Laser Technology: Applications in Adhesion and Related Areas

By K.L. Mittal

The acronym Laser is derived from Light Amplification by Stimulated Emission of Radiation. With the advent of the ruby laser in 1960, there has been tremendous research activity in developing novel, more versatile and more efficient laser sources or devices, as lasers applications are ubiquitous. Today, lasers are used in many areas of human endeavor and are routinely employed in a host of diverse fields: various branches of engineering, microelectronics, biomedical, medicine, dentistry, surgery, surface modification, to name just a few.
In this book (containing 10 chapters) we have focused on application of lasers in adhesion and related areas. The topics covered include:

• Topographical modification of polymers and metals by laser ablation to create superhydrophobic surfaces.
• Non-ablative laser surface modification.
• Laser surface modification to enhance adhesion.
• Laser surface engineering of materials to modulate their wetting behavior
• Laser surface modification in dentistry.
• Laser polymer welding.
• Laser based adhesion testing technique to measure thin film-substrate interface toughness.
• Laser surface removal of hard thin ceramic coatings.
• Laser removal of particles from surfaces.
• Laser induced thin film debonding for micro-device fabrication applications.

• Language: English
• Publisher: Wiley
• Release date: Jan 22, 2018
• ISBN: 9781119185048

Book preview

Preface

The acronym Laser is derived from Light Amplification by Stimulated Emission of Radiation. The first theoretical description of stimulated emission of radiation was given by Einstein in 1917. However, it took four decades for the technical realization of a laser source when in 1960 T.H. Maiman developed a solid state ruby laser emitting red laser radiation. Since the advent of the ruby laser, there has been an exponential progress in designing many different lasers with unique and specific characteristics, as lasers have found myriad applications in a host of industries for a legion of purposes. In fact, lasers are ubiquitously used and here an eclectric catalog of examples where lasers have been used efficiently and effectively should suffice to underscore the widespread utility of lasers: mechanical engineering operations (e.g., micromachining), adhesion promotion, plastics welding, surface modification, dentistry, surgery, microelectronics, patterning, MEMS (microelectromechanical systems) and NEMS (nanoelectromechanical systems).

Many laser parameters such as wavelength emitted, pulse duration, power, pulse repetition rate dictate the function and performance of a laser source. A panoply of laser sources is available for different tasks, and there is tremendous activity in ameliorating the existing laser sources as well as in devising more versatile and more efficient laser systems.

Considering the voluminous literature available dealing with laser science and technology, one will need a multi-volume compendium to cover all facets of lasers. However, in this book we have focused on the applications of laser technology in adhesion and allied areas. Lasers play a significant role in the domain of adhesion. For example, polymers are used for a chorus of industrial applications as polymers have a number of desirable bulk traits, but these materials suffer from lack of adhesion due to their low reactivity and low surface free energy. A number of different techniques (e.g., corona, plasma, flame, UV/ozone, wet chemical) are commonly harnessed to modify polymer surfaces and render them adhesionable. But laser surface treatment offers some distinctive features and advantages. These days one of the mantras is: green and laser technology offers a green (environmentally-benign) alternative without noxious emissions.

In the adhesion-related arena, two examples where lasers have played a very useful role can be cited as follows. One is directly from Nature’s treasure-trove and it is said that Nature does not waste time in trifling things and Nature is a great teacher. Here we are referring to the behavior and significant trait (self-cleaning) of the Lotus Leaf. Since the recognition/popularization of the Lotus Leaf Effect in 1997, there has been an explosive growth of interest in replicating the surface chemistry and topography of the Lotus Leaf using an array of techniques and in this venture laser technology has found much application. Another example is the removal of particles from surfaces. In the field of microelectronics, with the ever-shrinking feature size there is patent need to remove a few nanometer size particles and lasers have been found to be capable of removing such small particles. Apropos, the antonymous field of debonding has also benefited from the lasers, as lasers have been utilized to debond materials (e.g., thin films and coatings) from variegated substrates.

Now coming to this book which contains 10 chapters written by internationally renowned subject matter experts is divided into two parts: Part 1: Laser Surface Modification and Adhesion Enhancement, and Part 2: Other Applications. The topics covered include: Topographical modification of polymers and metals by laser ablation to create superhydrophobic surfaces; nonablative laser surface modification; laser surface engineering of materials to modulate their wetting behavior; laser surface modification to enhance adhesion; laser surface modification in dentistry; laser polymer welding; laser based adhesion testing technique to measure thin film-substrate interface toughness; laser induced thin film debonding for micro-device fabrication applications; laser surface removal of hard thin ceramic coatings; and laser removal of particles from surfaces.

This unique book should be of great interest, value and usefulness to those in materials science, chemistry, physics and engineering. The book is profusely illustrated and copiously referenced. The information consolidated in this book should be of much value and relevance to R&D personnel engaged in adhesion and adhesive bonding, surface modification (both physical and chemical) for a host of applications, polymer welding, cleaning (removal of hard thin coatings and nanometer size particles from surfaces), dentistry, device fabrication, micro and nanostructures formation, and unravelling thin film/substrate adhesion behavior.

Now comes the important and fun part of writing a Preface as it provides the opportunity to thank those who were instrumental in materializing this book. First and foremost, we are beholden to the authors for their interest, enthusiasm, unwavering cooperation and contributions which were a desideratum to bring out this book. Our appreciation is extended to Martin Scrivener (publisher) for his sustained commitment and steadfast support for this book project, and for giving this book a body form.

K. L. Mittal

Hopewell Jct., NY, USA

E-mail: [e-mail]

Wei-Sheng Lei

Applied Materials Inc.,

Sunnyvale, CA, USA

E-mail: [e-mail]

Part 1

LASER SURFACE MODIFICATION AND ADHESION ENHANCEMENT

Chapter 1

Topographical Modification of Polymers and Metals by Laser Ablation to Create Superhydrophobic Surfaces

Frank L. Palmieri* and Christopher J. Wohl

NASA Langley Research Center, Hampton, VA, USA

*Corresponding author: [e-mail]; [e-mail]

Abstract

The applications for superhydrophobic surfaces are nearly limitless: self-cleaning coatings, corrosion resistance, ice mitigation, non-stick cookware, and anti-fog surfaces to name a few. The last few decades of research have shown repeatedly that synergy of surface chemistry and topography must be harnessed to attain superhydrophobicity. Over the same time frame, laser technology has advanced such that fast and ultrafast lasers with sufficient power for surface ablation are now available to both researchers and high volume manufactures to modify the topography and chemistry of materials. In this chapter, laser processing methods to prepare hydrophobic and superhydrophobic surfaces are reviewed. Brief backgrounds in wetting theory and laser ablation are provided to prepare the reader. The preparation of superhydrophobic surfaces by laser ablation is divided into four sections based on substrate materials and hydrophobic coatings: 1) hydrophobic organic substrates, 2) hydrophilic organic substrates, 3) hydrophilic substrates with hydrophobic coatings, and 4) hydrophilic inorganic substrates.

Keywords: Surface free energy, contact angle, reentrant, hierarchical structures, wettability, laser ablation, superhydrophobicity

1.1 Introduction

Superhydrophobic surfaces have been the subject of thousands of research articles and patents since 1977 when scientists began studying and re-creating the properties of the lotus leaf [1–3]. Research on superhydrophobicity has exploded in the last few decades as scientists search for self-cleaning, anti-fouling, corrosion-resistant coatings to protect everything from industrial infrastructure to sunglasses [4, 5]. A keyword search for superhydrophobic on The Web of Science™ shows a dramatic increase in the number of new publications starting around 2002 (Figure 1.1), possibly spurred by a 1996 paper on superhydrophobic fractal surfaces and a 1997 review of hydrophobic plants [6, 7].

Figure 1.1 The number of publications using the word superhydrophobic each year since 1998.

Superhydrophobicity requires an advancing water contact angle (ACA) > 150° and a sliding angle (SA) < 10° (the angle with respect to gravity required for a drop to move on a surface). A low contact angle hysteresis (CAH, the difference between the ACA and receding contact angle (RCA)) is implicit with a low SA and is often used to classify surfaces as superhydrophobic. Smooth materials with low surface free energy have been prepared with contact angles (CAs) up to approximately 120° and low CAH, but cannot meet the superhydrophobicity requirements without additional topography [8–12]. For example, polished silicon wafers covered with densely packed trifluoromethyl groups achieved a water contact angle (WCA) of 119° and had a CAH of 6° [9]. Smooth surfaces covered with a liquid-like monolayer of covalently bound poly(dimethylsiloxane) resulted in an impressive CAH of 1° in some cases, but the ACA was less than 107° in all cases [8]. Porous solids filled with fluorinated fluid, given the name slippery liquid infused porous surfaces (SLIPS), resulted in effectively smooth surfaces with moderate ACA and CAH characteristics [12]. The scale and morphology of roughness for a SLIPS surface do not impact either the WCA or CAH, but have a profound effect on the retention of the infused liquid under high shear conditions [10]. Although these smooth, low energy surfaces demonstrated hydrophobicity and even low CAH in some cases, none of them achieved a WCA greater than 150° because the surfaces lacked topography.

The dependence of wettability on both the scale and form of surface topography has been thoroughly studied by observation of natural and synthetic surfaces [13]. Both random and regular topographies on the nanometer to micrometer scale impact the macroscopic wetting properties of solids. Naturally occurring superhydrophobic surfaces may contain structures with a combination of length scales (i.e. hierarchical structures). Synthetic surfaces, mimicking topographies found in nature, can be prepared using a litany of fabrication techniques which may be random (e.g. phase separation and abrasion) or precise (e.g. lithography, self-assembly and micromachining). Precise fabrication methods are widely used to prepare regular arrays on the nano- and micro- scales. A combination of regular, microscale structures and random, nanoscale structures is often used to prepare hierarchical topographies. One versatile technique for the fabrication of hierarchical nano/microstructures is micromachining by laser ablation.

Although conceived much earlier in science fiction, lasers were first developed in the 1960’s and were accompanied almost immediately by research on the ablation of materials [14]. As laser technology developed, systems capable of high precision micromachining were developed for applications from chemical analytics to microfabrication to medicine. Modern systems provide a fast, efficient, low environmental impact means of generating microstructured surfaces.

Although it is not the subject of this chapter, a great deal of research has been devoted to laser ablation as a means of cleaning surfaces, creating polar and reactive species, and creating topography to improve bonding with coatings and adhesives [15, 16]. Laser ablation can increase or decrease the surface free energy depending on the residual chemical species and resulting topography. For example, Lawrence and coworkers published a series of papers on laser ablation of stainless steel and aluminum to modify wettability. They showed that laser ablation with a defocused beam at high powers resulted in smoother surfaces with lower WCAs [17–19].

In this chapter, the basic theory for macroscopic wetting behavior is presented with an emphasis on the understanding and modeling of hydrophobic and superhydrophobic phenomena. The physics of laser ablation is described with specific surface modification examples of both inorganic and organic substrates. Finally, a literature review is presented for (super) hydrophobic surfaces prepared by laser ablation of inorganic substrates with and without hydrophobic coatings. Special attention is given to surfaces that exhibit (super)hydrophobicity without chemical modification after ablation (i.e. laser ablation results directly in (super)hydrophobicity).

1.2 Wetting Theory

This section provides the reader with a theoretical background of wetting phenomena. The development of macroscopic wetting theory with some attention to microscopic and stochastic models for prediction of wetting behavior will be emphasized. Many topics that are referenced in later sections of this chapter (i.e. wetting states, pinning theory, etc.) are described here in greater detail.

The basic theory that describes the interaction at a solid/liquid/vapor interface is given by Young’s equation from 1805,

(1.1)

which relates the CA, θ, to the interfacial free energies of the solid-liquid (γSL), solid-vapor (γSV), and liquid-vapor (γLV) interfaces [20]. The CA is a macroscopic, thermodynamic quantity because it is independent of intermolecular forces which are acting over much shorter distances than dimensions of the wetted interface. No information about microscopic shape of the contact profile can be derived from the macroscopic CA [21]. The apparent CA is observable by a wide variety of techniques[22] and is the basis of several models to calculate surface free energies[23–25]. Solids with a high γSV generally exhibit low CAs (< 90°), whereas low γSV surfaces exhibit high CAs (> 90°). Control over the solid surface free energy and, in turn, over the wetting properties is the goal of thousands of materials researchers.

Wenzel observed and published the first significant advance in understanding the impact of topography on CA [26]. His 1936 paper describes how surface roughness enhances the hydrophobicity of hydrophobic surfaces and the hydrophilicity of hydrophilic surfaces. His modification to Young’s equation is,

(1.2)

where θΑpp is the apparent CA observed on a rough surface, r is a roughness factor defined as the ratio of real surface area to flat surface area and θ⁰ is the intrinsic CA on an ideal surface, which replaces Young’s CA. An ideal surface is smooth, homogeneous, rigid, insoluble and non-reactive with the contacting liquid. Because r is always greater than one, adding roughness only increases the numerator of equation (Wenzel) which drives the apparent CA away from 90° [26].

The Wenzel model successfully predicts the apparent CA of rough, homogeneous surfaces. In 1944, Cassie and Baxter [27] proposed a model to predict the apparent CA on rough, heterogeneous surfaces composed of two different materials.

(1.3)

In equation (1.3), f1 and f2 are the complementary fractions of the real surface areas with intrinsic CAs given by θ1⁰ and θ2⁰, respectively [27]. For porous surfaces where f2 is the area fraction of air entrapped under a droplet, θ2⁰ is 180°, and equation (1.3) simplifies to equation (1.4).

(1.4)

This simplified form of the Cassie-Baxter equation is commonly used for describing superhydrophobic surfaces where f1 = f and is referred to as the Cassie-Baxter coefficient. When Wenzel’s roughness factor, r, is applied to the Cassie-Baxter model, a combined, Wenzel/Cassie-Baxter model can be written.

(1.5)

Unlike the Wenzel model, the Cassie-Baxter model predicts the possibility of an apparent CA greater than 90° even with an intrinsic CA < 90° which means intrinsically hydrophilic substrates can be topographically modified to be (super)hydrophobic without further chemical modification if the topography results in trapped air. Equating (1.2) with (1.5), we obtain a relationship between f, r, and the critical intrinsic CA (θc) predicted for the transition between the Wenzel and Cassie-Baxter wetting states.

(1.6)

Because f < 1 < r, the transition from the Wenzel state to a Cassie-Baxter state requires that θc > 90°, but (super)hydrophobicity has been demonstrated on substrates with θ⁰ < 90° which requires a Cassie-Baxter wetting state [28–30]. In fact, the waxy coating found on the lotus leaf and the surfaces of several other superhydrophobic plants have a θ⁰ ~ 75° [28]. The failure of equation (1.6) to predict superhydrophobicity on these surfaces has been attributed to the reentrant surface structures, characterized by sidewall angles > 90°, i.e. the surface normal vector intersects the surface interface more than once as in Figure 1.2.

Figure 1.2 The local angle of a surface topographical structure with respect to the average surface tangent is shown here as θTop. A non-reentrant feature (a) has θTop < 90° and only one intersection with the surface normal vector (arrows protruding from surface). Three possible reentrant structures are shown (b-d) which have θTop > 90° and intersect with the surface normal vector in at least two places.

The entrapment of air between reentrant structures occurs due to contact line pining at an outside corner of the surface structure where additional advancement of the contact line would reduce the microscopic CA below θ⁰. Wang and Chen established a set of criteria for air entrapment based on an energy balance. Air entrapment is predicted when the depth of a pore is greater than the depth of intrusion by a liquid [31]. This is depicted in Figure 1.3, where the liquid front may make contact with a neighboring structure and entrap air before contacting the bottom of the pore depending on the microscopic geometry and θ⁰ [32].

Figure 1.3 A representation of a liquid pinning on a reentrant surface with a θ⁰ of 70° leading to entrapped air. (a) shows a liquid front advancing on a reentrant surface. In (b), the liquid front maintains a CA of 70° as it advances around the reentrant surface feature. Pinning occurs at the point on the circular surface feature where further wetting forces the CA to deviate from 70°. In (c) the liquid front makes contact with the top of a neighboring structure before the liquid is intruded to the bottom of the pore, causing gas entrapment.

For the entrapment of air, the Cassie-Baxter state requires a liquid bridge to form between surface asperities. For stability, the three-phase contact line must remain pinned on the asperities, and the liquid bridge cannot make contact with any solid surface within the pore. As the span of the liquid bridge increases or the intrinsic CA of the surface decreases, the liquid bridge penetrates farther into the pore due to gravitational forces acting on the liquid. If this penetration depth is equal to the topography height, the entire surface will be wetted, and superhydrophobicity will be lost as the pore assumes a Wenzel state. The microscopic pinning of the contact line on surface topography can be used to explain the wetting behavior in the Wenzel state for the so-called rose petal effect where droplets simultaneously exhibit superhydrophobic CAs and very high CAH [33–35].

The design of topographies to entrap air on wetted surfaces has been the focus of many researchers [7, 28–30, 36]. Cao et al. stated that a reentrant surface with a topography angle (θTop) > (180 – θ⁰) would prevent a liquid from penetrating into a pore [30]. Wang and Chen went further to provide a model to predict intrusion depth (hi) into a pore of width dv in terms of an intrusion angle β and the θ⁰ of the substrate [31].

(1.7)

For air entrapment, surface topography must accommodate both θTop > β and depth of pore > hi. This model was applied to 1) square arrays of cylindrical pillars, 2) square arrays of square pillars, and 3) hexagonal arrays of square pillars all with θTop = 90°. The intrusion depth was smallest for geometry 2) and greatest for geometry 3). Wang and Chen’s model accurately predicted air entrapment for experimental results obtained by others on structures with θTop ~ 90° [31, 37, 38]. Finally, the model was applied to a reentrant geometry, allowing for β > 90°, which predicted entrapped air for θ⁰ < 90°.

Tuteja et al. [28] proposed a more generalized model to predict the stability and hydrophobicity of the Cassie-Baxter state on reentrant surfaces based on two dimensionless geometric parameters, H* and D*. The dimensionless height (H*) is the ratio of the maximum pore depth (h2), the vertical distance between the contact line and pore bottom, to the sagging depth (h1), the depth of liquid penetration past the contact line. When the liquid-air interface is farther from the bottom of the pore, the value of H* is greater and the Cassie-Baxter state is more stable. The dimensionless distance between asperities, D*, is the inverse of f. When the distance between asperities is greater, more of the surface is covered with entrapped air, the value of D* is greater and the surface is more hydrophobic

The forms of H* and D* are geometry dependent, and are given for a surface covered in cylindrical fibers and for micro-hoodoo structures (a specific form of reentrant structure). For cylinders:

(1.8)

(1.9)

Here, R is the radius of the fiber or hoodoo head, is the density, g is gravitational acceleration, and D is half the characteristic spacing between the fibers or hoodoo heads. For hoodoo structures:

(1.10)

(1.11)

where W is the width of the micro-hoodoo head and H is the height of the support column. These dimensionless parameters predicted the stability of tall, reentrant structures with large spacing (nanonails and micro-hoodoos) to have the greatest hydrophobicity and stability [28, 29].

Continuum wetting models based on macroscopic quantities, although greatly advanced since the time of Thomas Young, cannot explain all wetting phenomena. Stochastic models may provide insight into the microscopic mechanisms of observed wetting behaviors. Mean field theory (MFT) is a simplified probability model where the average effect of an ensemble on a body is statistically determined. Monson and coworkers used MFT with a lattice gas model to predict the 3-D density distribution for liquid droplets on smooth and textured surfaces [39, 40]. The density distribution provided information about CA and penetration of liquids into pores. It also predicted hybrid wetting states where Cassie-Baxter (CB) and Wenzel states occur simultaneously in distinct regions along the solid-liquid interface. This model might explain the observed petal state of some droplets that exhibit high CAs indicative of a CB wetting state while simultaneously exhibiting high CAH, indicative of a Wenzel wetting state.

MFT can be used to model wetting behavior for nanoscale droplets or topographical features, which is challenging to observe experimentally. Malonoski et al. used lattice density functional theory (DFT) to study the wetting of nanodroplets on nanostructured surfaces and showed that line tension, which was originally proposed by Gibbs but was neglected in Young’s equation, must be considered on the nanoscale [41]. Malonoski proposed a simple expression for θ⁰ based on Young’s equation.

(1.12)

The term is composed of the Gibbs line tension (τ), the strength of the interaction potential (σ), and the radius of the circle of contact between the drop and an ideal surface (RC) which acts to increase the intrinsic CA. As the drop size (and RC) increases to the macroscale, the line tension term becomes vanishingly small and Young’s equation dominates the expression. Other authors used a similar treatment to model the effect of line tension on θ⁰ [42, 43]. Checco and Guenoun used noncontact atomic force microscopy to measure CAs of nanoscopic alkane droplets on silicon wafers coated with octadecyltrichlorosilane, and found that the intrinsic CA increased with increasing drop size[44]. The experimental data did not correlate well with model systems based on line tension effects. Rosso and Virga stated that measuring line tension from these experiments was exceptionally difficult and values ranging from 10−12 to 10−5 N were reported [45, 46].

Wang et al. [47] used a DFT analysis and concluded that the Cassie-Baxter wetting of micro-rough surfaces with microscopic droplets was modeled well using macroscopic equations. In contrast, Wenzel and transitional (i.e. between Cassie-Baxter and Wenzel states) wetting state characteristics predicted by macroscopic models deviated significantly on micro-rough surfaces.

Given the developments over the past two centuries, wetting phenomena remain as a highly active area of fundamental and applied research. It should be noted that, almost exclusively, the study of wetting phenomena has been confined to relatively steady-state systems, i.e., low speed dynamics, in which case inertia effects can be ignored. Recently, research on the wetting and liquid transfer between two surfaces was investigated at different separation velocities where it was determined that the transfer ratio (the ratio of liquid transferred to the second surface relative to the amount remaining on the first) did not converge to 0.5 as a result of an asymmetric liquid bridge between the surfaces at high separation velocities [48]. The design of practical, superhydrophobic surfaces stands to benefit from better understanding of microscopic effects on macroscopic wetting behavior.

1.3 Laser Ablation Background

Laser ablation is the removal of material from a surface using laser radiation. Ablation can be performed on any material that absorbs the incident radiation making it a highly versatile technique for creating topography on polymers, ceramics and metals. Laser radiation can be focused to an ultimate resolution (R) given by the Rayleigh criterion where λ is wavelength, NA is the lens numerical aperture, and k is a system constant of order 1.

(1.13)

A focused beam combined with robotic motion control can be used to machine regular arrays of 3D microstructures which are often covered with irregular nanostructures leading to a hierarchical topography [49]. Radiation can also be used to create random micro- and nano- scaled structures which are inherently formed during many ablation processes due to the variety of physical and chemical mechanisms that occur simultaneously during substrate irradiation.

1.3.1 Ablation Mechanics

Photo-chemical, photo-physical, and photo-thermal processes can occur individually or in combination to cause ablation. Photo-chemical ablation (photoablation) is the disassociation of chemical bonds due to the absorption of photons which typically requires a fluence of 800–1000 mJ/cm² for organic materials. As much as an order of magnitude higher fluence is needed to cause photoablation in metals and ceramics [50]. Photons in the ultraviolet range (100–400 nm) are absorbed by most materials within one micrometer of the surface and have sufficient energy to disassociate covalent bonds. Additionally, lasers with short pulse duration (~10 picoseconds or less) can have sufficient peak power to enable multiphoton absorbtion and photoablation. A laser pulse of sufficient power can disassociate a large number of bonds creating a low-density, high-pressure plasma cell which can explode and eject material. If the time scale for these mechanisms is short (tens of nanoseconds or less) the process remains mostly photoablative rather than photothermal. The photoablative process generally results in less heating of the surface and therefore in a smaller heat affected zone (HAZ), less melting, and less debris.

Photothermal ablation (thermal ablation) occurs when laser radiation is absorbed and causes rapid heating of a material above the boiling point such that vapor is formed. Lasers in the visible to infrared spectral range (400 nm to 1 mm) have a photon energy less than 3 eV and are less likely to cause bond disassociation but can result in heating. Also lasers with longer pulse durations (nanoseconds and longer) can lead mostly to heating. The time scale for thermal ablation is also important to reduce thermal degradation of material near the site of ablation, i.e. the heat affected zone (HAZ). Rapid melting, boiling and solidification which occur as part of thermal ablation can lead to highly textured surfaces with applications in hydrophobic surfaces.

Photo-physical ablation (physical ablation) is the ejection of material caused by a shockwave created near the surface of a material by a laser pulse. The shockwave is generally caused by rapid thermal expansion of the substrate. If the laser pulse duration is significantly shorter than the thermal relaxation time of the material, a thermoelastic stress is induced and propagates out from the point of irradiation as a wave. The stress wave can cause the surface of the material to fracture and eject particles [51]. Physical ablation is significantly more energy efficient than other laser ablation mechanisms because particles with many thousands of atoms may be ejected without the requirement to disassociate or vaporize the atoms within the particles [52].

1.3.2 Ablation in Metals

The versatility of laser ablation patterning has led to extensive use of this technique to modify surface topography and chemistry. Laser parameters such as pulse duration, wavelength, laser fluence and pulse repetition rate, and other factors have been studied on various materials. Shorter pulse durations (τ < 10 ps) lead to plasma formation and photoablation which results in the removal of material without disturbing the surrounding substrate [53, 54]. An atomistic model for ultrafast lasers under the condition of stress confinement predicted the mechanical spallation of a molten layer or droplets at a critical fluence. Further increases in fluence lead to phase explosion characterized by atomized vapor in the ablation plume [55]. For τ > 50 ps, thermal ablation is dominant, so melting and freezing of the substrate can create topography in addition to the removal of material by vaporization.

Femtosecond lasers are better for controlled micromachining because they can create well-defined cavities and profiles with less thermal damage than a nanosecond laser [56]. Hwang et al. found that femtosecond laser pulses created protrusions, rims and spherules on copper, silver, and gold substrates with dimensions ranging from 20 to 250 nm [57]. Ripples, commonly known as laser induced periodic structures (LIPS) with a characteristic wavelength smaller than the laser wavelength were also observed on a variety of substrate types after ablation with femtosecond laser pulses [56, 58–61]. The ripples were attributed to direct surface plasmon-laser interference and grating-assisted surface plasmon-laser coupling [58]. A second model was developed for the formation of ripples based on the interference between reflected and incident light. In certain alloys of steel, the ripples initiated on grain boundaries [61]. With increasing fluence and number of pulses, the ripples are replaced by dense conical structures as seen in the center of the ablation spot in Figure 1.4 c and d, where the fluence is greatest [60].

Figure 1.4 Scanning electron micrographs depicting the formation of sub-wavelength (800 nm) surface ripples using a 470 fs pulse duration. (a) 9 pulses created ripples with little ablation. As the number of pulses was increased, the ripples became more pronounced (b, c) and eventually broke up into conical structures (d). (Reprinted from [56] with permission of Elsevier).

Kurselis et al. investigated the effects of beam polarization, line pitch, fluence, spot size, ambient gas, and substrate starting morphology on laser irradiation induced topography in metals using a 50 fs pulse duration, 800 nm wavelength, 1 kHz repetition rate and a spot size of approximately 40 µm. A distinct transition between sub-wavelength ripples and large amplitude corrugations depended on the spacing of laser lines (1–15 µm) and the fluence (0–5 J/cm²) [60]. In the same paper, sub-spot-size structures were produced from surface defects and seed patterns by creating a laser raster with variable pitch and fluence (Figure 1.5). In Figure 1.5, the upper portions of the structures appear to be covered with deposited ablated material ejected from the holes.

Figure 1.5 Scanning electron micrographs of seed holes patterns (1), deeper holes and pores (2) prepared at 0.87 J/cm², and larger corrugations (3) produced at 4.17 J/cm². (1) and (2) were prepared with a laser raster pitch of 7 µm, spot size of 40 µm and a scan speed of 800 µm/s). (Reprinted from [60] with permission of Elsevier).

Lasers with nanosecond pulse duration are widely used for industrial micromachining to manufacture integrated circuits and microelectromechanical devices. Yung et al. found that pulse energy and the number of passes had the largest effect on the quality of the kerf, the apparent width of an ablated line. A narrower kerf with less debris resulted from lower energy pulses and more passes. Thermal damage to the surface was also minimized in this case [62]. Nanosecond laser ablation often leads to self-organized conical structures with aspect ratios much greater than one (Figure 1.6). Cones with a height greater than 500 µm were produced in indium using 15 ns laser in a vacuum [63]. Dolgaev et al. used a 511 nm laser with a 20 ns pulse duration to prepare cones in silicon wafers with a height of tens of micrometers and sidewall angles of approximately 20°. Hydrodynamic instabilities create capillary waves in the molten silicon surface which leads to periodic surface undulations. These undulations develop into high aspect ratio cones during successive pulses (~10⁴) due to reflection of radiation from the sides of the cones and absorption in the valleys. Concentration of radiation between the cones deepens the valleys while ejected material tends to deposit on the peaks of the cones [64]. This explanation for the growth of conical structures may also be applicable for the formation of structures described by Kurselis et al. [60].

Figure 1.6 Scanning electron micrographs of conical structures formed in silicon (left) and germanium (right) using thousands of 20 ns pulses, λ = 511 nm. (Reprinted from [64] with permission of Springer).

1.3.3 Ablation in Polymers

The laser irradiation of organic polymers also leads to photo, thermal and mechanical decomposition which can alter surface topography and chemistry. The decomposition mechanism varies depending on polymer type and in many cases occurs by depolymerization as the polymer is heated above its ceiling temperature. Condensation polymers composed of imides and carbonates decompose into oligomers and fragments while chain reaction polymers tend to unzip into monomers [65]. The rapid production of volatiles during decomposition drives mechanical ablation processes.

Polymer optical, thermal and mechanical properties vary widely which affects both ablation rate and the resulting topography. For example, pyromellitic dianhydride (PMDA, Tg > 360 °C) based polyimide is much less sensitive to ablation with 308 nm radiation than Duramid (Tg = 285 °C), a photodefinable polyimide [65]. Pham et al. studied the effect of the Tg on the ablation of 14 polymers including step growth and free radical growth polymers with organic backbones as well as poly(dimethylsiloxane) using three different wavelengths. They found that the product of the ablation rate and the reduced glass transition temperature (TgR = Tg - 273 K) for all polymers correlated linearly and universally with laser fluence [66]. The ablation rate of Kapton HN at 248, 308, and 351 nm wavelengths follows a typical Arrhenius behavior, but at 193 nm, the ablation rate is linear with respect to fluence and has a clearly defined ablation threshold. The 193 nm photons cause photochemical ablation whereas thermal ablation dominates for the other three wavelengths [67].

Additives and impurities such as catalysts, fillers, UV and thermal stabilizers, and plasticizers can also impact the ablation process substantially. UV absorbing additives can render a transparent, insensitive polymer sensitive to ablation at specific wavelengths [65]. For example, poly(methyl methacrylate) (PMMA) absorbs weakly in the near UV, but was successfully ablated with minimal thermal damage by doping with (2–2’-hydroxy-3’, 5’-diisopentyl-phenyl) benzotriazole (Tinuvin™) [68]. Inorganic additives may have a significantly higher ablation threshold than the polymer which often leads to regular, conical microstructures [65, 69, 70]. The conical structures formed on a carbon filled acrylic polymer (Figure 1.7, left) and on a triazene polymer with a calcium containing organic additive (Figure 1.7 right) resemble those formed on a silicon wafer (Figure 1.6, left). Silvain et al. proposed that carbon particles agglomerate in the ablation plume and, by migration on a molten surface, form microscopic agglomerates which resist ablation and lead to the formation of cones [69]. A similar mechanism is likely to occur with the calcium containing polymer.

Figure 1.7 Cones are formed in a carbon filled acrylic resin by ablation with a 532 nm, 250 mJ/cm², 10 kHz laser (left) and in a triazene polymer with a calcium containing organic additive ablated with a 308 nm excimer laser (right). (Left image reprinted from [69] with permission from Elsevier. Right image reprinted from [65] with permission of Springer).

Kreutz et al. [71] investigated the development of surface roughness on various polymers using a 248 nm excimer laser with a fluence up to 30 J/cm². The roughness ratio (ablated roughness/initial roughness) increased with fluence approximately following a power law. Highly absorbing polymers with aromatic structures had small optical penetration depth and therefore less roughness than aliphatic polymers which had greater optical penetration depths. At high radiation dose, polymers also exhibited a smoothing effect attributed to cumulative heating which led to softening and flow.

Polytetrafluoroethylene (PTFE) is a particularly useful polymer for making superhydrophobic surfaces because of its low surface free energy and high intrinsic CA (approximately 120°); thus several groups have investigated its ablation properties. An 8 ns Nd:YAG laser with 1064 nm, 532 nm, and 355 nm wavelengths exhibited ablation thresholds in PTFE of 60.0, 40.0, and 17.4 J/cm², respectively, measured in air using a photothermal deflection technique. The variation in ablation threshold with wavelength was attributed mainly to a higher absorption coefficient at lower wavelengths [72]. The ablation characteristics followed a thermal model developed by Garrison [73]. An ultrafast, Ti:sapphire laser (780 nm, 110 fs) used to ablate PTFE produced microcone (3–4 µm diameter) structures using a single, 1 J/cm² pulse while 5 pulses resulted in a nanoporous surface. Multiple pulses improved the edge quality and ablation uniformity, therefore high repetition rates were recommended for laser machining of PTFE [74]. Below a fluence 1.2 J/cm², no shoulders were formed on the sides of ablated trenches; at 5 J/cm², a shoulder of up to 8 micrometers tall was observed. Hashida et al. [75] showed that picosecond and nanosecond laser ablation of expanded PTFE caused thermal damage and agglomeration of nanostructures, but 130 fs pulses at 800 nm and 8 J/cm² preserved the fine pore network in the starting