different conditions of laser-surface interaction was analyzed using a SEM.

1.4.2  RESULTS AND DISCUSSION The LIPSS orientation in these studies was always found to be orthogonal to the laser polarization. At a fixed position of the target, nanoripples were formed after 100–500 shots fired on the same spot of the semiconductor, depending on the fluence

27

28

CHAPTER 1  Periodic nanoripples formation on the semiconductors

FIGURE 1.16 (A) Experimental scheme. AP, Ablating pulse; FL, focusing lens; SW, silicon wafer; TS, translation stage. (B–D) SEM images of the ablation of silicon wafer by 3.5 fs, 760 nm, 1 kHz pulse repetition rate Ti:sapphire laser carried out at slow velocity of the translating stage (0.2 mm s−1) and at the fluence of 0.3 J cm−2. Reproduced with permission from R.A. Ganeev, D.Y. Lei, C. Hutchison, T. Witting, F. Frank, W.A. Okell, T.R. Roschuk, S.A. Maier, J.W.G. Tisch, J.P. Marangos, Extended homogeneous nanoripple formation during interaction of high-intensity few-cycle pulses with a moving silicon wafer, Appl. Phys. A 112 (2) (2013) 457–462, Springer.

and intensity of the laser radiation on the surface of the silicon wafer. It was also observed that the number of shots fired on the semiconductor did not have any effect on the ripple period. Most of the previous LIPSS studies were carried out using 50–200 fs pulses. As already mentioned, the ripple formation in those experiments was observed either at the central part of ablated spot or on the periphery of the ablated region. The surface of ablated semiconductors displayed two areas: one, which contained LIPSS, and another, which contained weakly ablated material (in the case of weak ablation of the edges of the irradiated spot) or overablated irregular structures (in the case of strong ablation of the central part of irradiated spot). In the latter case, the LIPSS being formed during initial shots disintegrated into irregular structures, such as nanodots, nanoholes, and nanowires. The silicon wafer was moved at different velocities orthogonally to the axis of the focusing beam propagation. This approach allowed to determine the conditions for homogeneous ripple formation in one set of ablation measurements. Moreover, during those studies, it was possible to produce extended ripples over the entire length of the sample (5 mm), which considerably increased the area of homogeneously formed periodic structures. Fig. 1.16 shows SEM images of an ablated Si wafer when the target was moved at a velocity of 0.2 mm s−1. When the target was moved continuously along the X axis, the number of overlapping shots on the same spot is given by (π/2)1/2 × (w0f/v) [70], where f is the pulse repetition rate (1 kHz in our case), v is the

1.4 Extended homogeneous nanoripple formation

translation speed (0.2 mm s−1 in the case presented in Fig. 1.16), and w0 is the spot radius. We note that ablation of moving targets by longer (150 fs) pulses [70] showed blunt irregular conical structures on the silicon surface. Meanwhile, in the discussed studies, the number of overlapping laser shots on the same spot was calculated to be ∼300, based on an ablated area of ∼0.1–0.2 mm in diameter. At the beginning of irradiation before the silicon wafer started moving, the ablated spot displayed the irregular structures due to multiple shots (see Fig. 1.16B, the large spot at the right side). Then the target motion started. In the case presented in Fig. 1.16, the relatively small target velocity led to “over-ablation” of the target surface caused by multiple shots on the same spot (intensity and fluence were 8 × 1013 W cm−2 and 0.3 J cm−2, respectively). Nanoripples formed during first few tens shots in the central part of the ablation spot were transformed by subsequent shots on the same area into randomly shaped nanodots (Fig. 1.16C). The ripples remained intact only at the periphery of ablation (Fig. 1.16D). The period of those ripples was 500 ± 40 nm. In that case the structures dominating along the whole line of ablation were randomly formed dots of ∼1  µm in size, which were irregularly distributed in the ablation area (Fig. 1.17). The translation of the target allowed the creation of such a line pattern along the whole sample length. This structure can be formed along the entire area of the silicon wafer by moving the sample along both X and Y axes. By adjusting the translation speed, a considerable improvement of the quality of the LIPSS on the Si wafer was achieved. The increase of speed from 0.2 to 1 mm s−1 led to the formation of an almost homogeneous structure along the entire ablation area (i.e., at both the periphery and central area). This homogeneity was observed

FIGURE 1.17  Chaotic Nanodots Formation at the Central Part of Focal Spot During OverHeating of Target Surface at the Velocity of the Translation Stage of 0.2 mm s−1 and at a Laser Fluence of 0.3 J cm−2. Reproduced with permission from R.A. Ganeev, D.Y. Lei, C. Hutchison, T. Witting, F. Frank, W.A. Okell, T.R. Roschuk, S.A. Maier, J.W.G. Tisch, J.P. Marangos, Extended homogeneous nanoripple formation during interaction of high-intensity few-cycle pulses with a moving silicon wafer, Appl. Phys. A 112 (2) (2013) 457–462, Springer.

29

30

CHAPTER 1  Periodic nanoripples formation on the semiconductors

FIGURE 1.18  Ablation of Silicon Wafer Using 3.5 fs, 760 nm, 1 kHz Pulses at Optimal Speed of the Target (1 mm s−1) at a Laser Fluence of 0.3 J cm−2. Reproduced with permission from R.A. Ganeev, D.Y. Lei, C. Hutchison, T. Witting, F. Frank, W.A. Okell, T.R. Roschuk, S.A. Maier, J.W.G. Tisch, J.P. Marangos, Extended homogeneous nanoripple formation during interaction of high-intensity few-cycle pulses with a moving silicon wafer, Appl. Phys. A 112 (2) (2013) 457–462, Springer.

both along the X and Y directions of the ablated surface. Fig. 1.18 shows an example of such ablation carried out along the whole length of a sample. As in the previous case, the ablation started with the static target (see the large ablated spot on the right side of Fig. 1.18A, but once the silicon wafer started moving at a speed of 1 mm s−1, the formation of extended ripples over the entire area of the focused beam was observed (Fig. 1.18B). The homogeneous shape of the ripples was maintained both in the central area of the focused radiation (Fig. 1.18C) and at the peripheries of ablation. These SEMs show well-defined, almost rectangular, ∼350-nm-thick ripples with sharp edges separated from each other by ∼150 nm. The period of LIPSS was ∼500 nm (Fig. 1.18D). The enlarged picture of these sharp-edged ripples is shown in Fig. 1.19A. The number of shots on the same spot at these conditions of target movement was ∼60. The multicycle 40 fs, 780 nm, 2 mJ pulses from another Ti:sapphire laser operating at 1 kHz pulse repetition rate were also used as the ablation source, maintaining approximately the same fluence on the target surface (0.27 J cm−2). The intensity of these laser pulses on the moving target surface was almost ten times smaller (7 × 1012 W cm−2) compared with few-cycle pulse case, while the number of shots on the same spot was in the range of 50–100. The extended LIPSS using the same technique of target movement were produced (Fig. 1.19B). However, the sharpness and homogeneity of the ripples were inferior to those produced using 3.5 fs pulses.

1.4 Extended homogeneous nanoripple formation

FIGURE 1.19 (A) Surface pattern containing regular nanoripples with sharp edges along the entire area of ablation (5 mm) at the conditions presented in Fig. 1.18. (B) LIPSS in the case of ablation using 40 fs pulses at approximately the same fluence as in the case of 3.5 fs pulses. Reproduced with permission from R.A. Ganeev, D.Y. Lei, C. Hutchison, T. Witting, F. Frank, W.A. Okell, T.R. Roschuk, S.A. Maier, J.W.G. Tisch, J.P. Marangos, Extended homogeneous nanoripple formation during interaction of high-intensity few-cycle pulses with a moving silicon wafer, Appl. Phys. A 112 (2) (2013) 457–462, Springer.

The increase in the number of overlapping shots on the same spot led to degradation of ripple structure, with its further disintegration. There is no exact threshold in breaking of ripples in the case of 3.5 fs pulses; however, one can consider the highest speed of 2 mm s−1 as a starting point in extended homogeneous ripples formation and maintenance at used experimental conditions. Also, at the speed of 4 mm s−1 the ripples do not appear due to small amount of overlapping shots on the same spot. The lowest speed at which the ripples start to break is approximately 0.4 mm s−1. We would like to reiterate that all these parameters are relevant with current experimental conditions, when very short pulses are applied for ripple formation. The images of nanoripples at different speeds of target movement are shown in Fig. 1.20. One can clearly see the dynamics of the beginning of nanoripples formation (upper panel, v = 5 mm s−1), extended ripples formation (middle panel, v = 1 mm s−1), and the beginning of their disintegration (bottom panel, v = 0.2 mm s−1). Using an optimized speed of target, the area of extended homogeneous LIPSS in the case of 3.5 fs ablating pulses can be considerably enlarged by further movement of the ablating surface along both the X and Y axes, thus covering a large area of Si wafer with such ripples. No degradation of the LIPSS has been observed in the months after the ablation. The use of cylindrical focusing optics [71] could further increase the patterning speed. The formation of these structures is attributed to the specific features of ablation using few-cycle pulses. As already mentioned, the extended LIPSS were previously formed by moving the target with regard to the focal spot of multicycle pulse. However, none of those studies have demonstrated homogeneous LIPSS along the entire ablation area. The extremely short period of ablation in the case of few-cycle

31

32

CHAPTER 1  Periodic nanoripples formation on the semiconductors

FIGURE 1.20 Dynamics of the beginning of nanoripples formation (upper panel, v = 5 mm s−1), extended ripples formation (middle panel, v = 1 mm s−1), and the beginning of their disintegration (bottom panel, v = 0.2 mm s−1). Reproduced with permission from R.A. Ganeev, D.Y. Lei, C. Hutchison, T. Witting, F. Frank, W.A. Okell, T.R. Roschuk, S.A. Maier, J.W.G. Tisch, J.P. Marangos, Extended homogeneous nanoripple formation during interaction of high-intensity few-cycle pulses with a moving silicon wafer, Appl. Phys. A 112 (2) (2013) 457–462, Springer.

1.5 Concluding comments

pulses is expected to allow the maintenance of LIPSS formation conditions, with only the central part (∼100  µm; see the size of ablated lines on Figs. 1.16B and 1.18A) of the focal spot (200 µm) ablating the surface to create the conditions of a metal-like plasma. In that case, the optimum number of overlapping shots on the same area of the Si wafer was determined by changing the speed of the translation stage, thus allowing the maintenance of a clean pattern of LIPSS, without debris and irregularities in both the central part and peripheries of the ablated area. The observed pattern of LIPSS formation can be explained by considering the transient metallic nature of the semiconductor surface during irradiation by intense femtosecond pulses [10]. Such irradiation eventually supports surface plasmon polariton excitation [72,73]. LIPSS formation can be considered to be the result of the interference of the laser radiation with the surface plasmon polariton generated at the silicon surface [9]. This consideration is consistent with the used experimental conditions where the surface of the semiconductor is irradiated during an extremely short period, which allows the formation of transient metal-like conditions. Formation of surface plasmon polariton waves is a well-known phenomenon in the context of metals, which have free electrons for excitation of this longitudinal charge density oscillation [9]. However, in the case of semiconductors, the origin of free electrons is the transient excitation of free electrons due to tunnel or multiphoton ionization. Though the threshold of surface breakdown increases with decreasing ablating pulse duration, the typical intensity of those experiments (8 × 1013 W cm−2) was well above the ablation threshold of silicon wafers. At these conditions, tunnel ionization is the dominant process, which causes the formation of a considerable amount of free electrons above the target surface. As the electron density generated on the surface after femtosecond pulse irradiation depends on various laser parameters, including the pulse duration, it is not surprising that significant changes in LIPSS formation are observed using some of the shortest high-power laser pulses currently available.

1.5  CONCLUDING COMMENTS We discussed different methods of periodic nanoripples formation. Particularly, we analyzed nanoripple formation on different bandgap semiconductor surfaces using femtosecond pulses, nanosecond LIPSS on WBG semiconductors, nanostructuring of semiconductor surfaces under the action of femtosecond pulses, fabrication of two-dimensional periodic nanostructures by two-beam interference of femtosecond pulses, and extended homogeneous nanoripple formation during interaction of highintensity few-cycle pulses with a moving silicon wafer. LIPSS formation from ultrashort laser pulse irradiation of semiconductors of different bandgaps has been studied using a Ti:sapphire laser with 8 mJ energy, 45 fs pulse duration, and 800 nm wavelength (1.5 eV) at a fluence in the range of ∼100 mJ cm−2–1 J cm−2. The effects of the number of laser shots, angle of incidence, laser polarization, fluence, incident laser wavelength, bandgap, and ambient medium

33

34

CHAPTER 1  Periodic nanoripples formation on the semiconductors

on the ripple period, have been studied. Depending upon the experimental parameters, nanoripple sizes varied in the range of λ/9–λ. The studies clearly show that narrower nanoripples are formed from WBG semiconductors. In addition, the width of the nanoripples decreases with the laser wavelength and fluence. The observed results were explained considering the transient metallic nature of the semiconductor surface on irradiation with intense femtosecond pulse, which excites surface plasmon leading to the LIPSS formation. The nanoripples get frozen due to fast cooling once the laser pulse is over. The critical role of the surface plasma electron density in deciding ripple period is identified, which helps in generation of narrow subwavelength nanoripples. LIPSS perpendicular to the laser polarization direction and with period of the order of the irradiation wavelength were obtained on the surface of WBG semiconductor wafers of InP, GaAs, GaP, and SiC upon repetitive irradiation at 266 nm with pulses of 6 ns. The ripples were fabricated at fluences below the ablation threshold for a single pulse irradiation and are produced mainly by optical interference effects due to the superposition of the incident radiation with a surface electromagnetic wave created on the wafer surface. Calculation of the temperature increase induced by laser irradiation indicated that a minimum fluence is necessary to assure that the surface temperature is high enough for allowing melting and material rearrangement. The increased surface heterogeneity caused by microdefects created by laser irradiation facilitates the feedback mechanism generating the ripple formation. The higher fluence and number of pulses needed for LIPSS formation as the bandgap increases is related to the higher melting temperatures corresponding to larger bandgaps. It was observed that the amplitude of the ripples increases with optical and thermal penetration depths. The nanostructures inspected by AFM are produced upon multiple pulse irradiation at fluences near the ablation threshold. It was observed that the accumulative effect of both fluence and number of pulses needed for LIPSS formation increased with the material bandgap energy. Those results, together with estimations of the growth of surface temperature, were discussed with reference to the semiconductor electrical, optical, and thermal properties. Two-dimensional periodic nanostructures on ZnO crystal surface were fabricated by two-beam interference of 790 nm femtosecond laser. The long period was, as usually reported, determined by the interference pattern of two laser beams. There were other short periodic nanostructures with periods of 220–270 nm embedding in the long periodic structures. The periods, orientation, and the evolution of the short periodic nanostructures were studied, and found them analogous to the self-organized nanostructures induced by single femtosecond laser beam. Finally, the study of extended nanoripple structures formed during the interaction of high-intensity 3.5 fs pulses with a moving silicon wafer was discussed. The optimization of laser intensity and sample moving velocity allowed the formation of long strips (∼5 mm) of homogeneous nanoripples along the whole area of laser ablation. The comparison of nanoripples produced on the silicon surfaces by few- and multicycle pulses was discussed. It was found that few-cycle pulses produce sharp and homogenous structures compared with multicycle pulses. Those images appeared at the speeds of 0.7–2 mm s−1, with the best observed results at 1 mm s−1 at the used

References

experimental conditions. The comparative studies using two fixed pulse durations (3.5 and 40 fs) were analyzed. Those studies opened the door to homogeneous ripple formation over extended areas of semiconductor surfaces. The comparison of the nanoripples produced on the moving target by few- and multicycle pulses has demonstrated analogous patterns in the case of longer pulses, though the LIPSS homogeneity and sharpness of ripples’ edges were considerably worse compared with few-cycle pulse induced LIPSS formation.

REFERENCES [1] Y. Vorobyev, C. Guo, Colorizing metals with femtosecond laser pulses, Appl. Phys. Lett. 92 (4) (2008) 041914. [2] Q.Z. Zhao, F. Ciobanu, S. Malzer, L.J. Wang, Enhancement of optical absorption and photocurrent of SiC by laser surface nanostructuring, Appl. Phys. Lett. 91 (12) (2007) 121107. [3] K. Minoshima, A.M. Kowalevicz, I. Hartl, E.P. Ippen, J.G. Fujimoto, Photonic device fabrication in glass by use of nonlinear materials processing with a femtosecond laser oscillator, Opt. Lett. 26 (19) (2001) 1516–1518. [4] E.N. Glezer, M. Milosavljevic, L. Huang, R.J. Finlay, T.H. Her, J.P. Callan, E. Mazur, Three-dimensional optical storage inside transparent materials, Opt. Lett. 21 (24) (1996) 2023–2025. [5] T.N. Kim, K. Campbell, A. Groisman, D. Kleinfeld, C.B. Schaffer, Femtosecond laserdrilled capillary integrated into a microfluidic device, Appl. Phys. Lett. 86 (20) (2005) 201106. [6] P.M. Fauchet, A.E. Siegman, Surface ripples on silicon and gallium arsenide under picosecond laser illumination, Appl. Phys. Lett. 40 (9) (1982) 824–826. [7] A.M. Ozkan, A.P. Malshe, T.A. Railkar, W.D. Brown, M.D. Shirk, P.A. Molian, Femtosecond laser-induced periodic structure writing on diamond crystals and microclusters, Appl. Phys. Lett. 75 (23) (1999) 3716–3718. [8] N. Yasumaru, K. Miyazaki, J. Kiuchi, Femtosecond-laser-induced nanostructure formed on hard thin films of TiN and DLC, Appl. Phys. A 76 (6) (2002) 983–985. [9] H. Raether, Surface Plasmons on Smooth and Rough Surfaces and on Gratings, SpringerVerlag, Berlin, (1988). [10] Y. Shimotsuma, P.G. Kazansky, J. Qiu, K. Hirao, Self-organized nanogratings in glass irradiated by ultrashort light pulses, Phys. Rev. Lett. 91 (24) (2003) 247405. [11] J. Reif, F. Costache, M. Henyk, S. Pandelov, Ripples revisited: non-classical morphology at the bottom of femtosecond laser ablation craters in transparent dielectrics, Appl. Surf. Sci. 197–198 (2002) 891–895. [12] Y. Dong, P. Molian, Coulomb explosion-induced formation of highly oriented nanoparticles on thin films of 3C–SiC by the femtosecond pulsed laser, Appl. Phys. Lett. 84 (1) (2004) 10–12. [13] X.J. Wu, T.Q. Jia, F.L. Zhao, M. Huang, N.S. Xu, H. Kuroda, Z.Z. Xu, Formation mechanisms of uniform arrays of periodic nanoparticles and nanoripples on 6H-SiC crystal surface induced by femtosecond laser ablation, Appl. Phys. A 86 (4) (2007) 491–495. [14] V.R. Bhardwaj, E. Simova, P.P. Rajeev, C. Hnatovsky, R.S. Taylor, D.M. Rayner, P.B. Corkum, Optically produced arrays of planar nanostructures inside fused silica, Phys. Rev. Lett. 96 (5) (2006) 057404.

35

36

CHAPTER 1  Periodic nanoripples formation on the semiconductors

[15] U. Chakravarty, R.A. Ganeev, P.A. Naik, J.A. Chakera, M. Babu, P.D. Gupta, Nanoripple formation on different band-gap semiconductor surfaces using femtosecond pulses, J. Appl. Phys. 109 (8) (2011) 084347. [16] M. Huang, F. Zhao, Y. Cheng, N. Xu, Z. Xu, Large area uniform nanostructures fabricated by direct femtosecond laser ablation, Opt. Exp. 16 (23) (2008) 19354–19365. [17] M. Huang, F. Zhao, Y. Cheng, N. Xu, Z. Xu, Mechanisms of ultrafast laser-induced deepsubwavelength gratings on graphite and diamond, Phys. Rev. B 79 (12) (2009) 125436. [18] M. Huang, F. Zhao, Y. Cheng, N. Xu, Z. Xu, Origin of laser-induced near-subwavelength ripples: interference between surface plasmons and incident laser, ACS Nano 3 (12) (2009) 4062–4070. [19] A. Borowiec, H.K. Haugen, Subwavelength ripple formation on the surfaces of compound semiconductors irradiated with femtosecond laser pulses, Appl. Phys. Lett. 82 (25) (2003) 4462–4464. [20] R.A. Ganeev, M. Baba, T. Ozaki, H. Kuroda, Long- and short-period nanostructure formation on semiconductor surfaces at different ambient conditions, J. Opt. Soc. Am. B 27 (5) (2010) 1077–1082. [21] M. Huang, F. Zhao, Y. Cheng, N. Xu, Z. Xu, The morphological and optical characteristics of femtosecond laser-induced large-area micro/nanostructures on GaAs, Si, and brass, Opt. Exp. 18 (S4) (2010) A600–A619. [22] J. Wang, C. Guo, Formation of extraordinarily uniform periodic structures on metals induced by femtosecond laser pulses, J. Appl. Phys. 100 (2) (2006) 023511. [23] Q.Z. Zhao, S. Malzer, L.J. Wang, Formation of subwavelength periodic structures on tungsten induced by ultrashort laser pulses, Opt. Lett. 32 (13) (2007) 1932–1934. [24] M. Birnbaum, Semiconductor surface damage produced by ruby lasers, J. Appl. Phys. 36 (11) (1965) 3688–3694. [25] Z. Guosheng, P.M. Fauchet, A.E. Siegman, Growth of spontaneous periodic surface structures on solids during laser illumination, Phys. Rev. B 26 (10) (1982) 5366–5372. [26] J.F. Young, J.S. Preston, H.M. van Driel, J.E. Sipe, Laser-induced periodic surface structure. II. Experiments on Ge, Si, Al, and brass, Phys. Rev. B 27 (2) (1983) 1155–1162. [27] S.E. Clark, D.C. Emmony, Ultraviolet-laser-induced periodic surface structures, Phys. Rev. B 40 (4) (1989) 2031–2039. [28] B. Kumar, R.K. Soni, Submicrometre periodic surface structures in InP induced by nanosecond UV laser pulses, J. Phys. D Appl. Phys. 41 (15) (2008) 155303. [29] D. Bäuerle, Laser Processing and Chemistry, fourth ed., Springer-Verlag, Berlin ­Heidelberg, 2011. [30] J. Bonse, J.M. Wrobel, K.W. Brzezinka, N. Esser, W. Kautek, Femtosecond laser irradiation of indium phosphide in air: Raman spectroscopic and atomic force microscopic investigations, Appl. Surf. Sci. 202 (3–4) (2002) 272–282. [31] F. Costache, M. Henyk, J. Reif, Surface patterning on insulators upon femtosecond laser ablation, Appl. Surf. Sci. 208–209 (2003) 486–491. [32] O. Varlamova, F. Costache, J. Reif, M. Bestehorn, Self-organized pattern formation upon femtosecond laser ablation by circularly polarized light, Appl. Surf. Sci. 252 (13) (2006) 4702–4706. [33] T. Tomita, Y. Fukumori, K. Kinoshita, S. Matsuo, S. Hashimoto, Observation of laserinduced surface waves on flat silicon surface, Appl. Phys. Lett. 92 (1) (2008) 013104. [34] E.M. Hsu, T.H.R. Crawford, H.F. Tiedje, H.K. Haugen, Periodic surface structures on gallium phosphide after irradiation with 150fs–7ns laser pulses at 800 nm, Appl. Phys. Lett. 91 (11) (2007) 111102.

References

[35] N. Yasumaru, K. Miyazaki, J. Kiuchi, Fluence dependence of femtosecond-laser-induced nanostructure formed on TiN and CrN, Appl. Phys. A 81 (5) (2005) 933–937. [36] M. Shinoda, R.R. Gattass, E. Mazur, Femtosecond laser-induced formation of nanometer-width grooves on synthetic single-crystal diamond surfaces, J. Appl. Phys. 105 (5) (2009) 053102. [37] J. Reif, O. Varlamova, M. Ratzke, M. Schade, H.S. Leipner, T. Arguinov, Multipulse feedback in self-organized ripples formation upon femtosecond laser ablation from silicon, Appl. Phys. A 101 (2) (2010) 361–365. [38] D. Bimberg, Semiconductor Nanostructures, Springer-Verlag, Berlin Heidelberg, (2008). [39] M. Sanz, E. Rebollar, R.A. Ganeev, M. Castillejo, Nanosecond laser-induced periodic surface structures on wide band-gap semiconductors, Appl. Surf. Sci. 278 (2) (2013) 325–329. [40] E.D. Palik, Handbook of Optical Constants of Solids, Academic Press, New York, (1997). [41] S.O. Kasap, P. Capper, Handbook of Electronic and Photonic Materials, Springer-Verlag, Berlin Heidelberg, (2006). [42] H. Qi, Q. Wang, X. Zhang, Z. Liu, S. Zhang, J. Chang, W. Xia, G. Jin, Theoretical and experimental study of laser induced damage on GaAs by nanosecond pulsed irradiation, Opt. Lasers Eng. 49 (2) (2011) 285–291. [43] T. Kondo, S. Matsuo, S. Juodkazis, V. Mizeikis, H. Misawa, Multiphoton fabrication of periodic structures by multibeam interference of femtosecond pulses, Appl. Phys. Lett. 82 (17) (2003) 2758–2760. [44] G. Liang, W. Mao, Y. Pu, H. Zou, H. Wang, Z. Zeng, Fabrication of two-dimensional coupled photonic crystal resonator arrays by holographic lithography, Appl. Phys. Lett. 89 (4) (2006) 041902. [45] A. Divliansky, A. Shishido, I. Khoo, T. Mayer, D. Pensa, S. Nishimura, C. Keating, T. Mallouk, Fabrication of two-dimensional photonic crystals using interference lithography and electrodeposition of CdSe, Appl. Phys. Lett. 79 (11) (2001) 3392–3394. [46] T. Kondo, S. Juodkazis, V. Mizeikis, H. Misawa, S. Matsuo, Holographic lithography of periodic two- and three-dimensional microstructures in photoresist SU-8, Opt. Exp. 14 (19) (2006) 7943–7953. [47] N. Lai, W. Liang, J. Lin, C. Hsu, C. Lin, Fabrication of two- and three-dimensional periodic structures by multi-exposure of two-beam interference technique, Opt. Exp. 13 (23) (2005) 9605–9611. [48] M. Campbell, D. Sharp, M. Harrison, R. Denning, A. Turberfield, Fabrication of photonic crystals for the visible spectrum by holographic lithography, Nature 404 (2000) 53–56. [49] J. Sipe, J. Young, J. Preston, H. Driel, Laser-induced periodic surface structure. Theory I, Phys. Rev. B 27 (5) (1983) 1141–1154. [50] T. Jia, H. Chen, M. Huang, F. Zhao, J. Qiu, R. Li, Z. Xu, X. He, J. Zhang, H. Kuroda, Phys. Rev. B 72 (12) (2005) 125429. [51] M. Huang, S. Mao, H. Feick, H. Yan, Y. Wu, H. Kind, E. Weber, R. Russo, P. Yang, Room temperature ultraviolet nanowire nanolasers, Science 292 (2001) 1897–1899. [52] P. Gao, Y. Ding, W. Mai, W. Hughes, C. Lao, Z. Wang, Conversion of zinc oxide nanobelts into superlattice-structured nanohelices, Science 309 (2005) 1700–1704. [53] T. Jia, M. Baba, M. Suzuki, R.A. Ganeev, H. Kuroda, J. Qiu, X. Wang, R. Li, Z. Xu, Fabrication of two-dimensional periodic nanostructures by two-beam interference of femtosecond pulses, Opt. Exp. 16 (3) (2008) 1874–1878. [54] T.H.R. Crawford, H.K. Haugen, Sub-wavelength surface structures on silicon irradiated by femtosecond laser pulses at 1300 and 2100 nm wavelengths, Appl. Surf. Sci. 253 (11) (2007) 4970–4977.

37

38

CHAPTER 1  Periodic nanoripples formation on the semiconductors

[55] T.-H. Her, R.J. Finlay, C. Wu, S. Deliwala, E. Mazur, Microstructuring of silicon with femtosecond laser pulses, Appl. Phys. Lett. 73 (12) (1998) 1673–1675. [56] Q. Wu, Y. Ma, R. Fang, Y. Liao, Q. Yu, X. Chen, K. Wang, Femtosecond laser-induced periodic surface structure on diamond film, Appl. Phys. Lett. 82 (11) (2003) 1703–1705. [57] S.I. Dolgaev, S.V. Lavrishev, A.A. Lyalin, A.V. Simakin, V.V. Voronov, G.A. Shafeev, Formation of conical microstructures upon laser evaporation of solids, Appl. Phys. A 73 (2) (2001) 177–181. [58] S. Lee, D. Yang, S. Nikumb, Femtosecond laser micromilling of Si wafers, Appl. Surf. Sci. 254 (10) (2008) 2996–3005. [59] A.J. Pedraza, J.D. Fowlkes, Y.F. Guan, Surface nanostructuring of silicon, Appl. Phys. A 77 (2) (2003) 277–284. [60] A. Sakellari, E. Kabouraki, D. Gray, V. Purlys, C. Fotakis, A. Pikulin, N. Bityurin, M. Vamvakaki, M. Farsari, Diffusion-assisted high-resolution direct femtosecond laser writing, ACS Nano 6 (3) (2012) 2302–2311. [61] N. Vasilantonakis, K. Terzaki, I. Sakellari, V. Purlys, D. Gray, C.M. Soukoulis, M. Vamvakaki, M. Kafesaki, M. Farsari, Three-dimensional metallic photonic crystals with optical bandgaps, Adv. Mater. 24 (8) (2012) 1101–1105. [62] R. Wagner, J. Gottmann, A. Horn, E.W. Kreutz, Subwavelength ripple formation induced by tightly focused femtosecond laser radiation, Appl. Surf. Sci. 252 (24) (2006) 8576–8579. [63] B. Tan, K. Venkatakrishnan, A femtosecond laser-induced periodical surface structure on crystalline silicon, J. Micromech. Microeng. 16 (5) (2006) 1080–1087. [64] T.H.R. Crawford, G.A. Botton, H.K. Haugen, Crystalline orientation effects on conical structure formation in femtosecond laser irradiation of silicon and germanium, Appl. Surf. Sci. 256 (6) (2010) 1749–1755. [65] S. Bashir, M.S. Rafique, W. Husinsky, Femtosecond laser-induced subwavelength ripples on Al, Si, CaF2 and CR-39, Nucl. Instrum. Meth. B 275 (2012) 1–6. [66] R.A. Ganeev, D.Y. Lei, C. Hutchison, T. Witting, F. Frank, W.A. Okell, T.R. Roschuk, S.A. Maier, J.W.G. Tisch, J.P. Marangos, Extended homogeneous nanoripple formation during interaction of high-intensity few-cycle pulses with a moving silicon wafer, Appl. Phys. A 112 (2) (2013) 457–462. [67] J.S. Robinson, C.A. Haworth, H. Teng, R.A. Smith, J.P. Marangos, The generation of intense, transform-limited laser pulses with tunable duration from 6 to 30 fs in a differentially pumped hollow fibre, Appl. Phys. B, 85 (4) (2006) 525–529. [68] T. Witting, F. Frank, C.A. Arrell, W.A. Okell, J.P. Marangos, J.W.G. Tisch, Characterization of high-intensity sub-4-fs laser pulses using spatially encoded spectral shearing interferometry, Opt. Lett. 36 (9) (2011) 1680–1682. [69] L. Gallmann, G. Steinmeyer, D.H. Sutter, T. Rupp, C. Iaconis, I.A. Walmsley, U. Keller, Spatially resolved amplitude and phase characterization of femtosecond optical pulses, Opt. Lett. 26 (2) (2001) 96–98. [70] T.H.R. Crawford, A. Borowiec, H.K. Haugen, Femtosecond laser micromachining of grooves in silicon with 800 nm pulses, Appl. Phys. A 80 (8) (2005) 1717–1724. [71] S.K. Das, K. Dasari, A. Rosenfeld, R. Grunwald, Extended-area nanostructuring of TiO2 with femtosecond laser pulses at 400 nm using a line focus, Nanotechnology 21 (15) (2010) 155302. [72] S.A. Maier, Plasmonics: Fundamentals and Applications, Springer, New York, (2007). [73] Y. Han, S. Qu, The ripples and nanoparticles on silicon irradiated by femtosecond laser, Chem. Phys. Lett. 495 (6-10) (2010) 241–244.