Matrix laser cleaning: a new technique for the removal of nanometer sized particles from semiconductors

In this paper a new laser-based technique for the removal of nanoparticles from silicon wafers, called matrix laser cleaning, is introduced. In contrast to the already existing technique dry laser cleaning damages of the substrate can be avoided. Furthermore no liquids are used, avoiding problems that occur, e.g. in steam laser cleaning and other wet cleaning techniques. We show that damage free particle removal of polystyrene particles with diameters of at least down to 50 nm is possible with a cleaning efficiency very close to 100% within a single shot experiment. Furthermore the cleaning threshold is independent of the particle size. PACS 64.70.Hz; 68.43.Vx; 81.65.Cf     

Acoustic substrate expansion in modelling dry laser cleaning of low absorbing substrates

Acoustic expressions have been derived for the thermal expansion of substrate surfaces due to irradiation by an exponential laser pulse. The result of acoustic effects on three substrates (silicon, glass and silica) with different absorptions has been calculated.It has been shown that for substrates having relatively low absorptions, like silica and glass, acoustic considerations substantially reduce thermal expansion of the substrate caused by irradiation by nanosecond laser pulses relative to a quasi-static expansion model. In particular, the expansion of the substrate occurs over a much longer time frame than when the quasi-static approximation holds. Consequently, acceleration of the substrate surface is greatly reduced and laser cleaning threshold fluences for particle removal are increased.The predictions of the model of Arnold et al. when developed for acoustic considerations give reasonable agreement with experimentally found threshold fluences for alumina particles on silica and glass substrates although it underestimates the ratio of the threshold cleaning fluences of silica and glass. This could be due to the model underestimating the contribution of surface expansion to the laser cleaning process. The influence of multiple reflections in the substrate and departure from one dimensionality in the heat conduction on the threshold fluence was found to be insignificant. Thermal contact between the particle and the substrate was also found to have little effect on laser cleaning threshold fluences. Another mechanism that may enhance surface expansion is the 3D focussing of radiation by the particles. PACS 42.62.Cf; 81.65.Cf; 42.55.Lt     

Enhanced laser cleaning via direct line beam irradiation

A fluence advantage was achieved in dry/damp laser cleaning by reduction of the laser beam dimensions. 0.1 m Al₂O₃ particles were removed from glass slides using a KrF excimer laser (248 nm). As the width of the rectangular beam was reduced, a decrease in the threshold fluence required for particle removal was observed. Modelling based on the simplified thermal–mechanical response of the substrate and particle to the laser pulse does not describe the experimental results presented here. The case of dynamic expansion must be further considered, but it is believed that other mechanisms are involved in a full explanation. PACS 42.62.Cf     

Modelling laser cleaning of low-absorbing substrates: the effect of near-field focussing

A three-dimensional model for laser cleaning of spherical, transparent particles on low-absorbing substrates has been developed. It takes into account near-field focussing of the laser radiation by the particles. The intensity distribution under a particle was found using Mie theory together with the geometrical optics approximation. This permits the estimation of the beam width at the substrate surface and the focal distance of the radiation coming from the spherical particle. These parameters are used to find the distribution of intensity within the low-absorbing substrate from the formula for a focussed Gaussian beam. This is in contrast with most other models of laser cleaning, which assume that all absorption occurs at the surface of the substrate. The energy criterion was used to calculate the threshold fluence. The model predicts threshold fluences of the order of 10³ J/cm² for silica spheres having a diameter of the order of a micron on silica substrates, assuming adhesion by van der Waals force. As this is well above the damage threshold for silica, it effectively predicts that laser cleaning of silica spheres from silica will be impossible. For glass slides the threshold fluence is predicted to be a factor of 10^-4 times smaller than that for silica slides (about 0.1 J/cm²). This is due to the much higher absorption of glass compared to that of silica at 248 nm. PACS 42.62.Cf; 81.65.Cf     

The coalescence of supported gold nanoparticles induced by nanosecond laser irradiation

Gold nanoparticles supported on a quartz substrate are fabricated by e-beam lithography and subjected to nanosecond laser irradiation to induce coalescence. In combination with ex-situ scanning electron microscopy and optical extinction spectra, transmission is monitored in-situ to determine the characteristic coalescence time, the threshold fluence for the onset of coalescence and the produced particle morphology. Parallel molecular dynamics simulation is employed and shown to reasonably predict the coalescence time, thus explaining the experimental results. PACS 82.60.Qr; 42.62.Cf; 81.07.-b; 64.70.Nd     

Removal mechanisms of micro-scale particles by surface wave in laser cleaning

This paper is to investigate the mechanisms of micro-scale particle removal by surface wave, which was induced by a short pulse laser in a cleaning process. The authors analyzed the adhesive forces of particles on substrate surface and the clearance force produced by surface wave in laser cleaning. The physical model of particle removal by laser-induced surface wave was established to predict the removal area and the processing conditions of laser cleaning. In this research, a KrF excimer laser was applied to irradiate 304 stainless steel specimen distributed with copper particles to generate surface wave for copper particle removal. Considering that a time-varying and uniformly distributed heat source irradiates on material surface with thermao-elastic behavior, the displacement and acceleration of substrate induced by a pulsed laser were solved by an uncoupled thermal–mechanical analysis based on the finite element method. The processing parameters such as laser energy, laser spot size are discussed, respectively. A series of laser cleaning experiments were designed to compare with computation results. The results show that the removal area by surface wave beyond the laser spot increases with the laser energy and that, the surface acceleration decreases with the increase of the laser spot size.     

Modeling laser-assisted particle removal using molecular dynamics

Laser-assisted particle removal, a method of cleaning nano- to micro-scale particles from surfaces, was modeled using molecular dynamics. A two-dimensional molecular model consisting of substrate, particle, and adsorbed fluid was used. In order to obtain statistical accuracy of cleaning efficiencies, over 1200 particle-removal simulations were conducted. The effects of fluid thickness and substrate temperature were both considered, and good qualitative agreement with experimental results was obtained. The molecular dynamics approach is shown to be an effective way to study these systems. PACS 81.65.Cf; 79.20.Ds     

Optical microscopy imaging and image-analysis issues in laser cleaning

In earlier work we characterised single-pulse laser cleaning of medium-density (areal coverage 10–20%) alumina particles from glass surfaces with three different lasers. The method of measuring particle-removal efficiency involves optical microscopy imaging (digital) and subsequent image analysis of the area covered by, and/or the number of, particles before and after the single-pulse laser cleaning treatment. The sample-preparation technique used leads to both single particles and agglomerates on the surface. The issues of depth of focus in the imaging and grey-scale thresholding in the image analysis of such samples, with a range of particle and agglomerate sizes, has been systematically investigated. A protocol for optimum imaging and illumination of such samples is described herein. It has been developed based on the results of a systematic investigation of the effect of the image focal plane position relative to the surface also described herein. The image analysis to quantify the particles on the surface involves a judgement of the best threshold grey-scale level in the image to define the boundary between particles and background substrate. A quantitative appraisal of the impact on the laser cleaning efficiency results of a threshold grey-scale level that is set too high or too low, and how these results compare with those obtained for the best-judgement threshold grey-scale level, has been completed. PACS 81.65.Cf; 42.62.Cf; 42.70.Ce     

Ultraviolet laser removal of small metallic particles from silicon wafers

Laser removal of small 1 μm sized copper, gold and tungsten particles from silicon wafer surfaces was carried out using ultraviolet radiation at 266 nm generated by Nd:YAG harmonic generation. Successful removal of both copper and gold particles from the surface was achieved whereas tungsten particles proved to be difficult to remove. The cleaning efficiency was increased with an increase of laser fluence. The optimum processing window for safe cleaning of the surface without any substrate damage was determined by measuring the damage threshold laser fluence on the silicon substrate and the required fluence for complete removal of the particles. The different cleaning efficiencies with particle type are discussed by considering the adhesion force of the particle on the surface and the laser-induced cleaning force for the three different particles.     

Colour changes of a historical Gotland sandstone caused by laser surface cleaning in ambient air and N<Subscript>2</Subscript> flow

The surface discoloration due to laser cleaning was investigated for a historical Gotland sandstone. The difference in discoloration for cleaning performed in air and in the shielding environment of N₂ flowing at low velocities was studied by means of colorimetry and scanning electron microscopy and energy dispersive X-ray spectroscopy techniques. For ablative removal of the natural as well as artificially applied encrustation a pulsed 1064-nm laser operated at a fluence of 0.5 J/cm² was applied. It was observed that the natural colour variations (ΔL^*=21; Δb^*=23) of the stone can completely screen the laser-induced changes. Under conditions of shielding with nitrogen, darkening and yellowing slightly stronger than those occurring in the ambient air were revealed for the laser-cleaned, artificially crusted samples and the effect was independent of the gas-flow velocity. The observed difference confirmed the contribution of iron oxidation to the laser-induced yellowing and showed that the presence of oxygen in the ambient air affects favourably the cleaning by supporting removal of a variety of combustible surface remnants and crust components of organic as well as inorganic origin. PACS  42.62.-b; 52.70.Kz; 81.65.Cf     

Enhanced cleaning of photoresist film on a transparent substrate by backward irradiation of a Nd:YAG laser

Laser cleaning of a photoresist (PR) on a glass substrate using ns-pulsed Nd:YAG laser was studied. The direction of the substrate facing the laser beam was varied as a main parameter as well as the power of the laser beam. The backward irradiation (BWI) of the third harmonic beam (355 nm) completely removed 1.2 μm thick PR layer with three pulses at 1.5 J/cm² leaving no residues behind; while the forward irradiation (FWI) at the same condition just partially cleaned it. To investigate the difference of removal mechanisms between irradiation directions, the size distributions of particulates generated during laser cleaning were observed using an optical particle counter. The concentration of micron-sized particulates increased with increasing laser fluence up to 1 J/cm² for FWI and 0.5 J/cm² for BWI and then decreased at higher fluences because the target was a very thin film. The concentration of larger particulates for BWI was much higher than that for FWI implying the difference in removal mechanisms. In consideration of the size characteristics of the particulates and the temperature profiles of the PR layer, the most probable distinct mechanism for the BWI would be a blasting due to high temperature at the PR/glass interface. The particulate number concentration decreased rapidly after the completion of cleaning, suggesting that the measurement of the particulate concentration could detect the progress of the cleaning. Our results demonstrated that the backward irradiation will be useful for the laser cleaning of film-type contaminants on an optically transparent substrate.     

Time-of-flight measurements of ejected particles during dry laser cleaning

We propose an experimental approach which allows the characterization of the dynamics of the ejected particles in dry laser cleaning. Submicron silica particles on silicon substrates were illuminated by single nanosecond laser pulses at fluences which lead to particle removal. Time- and space-resolved scattered signal detection was demonstrated as a suitable technique to perform time-of-flight analyses of the ejected particles. The determination of the resulting detachment velocity at the particle removal threshold fluence contributes to a better understanding of mechanisms involved in dry laser cleaning. In particular, the present study evidences that the removal efficiency of the laser process is not based on the thermal expansion of materials. PACS 42.62.b; 42.15.Eq     

Ablative thresholds in laser cleaning of substrates from particulates

Laser cleaning thresholds based on the local ablation of substrate material are studied theoretically. Results are compared with the experimental data on the cleaning of silicon wafers from spherical silica particles using laser wavelengths at 248, 532 and 1064 nm. Calculations take into account local enhancement in the laser-light intensity and are based on analytical solutions for the temperature distribution. Influence of vapor atmosphere on cleaning thresholds is studied experimentally and theoretically. Here cleaning is assisted by explosive vaporization of capillary condensed water. A possibility to increase the window for damage-free cleaning by varying the pulse duration and laser wavelength is also discussed. PACS 42.62.Cf; 81.65.Cf; 68.35.Np     

Excimer laser irradiation at 248 nm of wooden archaeological objects and polymeric consolidants used in conservation: a study of cone formation and optimum cleaning parameters

Samples of a prehistoric consolidated vegetable basketry, coming from an important pile building in the North-East of Italy (Fiavé Carrera, Trento), had been irradiated with excimer laser pulses to remove the consolidant layer and restore the artifact. Some problems, related to the nature of the consolidant resin, were found in the preliminary laser cleaning process. The two main obstacles were related to the unknown composition of the polymeric resin and to the irregular surface morphology of the archaeological find. We observed that large numbers of laser pulses, for selected laser parameters, strongly modify the resin surface morphology with formation of a large number of cones. After cone formation, the surface was so irregular that it was almost impossible to complete the laser cleaning procedure. To find a solution to the problem, we here try to understand the mechanisms of cone formation in the present polymeric material. Scanning electron microscopy (SEM) and Fourier transformation infrared spectroscopy (FTIR) have been used to study surface morphology and chemical modification after selected irradiation processes. To understand the influence of laser irradiation on surface modifications and to establish a cleaning procedure for this specific application, investigations have been carried out separately on models of resin consolidants and wood (both dried and buried wood sample). The consolidants were commercial resins, typically used in archaeological conservation (Plexisol and Paraloid), deposited on a silicon substrate. Silicon was used because it is an appropriate substrate in FT-IR analysis. Finally, we chose wood because of its wide use in prehistoric artefact (Viburnum lantana). PACS  81.65.Cf; 82.50.Hp; 87.64.Ee; 87.64.km     

Plume optical transmission studies of a thin 2-propanol layer lifting off from a laser-heated Si substrate

Plume velocities were measured in air by an optical transmission technique as a function of laser fluence using a KrF laser for explosive boiling and lift-off of a thin variable thickness 2-propanol layer on a laser-heated Si substrate of interest for laser-assisted particle removal. The plume velocities were found to diminish with an increase in the thickness of the 2-propanol layer. This dependence was explained on the basis of the the momentum conservation rule, assuming explosive expansion of the superheated alcohol layer under spinodal conditions. PACS 05.70; 47.40.-x; 64.60.My; 64.70.Fr; 81.65.Cf     

激光等离子体去除微纳颗粒的热力学研究

微杂质污染一直是影响精密器件制造质量和使用寿命的关键因素之一.对于微纳米杂质颗粒用传统的清洗方式(超声清洗等)难以去除,而激光等离子体冲击波具有高压特性,可以实现纳米量级杂质颗粒的去除,具有很大的应用潜力.本文主要研究了激光等离子体去除微纳米颗粒过程中的热力学效应:实验研究了激光等离子体在不同脉冲数下对Si基底上Al颗粒去除后的颗粒形貌变化,发现大颗粒会发生破碎而转变成小颗粒,一些颗粒达到熔点后发生相变形成光滑球体,这源于等离子体的热力学效应共同作用的结果.为了研究微粒物态转化过程,基于冲击波传播理论研究,得到冲击波压强与温度特性的演化规律;同时,利用有限元模拟方式研究激光等离子冲击波压强和温度对微粒作用规律,得到了颗粒内随时间变化的应力分布和温度分布,并在此基础上得到等离子体对颗粒的热力学作用机制.     

KrF excimer laser dry and steam cleaning of silicon surfaces with metallic particulate contaminants

KrF excimer laser-assisted dry and steam cleaning of single-crystal silicon wafers contaminated with three different types of metallic particles was studied. The laser fluence used was 0.3 J/cm². In the dry process, for samples cleaned with 100 laser pulses the cleaning efficiency was 91, 71 and 59% for Au, Cu and W particles, respectively, whilst in steam cleaning the efficiency is about 100% after 5 laser pulses, independently of the type of contaminant. The effects of laser irradiation on the Si surface are investigated by scanning electron microscopy (SEM), X-ray photoelectron spectroscopy (XPS) and atomic force microscopy (AFM). Laser processing at 0.3 J/cm² does not deteriorate the Si-wafer surface, either in dry or steam cleaning. However, the measured XPS intensity coming from the metallic component is greater on the cleaned surfaces than in the initial condition. Quantification of the XPS results, assuming a stratified overlayer model for the detected species and accounting for the presence of the metallic particles on the surface, showed that the obtained results can be explained by the formation of a fractional metallic monolayer on the cleaned surfaces, due to partial vaporisation of small particles initially present on the sample surface. This contamination of the substrate could be considered excessive for some applications and it shows that the process requires careful optimisation for the required efficiency to be achieved without degradation of the substrate. Received: 14 January 2001 / Accepted: 19 February 2001 / Published online: 20 June 2001     

Limit analysis for laser removal of micron contaminant colloidal silicon dioxide particles from the super-smooth optical glass substrate by pulse Nd:YAG laser

Multimode Nd:YAG pulse laser was applied to remove micron and submicron particles by vaporizing a thin paint film pre-coated on super-smooth optical substrate surface. By analyzing the poor absorption of the optical glass substrate to the irradiative Nd:YAG pulse laser, the removal mechanism of contaminated colloidal particles from the super-smooth surface through vaporization of a volatile solid film is described. A limit analysis was proposed to determine the lower and the upper threshold of laser fluence for cleaning the SiO₂ contaminants from super-smooth K8 optical substrate. Relevant experiments on laser cleaning of micron-polishing particles from super-smooth K8 optical substrate confirmed the usefulness of this method in assisting the selection of effective cleaning fluence for accomplishing high cleanliness, which was in a range of 80–90% of the predicted upper threshold.     

Polymer jacket stripping of optical fibres by laser irradiation

Laser jacket stripping of the two-layer polymer jacket coating of Corning SMF-28 silica fibres has been studied as an alternative approach to chemical and mechanical techniques. These polymer outer layers, although chemically similar, become discernable through laser ablation and depth per pulse experiments. Etch rate measurements using nanosecond UV excimer laser sources (F₂, ArF, KrF and XeCl lasers) reveal that, as expected, the threshold fluence (energy per unit area) for significant material removal drops as the laser wavelength becomes shorter. For some wavelengths and fluences, spontaneous cone formation has been observed, thus providing additional threshold data through apex angle determination. The possible occurrence of deleterious damage to the silica cladding has been assessed using electron microscopy and optical transmission measurements. A wide fluence range over which damage was not observed characterised all the interactions. Irradiation techniques for producing apertures in the polymer coating or complete jacket removal are demonstrated and discussed. Briefly, polymer jacket stripping using a femtosecond laser source (800 nm) has been demonstrated. PACS 42.62.Cf; 52.38.Mf