JP2019532815A - Laser processing apparatus and method of laser processing workpiece - Google Patents

Abstract

A method of processing a workpiece having a first surface and a second surface opposite the first surface includes a first laser pulse having a pulse duration of less than 200 ps with a pulse repetition rate higher than 500 kHz. A beam is generated, a first laser pulse beam is irradiated along a beam axis that intersects the workpiece, and the beam axis is scanned along a processing locus. The beam axis is scanned such that a continuously irradiated laser pulse strikes the workpiece with a non-zero byte size to form a feature on the first surface of the workpiece. One or more such as bite size, pulse duration, pulse repetition rate, laser pulse spot size and laser pulse energy to ensure that the feature has a processed workpiece surface with an average surface roughness (Ra) of 1.0 μm or less Parameters are selected.

Description

Cross-reference to related applications

  This application claims the benefit of US Provisional Patent Application No. 62 / 368,053, filed July 28, 2016, which is incorporated by reference in its entirety.

background

I. TECHNICAL FIELD The present disclosure relates generally to processing materials using pulsed lasers and high repetition rates.

Related applications

II. 2. Description of Related Art Materials processing applications such as thin silicon wafer dicing, printed circuit board (PCB) drilling, solar cell manufacturing, and flat panel display manufacturing use similar material processing methods and are similar Have a serious problem. Previous solutions used mechanical processing methods and lithographic processing methods. However, the industry is moving in the direction of adopting laser processing methods because device size is reduced, device complexity is increased, and environmental costs are associated with chemical processing. Nowadays, high-power diode-pumped solid-state lasers with a typical wavelength of 1 μm and wavelength-converted to green or UV are being used. One method that has been utilized in some applications is to gradually cut the workpiece using repetitive passes at a relatively fast scan speed. There are three main problems in such applications. That is, (a) Debris is generated and accumulated at or near the processing location, (b) Large heat affected zone (HAZ) is generated, (c) Commercially feasible sufficiently high material removal rate Is to realize. As used herein, the term “debris” shall mean workpiece material that is ejected from a processing location (in either solid, liquid, or gaseous form) during laser processing and is recast. , Slugs, redeposits, etc., and other terms are commonly used. HAZ refers to the area of the workpiece whose microstructure or other chemical, electrical or physical properties have been altered by the heat generated during laser processing.

  Various options have been proposed to perform efficient and high-quality laser machining on workpieces, for example, lasers are used to generate ultrashort pulse duration laser pulses at high repetition rates. . As a result, less debris is generated than when a laser pulse having a relatively long pulse width is used, and the HAZ generated in the workpiece is relatively small. However, debris still occurs in the method using an ultrashort laser pulse generated at a high repetition rate. In some applications, debris accumulation can be a problem when the resulting debris accumulates, resulting in undesirable rough or uneven surfaces, or undesirable stress concentrations.

  Conventionally, accumulated debris has been accumulated by exposing the processed workpiece to a chemical etchant, or by cleaning the processed workpiece in an ultrasonic bath (eg made of DI water), or similar methods. Can be removed. This problem can also be solved by coating the workpiece with a sacrificial material layer, in which case debris generated during laser processing accumulates on the sacrificial layer and the sacrificial layer is laser processed. Can be removed after. However, in such a method, additional processing steps and additional consumable materials are added, thereby reducing throughput and increasing costs. Therefore, a solution that does not require removal of such debris would be preferred.

Overview

  One embodiment of the present invention provides a workpiece having a first surface and a second surface opposite the first surface, with a pulse repetition rate greater than 500 kHz and a pulse duration of less than 200 ps. Characterized in that the first laser pulse beam is generated, irradiated with the first laser pulse beam along a beam axis intersecting the workpiece, and the beam axis is scanned along a processing locus. it can. The beam axis is scanned such that a continuously irradiated laser pulse strikes the workpiece with a non-zero byte size to form a feature on the first surface of the workpiece. Such as byte size, pulse duration, pulse repetition rate, laser pulse spot size and laser pulse energy to ensure that the feature has a processed workpiece surface with an average surface roughness (Ra) of 1.0 μm or less One or more parameters are selected.

  In one embodiment, the pulse duration of each of the laser pulses in the first laser pulse beam is 1 ps or less, 800 fs or less, 750 fs or less, 700 fs or less, 650 fs or less, or 600 fs or less.

  In one embodiment, the pulse repetition rate of the laser pulse in the first laser pulse beam is higher than 1200 kHz, higher than 1250 kHz, higher than 1300 kHz, higher than 1400 kHz, higher than 1500 kHz, 1600 kHz. It is higher, higher than 1700kHz, higher than 1800kHz, higher than 1900kHz, higher than 2000kHz, or higher than 3000kHz.

  In an embodiment, the average surface roughness (Ra) is 0.75 μm or less, 0.5 μm or less, 0.4 μm or less, 0.3 μm or less, 0.25 μm or less, 0.2 μm or less, 0.15 μm or less, or the like Between any of the values.

  In one embodiment, the method generates a second laser pulse beam (after the feature is formed on the first surface of the workpiece), and a laser pulse in the second laser pulse beam. To produce a beam waist, and the focused second laser pulse beam is applied to the processed laser beam so that the beam waist is disposed in the workpiece or on the second surface of the workpiece. It can be further characterized as including the operation of irradiating along the beam axis intersecting the workpiece surface and machining the workpiece at or near the beam waist. In one embodiment, the workpiece is more transparent to the wavelength of the laser pulse in the second laser pulse beam than to the wavelength of the laser pulse in the first laser pulse beam. is there.

FIG. 1 schematically shows an apparatus for processing a workpiece according to an embodiment of the present invention. 2 and 3 show photomicrographs (taken from the top) of the trenches formed in the silicon wafer. 2 and 3 show photomicrographs (taken from the top) of the trenches formed in the silicon wafer. FIG. 4 shows micrographs (taken from the top) of laser processed features each containing a set of intersecting scribe lines formed on the surface of the silicon wafer. FIG. 5 illustrates the average surface roughness (Ra) of the processed workpiece surface in the trench formed in the silicon wafer and the trench formation process by propagating laser pulses along the scan beam axis at different pulse repetition rates. FIG. 4 shows a set of graphs showing the relationship between the material removal rate as a function of byte size and fluence. FIG. 6 shows a set of graphs illustrating a process window for forming a trench in a silicon wafer that will result in a processed workpiece surface having certain characteristics. FIG. 7 shows a photomicrograph (taken from a side cross-section) of a silicon wafer that has been processed to form trenches in a manner that produces a smooth processed workpiece surface. 8A and 8B show micrographs (taken from the side cross-section) of the processed silicon wafer shown in FIG. 7 after the silicon wafer has been further processed to form trench cracks inside the silicon wafer. Yes. FIG. 8A shows a view across the width of the trench shown in FIG. 8A and 8B show micrographs (taken from the side cross-section) of the processed silicon wafer shown in FIG. 7 after the silicon wafer has been further processed to form trench cracks inside the silicon wafer. Yes. FIG. 8B is a view along the length direction of the trench shown in FIG. 9A-9D illustrate a method for machining a workpiece in one embodiment. 9A-9D illustrate a method for machining a workpiece in one embodiment. 9A-9D illustrate a method for machining a workpiece in one embodiment. 9A-9D illustrate a method for machining a workpiece in one embodiment.

Detailed description

  In this specification, examples of embodiments will be described with reference to the accompanying drawings. Unless explicitly stated otherwise, in the drawings, the size and position of components, features, elements, etc., and the distances between them, are not necessarily to scale and are exaggerated for clarity. Yes.

  The terminology used herein is for the purpose of describing particular example embodiments only and is not intended to be limiting. As used herein, the singular forms are intended to include the plural unless the content clearly dictates otherwise. Further, the terms “comprising” and / or “comprising”, as used herein, identify the presence of the stated feature, integer, step, action, element, and / or component. It should be understood, however, that it does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof. Except as otherwise noted, when a range of values is stated, the range includes not only the subrange between the upper and lower limits of the range, but also the upper and lower limits. Except where specifically indicated, terms such as “first” and “second” are only used to distinguish elements from each other. For example, one node can be referred to as a “first node”, and another node can be referred to as a “second node”, or vice versa. The section headings used herein are for organizational purposes only and are not to be construed as limiting the subject matter described.

  Except where specifically indicated, terms such as “about” and “before and after” are not accurate and need not be accurate in terms of quantity, size, formulation, parameters, and other quantities and characteristics, It may be rounded and / or larger and / or smaller as required or to reflect tolerances, conversion factors, fractional calculations, measurement errors, and other factors known to those skilled in the art. Means good.

  As used herein, spatially relative terms such as “lower”, “lower”, “lower”, “upper”, and “upper” refer to other elements or features as shown in the figures. Can be used for ease of explanation when describing relationships to elements or features. It should be understood that spatially relative terms are intended to include different orientations in addition to those shown in the figures. For example, an element described as being “below” or “below” another element or feature will be directed “above” the other element or feature if the object in the figure is flipped . Thus, the exemplary term “downward” is intended to include both upper and lower orientations. If the object is oriented in another direction (for example, rotated 90 degrees or in another direction), the spatially relative descriptors used in this document are interpreted accordingly. obtain.

  Like numerals refer to like elements throughout the drawings. Thus, the same or similar numbers may be described with reference to other drawings, even if not mentioned or described in the corresponding drawings. In addition, even elements that are not given reference numerals may be described with reference to other drawings.

  It is understood that many different forms, embodiments, and combinations are possible without departing from the spirit and teachings of the present disclosure and that the present disclosure should not be construed as limited to the embodiments set forth herein. I can do it. Rather, these examples and embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

I. Overview The embodiments described herein generally include a method for laser processing (herein also referred to as laser processing of a workpiece, processing with a laser, or most simply “processing”) and It relates to the device. Generally, the whole or part of the processing is performed by irradiating the workpiece with laser radiation to heat, melt, evaporate, ablate, crack, or polish the workpiece. Done. Specific examples of processes that can be performed by the illustrated apparatus include via drilling, scribing, dicing, engraving and the like. Thus, features that can be formed on or in the workpiece as a result of processing include openings, vias (eg, non-through vias, through vias, slot vias), grooves, trenches, scribe lines, cuts, and the like. It may include grooves, recesses, etc., or any combination thereof.

Workable workpieces can be comprehensively characterized as metals, polymers, ceramics, or any combination thereof. Specific examples of workable workpieces include integrated circuits (IC), IC packages (ICP), light emitting diodes (LEDs), LED packages, semiconductor wafers, electronic or optical device substrates (eg, Al ₂ O ₃ , AlN, A substrate formed of BeO, Cu, GaAS, GaN, Ge, InP, Si, SiO ₂ , SiC, Si _1-x Ge _x (0.0001 <x <0.9999), or any combination thereof or an alloy thereof ), Plastic, glass (eg, non-tempered glass or heat tempered glass or chemically tempered glass or others), quartz, sapphire, plastic, silicon and the like. Therefore, the material that can be processed is one or more metals (for example, Al, Ag, Au, Cu, Fe, In, Mg, Pt, Sn, Ti, etc., or any combination thereof or an alloy thereof), conductive Metal oxides (eg ITO), transparent conductive polymers, ceramics, waxes, resins, substrate materials (eg Al ₂ O ₃ , AlN, BeO, Cu, GaAs, GaN, Ge, InP, Si, SiO ₂ , SiC, Si _1-x Ge _x, etc., or combinations or alloys thereof) (for example, silicon oxide, silicon nitride, silicon oxynitride, etc., or any combination thereof) Inorganic dielectric materials used, low-k dielectric materials (eg, methyl silsesquioxane (MSQ), hydrogen silsesquioxane (HSQ), tetraethyl fluoroorthosilicate (FTEOS), etc., or any combination thereof Food) Organic dielectric materials (eg SILK, benzocyclobutene, Nautilus (all manufactured by Dow), polyfluorotetraethylene (manufactured by DuPont), FLARE (manufactured by Allied Chemical), etc. Or any combination of these), fiber glass, polymer materials (polyamide, polyimide, polyester, polyacetal, polycarbonate, modified polyphenylene ether, polybutylene terephthalate, polyphenylene sulfide, polyethersulfone, polyetherimide, polyetherether) Ketones, liquid crystal polymers, acrylonitrile butadiene styrene, and any compounds, composites, or mixtures thereof), or any combination thereof.

II. System-Overview FIG. 1 schematically illustrates an apparatus for processing a workpiece, according to one embodiment of the present invention.

  Referring to the embodiment shown in FIG. 1, an apparatus 100 for processing a workpiece 102 includes a laser source 104 for generating laser pulses, a first positioner 106, a second positioner 108, and a first positioner. 3 positioners 110, a scan lens 112, and a controller 114. In view of the following description, including the first positioner 106 is optional if the device 100 includes the second positioner 108 (ie, the device 100 needs to include the first positioner 106). It should be understood. Similarly, if the device 100 includes the first positioner 106, the inclusion of the second positioner 108 is optional (ie, the device 100 need not include the second positioner 108). Should be understood. Finally, it should be understood that the inclusion of the third positioner 110 is optional (ie, the device 100 need not include the third positioner 108).

  Although not shown, the apparatus 100 focuses, expands, collimates, and laser pulses generated by the laser source 104 along one or more beam paths (eg, beam path 116) to the scan lens 112, One or more optical elements (eg, beam expanders, beam shapers, apertures, harmonics) for shaping, polarizing, filtering, splitting, combining, cropping, modifying, adjusting, and directing Generating crystal, filter, collimator, lens, mirror, polarizer, wave plate, diffractive optical element, or any combination thereof). U.S. Patent Nos. 4,912,487, 5,633,747, 5,638,267, 5,751,585, 5,847,960, 5,917,300, 6,314,473, 6,430,465, 6,700,600, 6,706,998, 6,706,999, 6,816,294 No. 6,947,454, No. 7,019,891, No. 7,027,199, No. 7,133,182, No. 7,133,186, No. 7,133,187, No. 7,133,188, No. 7,245,412, No. 7,259,354, No. 7,611,745, No. 7,834,8,8,8 No. 8,158,493, No. 8,288,679, No. 8,404,998, No. 8,497,450, No. 8,648,277, No. 8,680,430, No. 8,847,113, No. 8,896,909, No. 8,928,853 or the above-mentioned U.S. Patent Application Publication No. 2014/0026351, 2014/0197140, 2014/0263201, 2014/0263212, 2014/0263223, 2014/0312013, or German Federal Patent DE102013201968B4, or International Patent Publication WO2009 / 087392, or these As disclosed in any combination of the above It may be provided with one or more of cement, or could further understood that the apparatus 100 may comprise one or more additional components. Each of these publications is incorporated herein by reference in its entirety.

Laser pulses propagating through the scan lens 112 propagate along the beam axis and irradiate the workpiece 102. The laser pulse applied to the workpiece 102 may be characterized as having a Gaussian or shaped (eg, “top hat”) spatial intensity profile. The spatial intensity profile may also be characterized as the cross-sectional shape of the laser pulse propagating along the beam axis (or beam path 116). The cross-sectional shape may be a circle, an ellipse, a rectangle, a triangle, a hexagon, a ring shape, or any other shape. In addition, the laser pulses so irradiated can irradiate the workpiece 102 with a spot size in the range of 2 μm to 200 μm. As used herein, the term “spot size” refers to the area of the workpiece 102 (“work spot”, “process spot”, “spot position” or more simply processed by the irradiating laser pulse). It means the diameter or spatial width of the irradiating laser pulse at a position where the beam axis crosses (also called “spot”). For the discussion herein, the spot size is measured as the radial or transverse distance from the beam axis to where the light intensity drops to 1 / e ² of the light intensity at the beam axis. In general, the spot size of the laser pulse is minimized at the beam waist. However, it will be understood that the spot size can be smaller than 2 μm or larger than 200 μm. In this way, at least one laser pulse irradiated to the workpiece 102 is 2 μm, 3 μm, 5 μm, 7 μm, 10 μm, 15 μm, 30 μm, 35 μm, 40 μm, 45 μm, 50 μm, 55 μm, 80 μm, 100 μm, 150 μm, 200 μm, etc. Can have a smaller spot size, a larger spot size, a spot size equal to these, or a spot size between any of these values. In one embodiment, the laser pulse applied to the workpiece 102 may have a spot size in the range of 25 μm to 60 μm. In other embodiments, the laser pulse applied to the workpiece 102 may have a spot size in the range of 35 μm to 50 μm.

A. Laser Source Generally, the laser source 104 can generate laser pulses. Thus, the laser source 104 can include a pulsed laser source, a QCW laser source, or a CW laser source. If the laser source 104 includes a QCW laser source or a CW laser source, the laser source 104 may be a pulse gating unit (eg, acousto-optic) that temporally modulates the beam of laser radiation output from the QCW laser source or CW laser source. (AO) modulator (AOM), beam chopper, etc.). Although not shown, the apparatus 100 requires one or more harmonic generation crystals (also known as “wavelength conversion crystals”) configured to convert the wavelength of the light output by the laser source 104. Can be included. Therefore, the laser pulse finally irradiated to the workpiece 102 is ultraviolet light (UV), visible light (for example, green), infrared light (IR), near infrared light (NIR), short wavelength infrared light ( SWIR), characterized as having one or more wavelengths in one or more of the electromagnetic spectrum in the range of mid-wavelength infrared light (MWIR), or long-wavelength infrared light (LWIR), or any combination thereof It may be done.

  The laser pulses output by the laser source 104 can have a pulse width or pulse duration in the range of 30 fs to 500 ps (ie, based on the full width at half maximum (FWHM) of optical power over time). However, it will be appreciated that the pulse duration may be shorter than 10 fs or longer than 500 ps. In this way, at least one laser pulse output from the laser source 104 is 10fs, 15fs, 30fs, 50fs, 75fs, 100fs, 150fs, 200fs, 300fs, 500fs, 700fs, 750fs, 800fs, 850fs, 900fs, 1ps, 2 ps, 3 ps, 4 ps, 5 ps, 7 ps, 10 ps, 15 ps, 25 ps, 50 ps, 75 ps, 100 ps, 200 ps, 500 ps, shorter pulse durations, longer pulse durations, pulse durations equal to these, or these Can have a pulse duration between any of the values. In one embodiment, the laser pulse output by the laser source 104 has a pulse duration in the range of 10 fs to 1 ps. In other embodiments, the laser pulse output by the laser source 104 has a pulse duration in the range of 500 fs to 900 fs.

  The laser pulse output by the laser source 104 can have an average power in the range of 100 mW to 50 kW. However, it will be appreciated that the average power may be less than 100 mW or greater than 50 kW. Thus, the laser pulses output from the laser source 104 are 100 mW, 300 mW, 500 mW, 800 mW, 1 W, 2 W, 3 W, 4 W, 5 W, 6 W, 7 W, 10 W, 15 W, 25 W, 30 W, 50 W, 60 W, 100 W. , 150 W, 200 W, 250 W, 500 W, 2 kW, 3 kW, 20 kW, 50 kW, etc., or have an average power greater than or equal to, or an average power between any of these values.

  Laser pulses can be output by the laser source 104 at a pulse repetition rate in the range of 5 kHz to 1 GHz. However, it will be appreciated that the pulse repetition rate may be lower than 5 kHz or higher than 1 GHz. Thus, the laser pulse is generated by the laser source 104 at 5 kHz, 50 kHz, 100 kHz, 250 kHz, 500 kHz, 800 kHz, 900 kHz, 1 MHz, 1.5 MHz, 1.8 MHz, 1.9 MHz, 2 MHz, 2.5 MHz, 3 MHz, 4 MHz, 5 MHz, 10 MHz, Pulse repetition rate lower than 20MHz, 50MHz, 70MHz, 100MHz, 150MHz, 200MHz, 250MHz, 300MHz, 350MHz, 500MHz, 550MHz, 700MHz, 900MHz, 2GHz, 10GHz, etc., higher repetition rates, equal pulse repetition rates Or a pulse repetition rate between any of these values. In some embodiments, the pulse repetition rate can range from 1.5 MHz to 10 MHz.

In addition to wavelength, pulse duration, average power, and pulse repetition rate, the laser pulse irradiated to the workpiece 102 can be characterized by one or more other characteristics such as pulse energy, peak power, and the like. The laser pulse is a light intensity (measured in W / cm ² ) sufficient to process the workpiece 102 or its components to form one or more features having one or more desired characteristics (J It can be selected based on one or more other characteristics for illuminating the workpiece 102 at the process spot, such as at fluence (measured at / cm ² ). Examples of such other parameters include the material properties of workpiece 102, bite size, desired processing throughput, etc., or any combination thereof, wavelength, pulse duration, average power and pulse repetition rate. One or more of the above-described characteristics such as As used herein, the term “byte size” refers to the center-to-center distance between spot areas that are struck by continuously irradiated laser pulses.

  For example, a laser pulse applied to workpiece 102 may have a pulse energy in the range of 1 μJ to 20 μJ. In one embodiment, any laser pulse irradiated can have a pulse energy in the range of 2 μJ to 10 μJ. In other embodiments, any laser pulse irradiated may have a pulse energy in the range of 3 μJ to 6 μJ. However, it will be understood that the pulse energy of the irradiated laser pulse may be lower than 1 μJ or higher than 20 μJ. In other examples, the laser pulse applied to the workpiece 102 may have a fluence in the range of 1 μJ to 20 μJ. In one embodiment, any laser pulse irradiated can have a pulse energy in the range of 2 μJ to 10 μJ. In other embodiments, any laser pulse irradiated may have a pulse energy in the range of 2 μJ to 6 μJ. However, it will be understood that the pulse energy of the irradiated laser pulse may be lower than 1 μJ or higher than 20 μJ.

  Examples of laser types that can characterize the laser source 104 include gas lasers (eg, carbon dioxide laser, carbon monoxide laser, excimer laser, etc.), solid state lasers (eg, Nd: YAG laser, etc.), rod lasers, fiber lasers, etc. , Photonic crystal rod / fiber laser, passive mode-locked solid bulk or fiber laser, dye laser, mode-locked diode laser, pulsed laser (eg ms pulse laser, ns pulse laser, ps pulse laser, fs pulse laser), CW laser , QCW laser, etc., or any combination thereof. Specific examples of laser sources that can be provided as laser source 104 include BOREAS, HEGOA, SIROCCO or CHINOOK series lasers manufactured by EOLITE, PYROFLEX series lasers manufactured by PYROPHOTONICS, PALADIN manufactured by COHERENT Advanced 355 or DIAMOND series lasers, TRUFLOW series lasers manufactured by TRUMPF (eg, TRUFLOW 2000, 2700, 3200, 3600, 4000, 5000, 6000, 7000, 8000, 10000, 12000, 15000, 20000), TRUDISK Series, TRUPULSE series, TRUDIODE series, TRUFIBER series, or TRUMICRO series lasers, FCPAμJEWEL or FEMTOLITE series lasers manufactured by IMRA AMERICA, TANGERINE and SATSMA series lasers manufactured by AMPLITUDE SYSTEMES (and MIKAN and T-) PULSE series oscillators), CL series, CLPF series, CLPN series manufactured by IPG PHOTONICS, CLPNT series, CLT series, ELM series, ELPF series, ELPN series, ELPP series, ELR series, ELS series, FLPN series, FLPNT series, FLT series, GLPF series, GLPN series, GLR series, HLPN series, HLPP series, RFL series , TLM series, TLPN series, TLR series, ULPN series, ULR series, VLM series, VLPN series, YLM series, YLPF series, YLPN series, YLPP series, YLR series, YLS series, FLPM series, FLPMT series, DLM series, BLM Series or DLR series lasers (for example, including GPLN-100-M, GPLN-500-QCW, GPLN-500-M, GPLN-500-R, GPLN-2000-S, etc.), or the like, Alternatively, one or more laser sources such as an arbitrary combination thereof may be mentioned.

B. First Positioner The first positioner 106 is disposed, positioned or installed in the beam path 116 and diffracts, reflects, refracts, or the like, the laser pulses generated by the laser source 104. Or any combination of these can be moved to move the beam path 116 so that the beam axis is moved relative to the workpiece 102. In general, the first positioner 106 is configured to move the beam axis relative to the workpiece 102 along the X axis (or X direction) and the Y axis (or Y direction). Although not shown, the Y axis (or Y direction) may be understood to mean an axis (or direction) orthogonal to the illustrated X axis (or X direction) and Z axis (or Z direction).

The movement of the beam axis relative to the workpiece 102 made by the first positioner 106 generally causes a process spot within a first scan field or “first scanning range” extending from 0.01 mm to 4.0 mm in the X and Y directions. Limited to scan, move or position. However, the first scanning range includes (for example, the configuration of the first positioner 106, the position of the first positioner 106 along the beam path 116, the beam size of the laser pulse incident on the first positioner 106, the spot size. It will be understood that it may extend less than 0.01 mm in either the X or Y direction (depending on one or more factors such as, etc.) or may extend longer than 4.0 mm. Thus, the first scanning range is 0.04 mm, 0.1 mm, 0.5 mm, 1.0 mm, 1.4 mm, 1.5 mm, 1.8 mm, 2 mm, 2.5 mm, 3.0 mm, in either the X direction or the Y direction. It may extend a distance longer than 3.5 mm, 4.0 mm, 4.2 mm, etc., a distance equal to these, or a distance between any of these values. As used herein, the term “beam size” means the diameter or width of a laser pulse, where the optical intensity drops from the beam axis to 1 / e ² of the light intensity at the beam axis. Can be measured as a radial distance up to or a distance across.

  In general, the bandwidth by which the first positioner 106 can move the beam axis and thereby position the process spot (ie, the first positioning bandwidth) is from 50 kHz (or before and after) to 10 MHz (or around it). Is in range. In this way, the first positioner 106 has a positioning speed (obtained from the first positioning bandwidth) ranging from one spot position every 20 μs (or before and after) to one spot position every 0.1 μs (or before and after). The process spot can be positioned at any position within the first scanning range. The reciprocal of the positioning speed is referred to herein as “positioning time” and is necessary to change the position of the process spot from one position within the first scanning range to any other position within the first scanning range. Mean period. Thus, the first positioner 106 can be characterized by a positioning time ranging from 20 μs (or before and after) to 0.1 μs (or around). In one embodiment, the first positioning bandwidth is in the range of 100 kHz (or around) to 2 MHz (or around). For example, the first positioning bandwidth is 1 MHz (or around it).

The first positioner 106 incorporates a microelectromechanical system (MEMS) mirror or mirror array, an AO deflector (AOD) system, an electro-optic deflector (EOD) system, a piezoelectric actuator, an electrostrictive actuator, a voice coil actuator, etc. It may be a Fast Steering Mirror (FSM) element, or something similar, or any combination thereof. In one embodiment, the first positioner 106 includes at least one (eg, one, two, etc.) single element AOD system, at least one (eg, one, two, etc.) phased array AOD. It is an AOD system that includes a system or any combination thereof. Both AOD systems include AO cells formed from materials such as crystalline Ge, PbMoO ₄ or TeO ₂ , glassy SiO ₂ , quartz, As ₂ S ₃ . However, the former includes a single ultrasonic transducer acoustically coupled to the AO cell, and the latter includes a phased consisting of at least two ultrasonic transducers acoustically coupled to a common AO cell. Contains an array.

  Any of the AOD systems can be configured as a single axis AOD system (eg, configured to move the beam axis along a single direction) by deflecting the beam path 116 or (eg, in multiple directions) (Eg, configured to move the beam axis along the X and Y directions) may be provided as a multi-axis AOD system. In general, the multi-axis AOD system can be a multi-cell system or a single-cell system. Multi-cell multi-axis systems typically include multiple AOD systems that are configured to move the beam axis along different axes. For example, a multi-cell multi-axis system includes a first AOD system (eg, an “X-axis AOD system”) configured to move the beam axis along the X direction (eg, a single element or phased array AOD system); , A second AOD system (eg, a “Y-axis AOD system”) configured to move the beam axis along the Y direction (eg, a single element or phased array AOD system). Single cell multi-axis systems (eg, “X / Y-axis AOD systems”) typically include a single AOD system configured to move the beam axis along the X and Y directions. . For example, a single cell system can include at least two ultrasonic transducers acoustically coupled to different planes, facets, sides, etc. of a common AO cell.

C. Second Positioner Similar to the first positioner 106, the second positioner 108 is installed in the beam path 116, diffracts and reflects the laser pulse generated by the laser source 104 and passed through the first positioner 106. Refracting, or doing something similar, or any combination thereof, relative to the workpiece 102 (eg, along the X and Y directions) via movement of the beam path 116 relative to the scan lens 112 And can be operated to move the beam axis. The movement of the beam axis relative to the workpiece 102 performed by the second positioner 108 is generally a second scan field or “scanning range” that extends in the X and / or Y direction over a larger area than the first scanning range. The process spot is limited to be scanned, moved or positioned within. It should be understood that in the configuration described herein, the beam axis movement performed by the first positioner 106 can be superimposed on the beam axis movement performed by the second positioner 108. In this way, the second positioner 108 is operable to scan the first scanning range within the second scanning range.

  In one embodiment, the second scanning range extends from 1 mm to 50 mm in the X and / or Y direction. However, it will be appreciated that the second positioner 108 may be configured such that the second scanning range extends less than 1 mm or longer than 50 mm in either the X direction and / or the Y direction. Thus, in some embodiments, the maximum dimension of the second scanning range (e.g., in the X or Y direction, or other directions) is the feature formed on the workpiece 102 (e.g., via, trench, Or larger than the corresponding maximum dimension (measured in the XY plane) of the scribe line, recess, conductive trace, etc.). However, in other embodiments, the maximum dimension of the second scanning range may be less than the maximum dimension of the feature being formed.

  Generally, a bandwidth (ie, a second positioning bandwidth) that allows the second positioner 108 to move the beam axis to position the process (and thereby scan the first scanning range within the second scanning range). Is narrower than the first positioning bandwidth. In one embodiment, the second positioning bandwidth is in the range of 900 Hz to 5 kHz. In other embodiments, the first positioning bandwidth is in the range of 2 kHz to 3 kHz (eg, about 2.5 kHz). For example, the second positioner 108 is provided as a galvanometer mirror system that includes two galvanometer mirror components. One galvanometer mirror component is configured to move the beam axis along the X direction relative to the workpiece 102, and the other galvanometer mirror component is positioned along the Y direction relative to the workpiece 102. Configured to move. However, in other embodiments, the second positioner 108 may be provided as a rotating polygon mirror system or the like. Thus, it will be appreciated that depending on the particular configuration of the second positioner 108 and the first positioner 106, the second positioning bandwidth may be greater than or equal to the first positioning bandwidth.

D. Third Positioner The third positioner 110 is operable to move the workpiece 102 relative to the scan lens 112 and, as a result, move the workpiece 102 relative to the beam axis. Movement of the workpiece 102 relative to the beam axis generally scans, moves, or moves the process spot within a third scan field or “scanning range” that extends in the X and / or Y direction over an area larger than the second scanning range. Limited to allow positioning. In one embodiment, the third scanning range extends from 25 mm to 2 m in the X and / or Y direction. In other embodiments, the third scanning range extends from 0.5 m to 1.5 m in the X and / or Y direction. In general, the maximum dimension of the third scanning range (eg, in the X or Y direction, or other direction) is the corresponding maximum dimension of the feature formed on the workpiece 102 (measured in the XY plane). That's it. If desired, the third positioner 110 may be configured to move the workpiece 102 relative to the beam axis within a scanning range that extends in the Z direction (eg, over a range of 1 mm to 50 mm). For this reason, the third scanning range may extend along the X direction, the Y direction, and / or the Z direction.

  It is understood that in the configurations described herein, the movement of the beam axis performed by the first positioner 106 and / or the second positioner 108 can be superimposed on the movement of the workpiece 102 performed by the third positioner 110. Should. Thus, the third positioner 110 is operable to scan the first scanning range and / or the second scanning range within the third scanning range. Generally, the bandwidth over which the third positioner 110 can position the process spot (thus scanning the first scanning range and / or the second scanning range within the third scanning range) is the second positioning band. It is narrower than the width (for example, 10 Hz, before and after, or less).

  In one embodiment, the third positioner 110 includes one or more linear stages (eg, each capable of translating the workpiece 102 along the X, Y, and / or Z directions, respectively) (eg, each One or more rotary stages (similar to or capable of providing the workpiece 102 with rotational movement about an axis parallel to the X, Y and / or Z directions), or similar, or any combination thereof Offered as a thing. In one embodiment, the third positioner 110 is supported by an X-axis stage for moving the workpiece 102 along the X direction and the X-axis stage (which moves along the X direction by the X-axis stage). And a Y-axis stage for moving the workpiece 102 along the Y direction. Although not shown, the apparatus 100 may include an optional chuck coupled to a third positioner 110, on which the workpiece 102 can be clamped, secured, held, fixed or supported. . Although not shown, the apparatus 100 may include an optional base that supports the third positioner 110.

  As has been described, the apparatus 100 uses a so-called “stacked” positioning system. That is, the position of components such as the first positioner 106, the second positioner 108, and the scan lens 112 are relative to the workpiece 102 that is moved through the third positioner 110 (eg, as known in the art). Remain stationary within device 100 (via one or more supports, frames, etc.). In other embodiments, the third positioner 110 may be arranged and configured to move one or more components, such as the first positioner 106, the second positioner 108, the scan lens 112, and the workpiece 102. May remain stationary. In still other embodiments, the apparatus 100 includes one or more components such as a first positioner 106, a second positioner 108, a scan lens 112, etc. conveyed by one or more linear or rotary stages. A split axis positioning system may be used that is arranged and configured so that a linear or rotary stage moves the workpiece 102. In this manner, the third positioner 110 causes movement of the workpiece 102 in addition to movement of one or more of the first positioner 106, the second positioner 108, the scan lens 112, and the like. Examples of split axis positioning systems that can be advantageously or effectively utilized in apparatus 100 include U.S. Patent Nos. 5,751,585, 5,798,927, 5,847,960, 6,706,999, 7,605,343, 8,680,430, 8,847,113, Or any of those disclosed in US Patent Application Publication No. 2014/0083983, or any combination thereof. Each of these publications is incorporated herein by reference in its entirety.

  In other embodiments, one or more components, such as the first positioner 106, the second positioner 108, and the scan lens 112, are multi-axis articulated robot arms (eg, 2 axes, 3 axes, 4 axes, 5 axes, Or a 6-axis arm). In such an embodiment, the second positioner 108 and / or the scan lens 112 can be transported as an end effector of a robot arm, if desired. In still other embodiments, the workpiece 102 can be transported directly on the end effector of the multi-axis articulated robotic arm (ie, without the third positioner 110). In still other embodiments, the third positioner 110 can be transported on the end effector of a multi-axis articulated robot arm.

D. A scan lens (eg, provided as either a simple lens or a compound lens) scan lens 112 is typically a beam that typically produces a beam waist that can be positioned at a desired process spot. A laser pulse directed along the path is configured to focus. The scan lens 112 is an f-theta lens, a telecentric lens, an axicon lens (in this case, a series of beam waists are generated and a plurality of process spots shifted from each other along the beam axis), or the like Can be provided in any combination.

E. Controller Generally, the controller 114 includes one or more components of the device 100 such as a laser source 104, a first positioner 106, a second positioner 108, a third positioner 110, a lens actuator (eg, USB, Ethernet, These components are communicatively linked (via one or more wired or wireless communication links, such as Firewire, Wi-Fi, RFID, NFC, Bluetooth, Li-Fi, etc., or any combination thereof) Are operable in response to one or more control signals output by the controller 114.

  For example, the controller 114 performs relative movement between the beam axis and the workpiece, and moves the process spot and workpiece 102 along a trajectory (also referred to herein as a “process trajectory”) within the workpiece 102. The operation of the first positioner 106, the second positioner 108, or the third positioner 110 may be controlled to cause relative movement therebetween. Any two of these positioners, or all three of them, may have two positioners (eg, first positioner 106 and second positioner 108, first positioner 106 and third positioner 110, The second positioner 108 and the third positioner 110) or three positioners simultaneously cause relative movement between the process spot and the workpiece 102 (so that "composite relative movement between the beam axis and the workpiece" It will be understood that it may be controlled to produce “ Of course, at any point in time, a single positioner will cause a relative movement between the process spot and the workpiece 102 (which will result in a “non-composite relative movement” between the beam axis and the workpiece). It is also possible to control only (for example, the first positioner 106, the second positioner 108, or the third positioner 110). A control signal instructing compound or non-compound relative movement may be calculated in advance or determined in real time.

  In general, the controller 114 includes one or more processors configured to generate the control signals described above when executing instructions. A processor may be provided as a programmable processor (eg, including one or more general purpose computer processors, microprocessors, digital signal processors, etc., or any combination thereof) when configured to execute instructions. Instructions executable by the processor include software, firmware, etc., or programmable logic devices (PLDs), field programmable gate arrays (FPGAs), field programmable object arrays (FPOAs), application specific integrated circuits (ASICs) (digital) The circuit may be implemented in any suitable form (including circuits, analog circuits, mixed analog / digital circuits), or any combination thereof. Instruction execution may be performed on a single processor, distributed across multiple processors, performed in parallel across multiple processors within a device or across a network of devices, or the like. Or these may be combined arbitrarily.

  In one embodiment, the controller 114 includes a tangible medium such as a computer memory that is accessible by the processor (eg, via one or more wired or wireless communication links). As used herein, “computer memory” refers to magnetic media (eg, magnetic tape, hard disk drive, etc.), optical disk, volatile or non-volatile semiconductor memory (eg, RAM, ROM, NAND flash). Memory, NOR type flash memory, SONOS memory, etc.), which may be locally accessible, remotely accessible (eg, via a network), or a combination thereof. In general, instructions may be stored as computer software (eg, executable code, files, instructions, library files, etc.). Such computer software is written by, for example, C, C ++, Visual Basic, Java, Python, Tel, Perl, Scheme, Ruby, etc., and can be easily created from the descriptions given herein by those skilled in the art. it can. Computer software is typically stored in one or more data structures transmitted by computer memory.

  Although not shown, one or more drivers (eg, an RF driver, a servo driver, a line driver, a power supply, etc.) are connected to the laser source 104, the first positioner 106, the second positioner 108, the third positioner 110, It can be communicatively coupled to the input of one or more components, such as a lens actuator. In one embodiment, each driver typically includes an input to which the controller 114 is communicatively coupled so that the controller 114 generates one or more control signals (eg, a trigger signal, etc.). It is possible. This control signal may be communicated to one or more driver inputs associated with one or more components of device 100. In this manner, components such as the laser source 104, the first positioner 106, the second positioner 108, the third positioner 110, and the lens actuator are adapted to respond to control signals generated by the controller 114.

  In other embodiments, not shown, one or more additional controllers (e.g., component specific controllers) may be added as needed to laser source 104, first positioner 106, second positioner. 108, a third positioner 110, a lens actuator, etc. may be communicatively coupled to a driver input communicatively coupled (and associated with the component). In this embodiment, each component specific controller is communicatively coupled to the controller 114 and can generate one or more control signals (eg, trigger signals, etc.) in response to one or more control signals received from the controller 114. It may be. The one or more control signals can then be communicated to an input of a driver communicatively coupled thereto. In this embodiment, the component specific controller may be configured as described with respect to controller 114.

  In other embodiments in which one or more component specific controllers are provided, a component specific controller associated with a component (eg, laser source 104) is associated with a component (eg, first positioner 106, etc.). It can be communicatively coupled to a component specific controller. In this embodiment, one or more of the component-specific controllers are responsive to one or more control signals received from one or more other component-specific controllers, such as one or more control signals (eg, trigger signals, etc.). Can be generated.

III. Experimental Results for Removal of Workpiece Material According to certain embodiments, as described in more detail below, apparatus 100 removes a portion of workpiece 102 to provide one or more features (eg, openings, slots, vias). A laser source 104 configured to process the workpiece 102 by forming a groove, a trench, a scribe line, a kerf, a recessed region, etc., or any combination thereof. Surfaces generated as a result of processing are hereinafter referred to as “machined workpiece surfaces” and may include sidewalls, bottom surfaces, etc., or any portion or any combination thereof. In these embodiments, material is removed from the workpiece 102 by irradiating the workpiece 102 with a laser pulse having an ultrashort pulse duration at a high repetition rate.

  According to various studies, laser material processing in ultrashort pulse mode (using laser pulses with pulse durations of less than tens of ps) has many advantages over longer pulses It is shown. The thermal shock of laser interaction in picoseconds and femtoseconds is very limited, and the dissipation of laser energy can be limited to a small optical penetration depth while minimizing incidental damage. This precisely limited laser “heating” minimizes energy loss to the underlying bulk material and allows for an efficient and controllable ablation process. The ultrashort pulse duration also ensures that a large portion of the laser energy is transferred to the workpiece 102 before a significant amount of ablation plume and / or plasma is generated. Longer pulse duration laser pulses cannot achieve such efficient energy coupling due to plasma reflection, plasma and plume scattering, and plume heating. When an ultrashort laser pulse is irradiated at a high pulse repetition rate (ie, greater than 100 kHz), the heat generated by the laser pulse previously irradiated to the process spot is not completely emitted from the spot and the heat It is also generally known that at least a portion of is present in the vicinity of the spot in the workpiece 102 until the next laser pulse is irradiated. Thus, there is a tendency for heat to accumulate in the area of the workpiece 102 in the vicinity of the previously irradiated process spot, resulting in subsequent irradiation of the irradiated area with a laser pulse to be irradiated. Can do. When an ultrashort laser pulse is applied to a heated area, the temperature rise will have a positive effect on the interaction between the laser and the material, while reducing the generation of debris, and an efficient material Can help increase removal.

  However, the inventors have found that parameters such as fluence, average power, pulse energy, byte size and spot size (in addition to pulse repetition rate) and various combinations of two or more of these parameters have been processed. It has been discovered that the surface morphology of the workpiece surface can be affected, and in some cases, the production of debris during processing. The following list provides examples of new and unexpected relationships discovered during our extensive experimental work. In these experiments, the material of the workpiece 102 being processed was not “transparent” (or “non-transparent”) to the wavelength of light in the irradiated laser pulse. In this context, the material has a linear absorption spectrum and thickness within a specific bandwidth of the irradiated laser pulse, and the proportion of light transmitted through the material (along the beam axis) is less than 99%, A material is “non-transparent” if it is less than 97%, less than 95%, less than 90%, less than 75%, less than 50%, less than 25%, less than 15%, less than 10%, less than 5%, less than 1% Can be said.

A. Relationship between the byte size and the debris generated Figure 2, the trench formed in the provided surfaces of the workpiece 102 as the silicon wafer in this example (a) to (e) and (taken from the top) micrographs Show. A laser pulse is emitted while causing a relative movement between the beam axis and the workpiece 102, and a laser pulse is emitted along a process trajectory extending from left to right. The starting end of each trench is shown on the left side of the micrograph, and the ending ends of trenches (a) to (c) are shown on the right side. The appearance of the end ends of the trenches (d) and (e) was substantially the same as the appearance of the end ends of the trench (c).

  Each of the trenches (a) to (e) was formed by propagating a laser pulse having a spot size of 35 μm, a pulse duration of 800 fs, and a pulse energy of 6 μJ along the beam axis at a pulse repetition rate of 1855 kHz. Continuously irradiated laser pulses with a byte size of 0.5 μm for trench (a), byte size of 0.475 μm for trench (b), byte size of 0.5 μm for trench (c), and trench (d) Was a bite size of 0.425 μm, and for the trench (e), a relative movement was caused between the beam axis and the workpiece 102 so as to hit the workpiece 102 with a byte size of 0.4 μm. The trenches (a) to (e) were formed by scanning the irradiated laser pulse along the process trajectory with a single pass.

  As shown in FIG. 2, the parameters were selected to form a trench (a) in which debris was noticeably formed both inside and outside the trench. This resulted in a roughened workpiece surface and a roughened workpiece surface outside the processed region along the edge of the trench. When the byte size is reduced from 0.5 μm to 0.475 μm, the trench formation process causes debris along many parts of the length of the trench (b), but debris is detected at and near the end of the trench (b). I understand that it is not done. It is estimated that about 850 μs has elapsed since the start of formation of the trench (b) until there was no significant debris generation (ie, the debris transition period was about 850 μs). If the byte size is further reduced from 0.45 μm to 0.425 μm and 0.4 μm, respectively, during the formation of trenches (c), (d) and (e), the debris transition period is shortened to about 600 μs, about 320 μs and about 305 μs, respectively. It was.

  While not wishing to be bound by any particular theory, we have made work by reducing the byte size (while keeping spot size, pulse duration, pulse energy and pulse repetition rate constant). It is considered that the debris transition period is shortened because the spatial region in which the laser pulse is continuously irradiated on the piece decreases. This can cause heat to accumulate in a region within the workpiece 102 that locally surrounds the irradiated process spot. After the debris transition period has elapsed, the temperature of some of these regions (ie, the region located along the process trajectory) remains elevated (ie, between the melting and vaporization temperatures of the material being removed). ) Residual heat remaining in these regions of the workpiece 102 allows for efficient ablation of the internal material without causing noticeable debris.

B. Relationship between Pulse Energy and Debris Generation FIG. 3 shows micrographs (taken from the top) of trenches (a) to (e) formed in the surface of a workpiece 102 provided as a silicon wafer in this example. Show. A laser pulse is emitted while causing a relative movement between the beam axis and the workpiece 102, and a laser pulse is emitted along a process trajectory extending from left to right. Only the starting end of each trench is shown.

  Each of the trenches (a) to (e) was formed by propagating a laser pulse having a spot size of 35 μm and a pulse duration of 800 fs along the beam axis at a pulse repetition rate of 1979 kHz. Relative movement was caused between the beam axis and the workpiece 102 so that the continuously irradiated laser pulses hit the workpiece 102 with a byte size of 0.5 μm for each trench. The laser pulse applied to the workpiece 102 has a pulse energy of 6 μJ during the formation of the trench (a), has a pulse energy of 5 μJ for the trench (b), and is applied to the trench (c). Had a pulse energy of 4 μJ, a pulse energy of 3 μJ for the trench (d) and a pulse energy of 2 μJ for the trench (e). The trenches (a) to (e) were formed by scanning the irradiated laser pulse along the process trajectory with a single pass.

  As shown in FIG. 3, the parameters were selected so as to form a trench (a) in which debris was conspicuously formed in the vicinity of the start end outside the trench. This resulted in a roughened workpiece surface outside the machined region (less noticeable along the trench away from the start end) and the machined workpiece surface in the trench was relatively smooth. When the pulse energy is reduced from 6 μJ to 5 μJ, the trench formation process causes debris to be prominently formed in the vicinity of the start end outside the trench (debris is also prominent along the trench away from the start end), and the processed region It can be seen that the outer surface of the workpiece becomes rough. The processed workpiece surface in the trench (b) was not as smooth as the processed workpiece surface in the trench (a). When the pulse energy is further reduced to 4 μJ, the trench formation process produces noticeable debris at the start of the trench (c), roughening the workpiece surface outside the processed region, and processing in the trench (c). The finished workpiece surface is also roughened. Moreover, debris existed in the form of ridgelines of the recast material from the start end to the end end along the long sides of the trenches (a) to (c). When the pulse energy is further reduced to 3 μJ, debris is conspicuously formed in and around the start end of the trench (d) by the trench formation process, and the workpiece surface outside the processed region becomes rough, and the trench (d) It can be seen that the surface of the processed workpiece is also roughened. However, after a relatively short debris transition period, no noticeable debris formation was seen. As the distance from the start of the trench (d) increased, the ridgeline of the recast material disappeared. When the pulse energy is further decreased to 2 μJ, it can be seen that a very small amount of debris is generated outside the trench (e) at or near the start of the trench (e), and there is no noticeable debris in the trench. However, the processed workpiece surface in the trench (e) was smooth and no significant debris was seen. No edge of the recast material was seen outside the trench (e).

C. Effect of scaling on debris generation FIG. 4 shows micrographs (taken from the top) of laser machined features (a) and (b). Each laser machined feature includes a set of intersecting scribe lines formed on the surface of the workpiece 102, which in this example is provided as a silicon wafer.

  To form features (a) and (b), a laser pulse was applied to the workpiece 102 while causing a relative movement between the beam axis and the workpiece 102. Each scribe line was irradiated with a laser pulse along a process trajectory that includes three parallel scan lines, and each scan line was performed in a single pass. Each of features (a) and (b) was formed by propagating a laser pulse with a pulse duration of 800 fs along the beam axis at a pulse repetition rate of 1855 kHz. The laser pulse irradiated to the workpiece 102 during the formation of the feature (a) has a spot size of 25 μm and a pulse energy of 3.14 μJ, and the continuously irradiated laser pulse has a byte size of 0.1 μm. A relative movement was caused between the beam axis and the workpiece 102 so as to hit the piece 102. The laser pulse irradiated to the workpiece 102 during the formation of the feature (b) has a spot size of 35 μm and a pulse energy of 6.16 μJ, and the continuously irradiated laser pulse has a byte size of 0.25 μm. A relative movement was caused between the beam axis and the workpiece 102 so as to hit the piece 102.

  As is apparent from FIG. 4, a significant amount of debris was generated during the formation of feature (a), the processed workpiece surface of the scribe line became rough, and holes and other damage were visible. In contrast, substantially no debris was produced during the formation of feature (b), and the processed workpiece surface of the resulting scribe line was smooth and substantially free of debris. Note: Significant debris was generated and deposited in the area of feature (b) surrounded by a dotted circle. This area corresponds to the area of the feature machined twice.

D. Relationship between bite size, fluence and pulse repetition rate with surface roughness and material removal rate FIG. 5 shows two pulse repetition rates (ie, 927 kHz and below and 1855 kHz and below) on a workpiece 102 which is provided as a silicon wafer in this example. The average of the processed workpiece surface in the trench formed by scanning the irradiated laser pulse in a single pass along the machining trajectory while propagating the laser pulse along the beam axis in one of the Relationship between surface roughness (Ra) and material removal rate (area um2) during trench formation process, byte size (measured in μm) and fluence (measured in J / cm ² ) A set of graphs shown as a function of The average surface roughness (Ra) was measured using a Keyence 3D confocal scanning microscope having a 50 × objective lens.

  As can be seen from FIG. 5, at bite sizes greater than 0.2 μm, the average surface roughness of the processed workpiece surface drops below about 0.25 μm, approaching a mirror-like smooth surface finish. For all bite sizes and fluence levels tested, the average surface roughness of the processed workpiece surface obtained using a laser pulse irradiated at a pulse repetition rate of 1855 kHz or less is irradiated at a pulse repetition rate of 927 kHz or less. Generally less than the average surface roughness of the corresponding machined workpiece surface obtained using a laser pulse. The um2 area value represents the cross-sectional area of the scribe, and indicates that the material removal rate decreases as the bite size increases. The material removal rate during trench formation at a pulse repetition rate of 927 kHz or less is similar to the material removal rate during trench formation at a pulse repetition rate of 1855 kHz or less.

E. Relationship Between Byte Size, Fluence, Pulse Repeat Rate, and Average Power Debris Generation FIG. 6 shows a set of graphs showing process windows for forming trenches in a workpiece 102, which in this example is provided as a silicon wafer. ing. i) a processed workpiece surface without noticeable debris formation (ie, a processed workpiece surface without noticeable debris accumulation as described with respect to FIGS. 2-5) is formed, and ii) the debris is noticeable Produced workpiece surface (i.e., a workpiece surface with significant debris deposition as described with respect to FIGS. 2-5). The region marked with the pattern indicated by reference number 600 represents a parameter space where debris is generated conspicuously, and the region marked with the pattern indicated by reference number 602 is a parameter space where no conspicuous debris is formed. Represents. The observed trench processes the irradiated laser pulse while propagating the laser pulse along the beam axis at one of five pulse repetition rates (ie, 927.55kHz, 1264kHz, 1855kHz, 2022kHz and 3051kHz). Formed by scanning in a single pass along the trajectory. A plurality of trenches were formed at each pulse repetition rate. Each trench is formed using a different combination of byte size (measured in μm), fluence (measured in J / cm ² units) and average power (measured in W). It was.

  As shown in FIG. 6, at 927.55 kHz and 1264 kHz, it was found that all combinations of byte size, fluence and average power values tested produced a substantial amount of debris from moderate amounts. On the other hand, at 1855 kHz, 2022 kHz and 3051 kHz, it was found that processed workpiece surfaces with little or no debris accumulation were obtained with some (but not all) parameter value combinations. This finding tends to indicate that there is a pulse repetition rate threshold below which debris generation is unavoidable for the specific material being processed. However, other general observations can be obtained above the pulse repetition rate threshold. That is, at a relatively low fluence or average power value, the workpiece 102 can be machined using a relatively wide range of bite sizes to form features without producing moderate or substantial amounts of debris. it can. Also, as the fluence or average power increases, this byte size range becomes narrower.

  There is a parameter space in which the region 600 and the region 602 overlap. For example, see the region marked by the pattern indicated by reference numeral 604. This overlap can be (1) there is a transition between the generation of substantial or noticeable debris and the generation of minor or inconspicuous debris, or (2) for a given fluence, power and byte size. It can be generally understood that this indicates that there are processes that can produce clean or smooth features or features with debris generation. As an example, at 1855 kHz, there is a combination of pot size and pulse energy that produces different results (ie, features with clean or smooth features or debris generation) for the process of matching power and fluence. In other words, at a given coordinate in the parameter space where region 600 and region 602 overlap, a feature with clean or smooth surface or debris generation depending on the spot size and pulse energy of the irradiated laser pulse. Can be formed.

IV. Example Embodiments Based on Experimental Results Based on the experimental results described in Section III, Sections A through E above, one embodiment of the present invention can be used during the removal process (irradiating the impervious workpiece 102). It can be characterized as a laser process for forming features (eg, scribe or other trenches or recesses) in the workpiece 102 by removing material (relative to the wavelength of the laser pulse light). For example, referring to the embodiment shown in FIG. 9A, the workpiece 102 may be a semiconductor wafer having an upper surface (eg, surface 900a) and a lower surface opposite the upper surface (eg, surface 900b). The semiconductor wafer is formed from a material such as silicon, germanium, Si _1-x Ge _x (0.0001 <x <0.9999), GaAs, GaN, InP, etc., or any combination thereof. And a device layer 904 (eg, formed from a field effect transistor, dielectric layer, interconnect metallization structure, passivation layer, etc., or any combination thereof). It should be appreciated that the workpiece 102 may be provided by any method other than the semiconductor wafer described above. For example, the workpiece 102 may be Al ₂ O ₃ , AlN, BeO, Cu, GaAs, GaN, Ge, InP, Si, SiO ₂ , SiC, Si _1-x Ge _x (0.0001 <x <0.9999), or the like. Any single or multi-layer structure, plastic, glass (eg, unreinforced glass) including substrates (eg, electronic substrates, semiconductor substrates, optical substrates, etc.) formed of any combination of these or any alloy thereof Or heat tempered glass or chemically tempered glass or other glass), quartz, sapphire, plastic, silicon, one or more metals (eg, Al, Ag, Au, Cu, Fe, In, Mg, Pt, Sn, Ti, or any combination thereof or any alloy thereof), conductive metal oxide (eg, ITO), transparent conductive polymer, ceramic, wax, resin (eg, silicon oxide, silicon nitride) , Silicon oxynitride, etc. Or inorganic dielectric materials (used as interlayer dielectric structures like any combination of these), low-k dielectric materials (eg, methyl silsesquioxane (MSQ), hydrogen silsesquioxane (HSQ), Tetraethyl fluoroorthosilicate (FTEOS) or any combination thereof, organic dielectric materials (eg SILK, benzocyclobutene, Nautilus (all manufactured by Dow), polyfluorotetraethylene ( Manufactured by DuPont), FLARE (manufactured by Allied Chemical), or any combination thereof, fiber glass, polymer materials (polyamide, polyimide, polyester, polyacetal, polycarbonate, modified polyphenylene) Ether, polybutylene terephthalate, polyphenylenesulfur , Polyethersulfone, polyetherimide, polyetheretherketone, liquid crystal polymer, acrylonitrile butadiene styrene, and any compounds, composites, or mixtures thereof), or any combination thereof. Can be provided.

  Removal process parameters (eg, fluence, average power, pulse repetition rate, pulse energy, spot size, etc.) to ensure removal of a portion of the workpiece 102 in a manner that advantageously achieves one or more of the following: One or more of the byte sizes, etc.) is selected, controlled, or set. Minimizing or eliminating the generation of debris during processing, generating a smooth processed workpiece surface, generating a processed workpiece surface with fewer defects, scratches or cracks, processed workpiece Produce a uniform HAZ in the workpiece 102 adjacent to the surface. For example, during the removal process, a laser pulse beam may be irradiated along a beam axis that intersects the workpiece 102, and a continuously irradiated laser pulse strikes the workpiece 102 in a non-zero byte size, and the workpiece 102 The laser pulse beam may be scanned to form a feature (eg, feature 906 as shown in FIG. 9B, which may be a recess, a trench, etc.) on the top surface 900a of the substrate.

  In the embodiment shown in FIG. 9B, feature 906 extends through device layer 904 and extends to a portion of substrate 902 (eg, from the top surface of substrate 902 to a measured depth d). In some embodiments, the depth d can range from 5 μm (or before and after) to 22 μm (or around). For example, the depth d is 5 μm, 5.5 μm, 6.0 μm, 6.5 μm, 7.0 μm, 7.5 μm, 8.0 μm, 8.5 μm, 10 μm, 12 μm, 15 μm, 17 μm, 20 μm, 22 μm, or any of these values Can be between. However, it should be appreciated that the depth d may be less than 5 μm or greater than 22 μm. In other embodiments, feature 906

  If the parameters are selected to form a sufficiently smooth machined workpiece surface (eg, machined workpiece surface 906), such as internal machining of workpiece 102, workpiece penetration of workpiece 102, etc., or The processed workpiece surface may be used to facilitate subsequent processes such as any combination thereof. If the processed workpiece surface has an average surface roughness (Ra) of 1.0 μm or less, the processed workpiece surface (eg, processed workpiece surface 906) is “sufficient to facilitate subsequent processes. It can be thought of as “smooth”. In some embodiments, the processed workpiece surface is less than 1.0 μm, less than 0.75 μm, less than 0.5 μm, less than 0.4 μm, less than 0.3 μm, less than 0.25 μm, less than 0.2 μm, less than 0.15 μm, etc., or these values Can have an average surface roughness (Ra) between

  Internal machining of the workpiece 102 can be performed by irradiating another laser pulse beam first through the machined workpiece surface and then into the workpiece. In this case, the laser pulse beam is focused so that the beam waist of the laser pulse is located inside the workpiece 102. The laser pulses used during internal machining of the workpiece 102 have a wavelength that is more transparent to the material within the workpiece 102 being machined than the wavelength used during the initial formation of the machined workpiece surface. Parameters associated with such internal processing (eg, one or more of fluence, average power, pulse repetition rate, pulse energy, spot size, bite size, etc.) are determined by the laser pulses emitted by the material in the workpiece 102. Induces non-linear absorption, thereby machining (eg, melting, or processing) a portion of material within the workpiece 102 at or near the beam waist of the irradiated laser pulse (eg, portion 908 shown in FIG. 9C) Evaporate, ablate, crack, decolorize, etc. or chemical composition, crystal structure, electronic structure, microstructure, nanostructure, concentration, viscosity, refractive index, magnetic permeability, relative dielectric constant, etc. To change one or more attributes or characteristics). For example, after a trench has been formed in a workpiece 102, such as a silicon wafer, to form a sufficiently smooth processed workpiece surface (eg, as shown in the micrograph of FIG. 7), as described above. (Eg, as shown in the micrographs of FIGS. 8A and 8B (FIG. 8A is a view across the width of the trench shown in FIG. 7 and FIG. 8B is shown in FIG. 7). It is a view along the length of the trench))) A series of cracks may be formed inside the silicon wafer.

  Referring to FIG. 9D, the workpiece through-penetration of workpiece 102 causes another laser pulse beam to first pass through the processed workpiece surface (eg, processed workpiece surface 906a) and then into the workpiece. Can be carried out by irradiating the film. The irradiated laser pulse beam is focused so that the beam waist of the laser pulse is located at or near the lower surface 900b of the workpiece 102. The laser pulses used during internal machining of the workpiece 102 have a wavelength that is more transparent to the material within the workpiece 102 being machined than the wavelength used during the initial formation of the machined workpiece surface. Parameters associated with such workpiece penetration (eg, one or more of fluence, average power, pulse repetition rate, pulse energy, spot size, bite size, etc.) depend on the material of workpiece 102 at lower surface 900b. The material of the workpiece 102 in or near the beam waist of the irradiated laser pulse (eg, so as to form a trench or recess 910 on the lower surface 900b) that induces linear or nonlinear absorption of the irradiated laser pulse. Is selected to machine part of. An example of possible penetrating workpiece machining of workpiece 102 is described in US Pat. No. 9,610,653, which is incorporated herein by reference in its entirety.

V. CONCLUSION The foregoing describes embodiments and examples of the present invention and should not be construed as limiting. Although several specific embodiments and examples have been described with reference to the drawings, those skilled in the art will recognize that the disclosed embodiments and examples and other implementations may be practiced without departing from the novel teachings and advantages of the present invention. It will be readily appreciated that many improvements to the form are possible.

For example, the experiment described in Chapter III above was performed on a bare silicon wafer. If the material to be processed is opaque to the wavelength of the irradiation laser pulse, an ultrashort laser pulse is used ( It should be recognized that similar effects can be seen when processing workpieces containing materials (other than silicon wafers). Thus, a semiconductor wafer, an electronic device substrate, or an optical device substrate (for example, Al ₂ O ₃ , AlN, BeO, Cu, GaAs, GaN, Ge, InP, Si, SiO ₂ , SiC made of a material other than silicon) Si _1-x Ge _x (0.0001 <x <0.9999), etc., or any combination thereof, or a substrate formed from any alloy thereof, plastic, glass (non-tempered glass or heat tempered glass or chemically tempered glass) Or any other glass), quartz, sapphire, plastic, silicon, etc., one or more metals (eg, Al, Ag, Au, Cu, Fe, In, Mg, Pt, Sn, Ti, etc., or any of these) In combination or any alloy thereof), conductive metal oxides (for example, ITO), transparent conductive polymers, ceramics, waxes, resins (for example, silicon oxide, silicon nitride, silicon oxynitride, etc.) Is used as an interlayer dielectric structure, such as any combination of these), low-k dielectric materials (eg, methyl silsesquioxane (MSQ), hydrogen silsesquioxane (HSQ), fluorine Tetraethyl orthosilicate (FTEOS) or any combination thereof, organic dielectric materials (eg SILK, benzocyclobutene, Nautilus (all manufactured by Dow), polyfluorotetraethylene (DuPont) Manufactured by the company), FLARE (manufactured by Allied Chemical), or any combination thereof, fiber glass, polymer materials (polyamide, polyimide, polyester, polyacetal, polycarbonate, modified polyphenylene ether) , Polybutylene terephthalate, polyphenylene sulfide Polyethersulfone, polyetherimide, polyetheretherketone, liquid crystal polymer, acrylonitrile butadiene styrene, and any compounds, composites, or mixtures thereof), or any combination thereof. It should be appreciated that the above-described embodiments can be advantageously adapted to process.

  Accordingly, all such modifications are intended to be included within the scope of this invention as defined in the claims. For example, those skilled in the art will recognize the subject matter of any sentence or paragraph, example or embodiment as part or all of another sentence or paragraph, example or embodiment, unless such combinations are mutually exclusive. It will be understood that it can be combined with the subject matter. Accordingly, the scope of the invention should be determined by the following claims and the equivalents of the claims to be included therein.

Claims (15)

Providing a workpiece having a first surface and a second surface opposite to the first surface;
Generating a first laser pulse beam having a pulse duration of less than 200 ps with a pulse repetition rate higher than 500 kHz, a spot size, and a pulse energy;
Irradiating the first laser pulse beam along a beam axis intersecting the workpiece;
Scanning the beam axis along a machining trajectory so that a continuously irradiated laser pulse strikes the workpiece with a non-zero byte size to form a feature on the first surface of the workpiece; Is characterized as having a processed workpiece surface with an average surface roughness (Ra) of less than 1.0 μm,
Method.   The method of claim 1, wherein the pulse duration is 1 ps or less.   The method according to claim 1, wherein the pulse duration is 800 fs or less.   The method according to claim 1, wherein the pulse repetition rate is higher than 1264 kHz.   The method according to claim 1, wherein the pulse repetition rate is 1800 kHz or more.   The method according to claim 1, wherein the pulse repetition rate is 1900 kHz or more.   The method according to claim 1, wherein the pulse repetition rate is 2000 kHz or more.   The method according to claim 1, wherein the pulse repetition rate is 3000 kHz or more.   The method according to claim 1, wherein the average surface roughness (Ra) is less than 0.75 μm.   The method according to claim 1, wherein the average surface roughness (Ra) is less than 0.5 μm.   The method according to claim 1, wherein the average surface roughness (Ra) is less than 0.4 μm.   The method according to claim 1, wherein the average surface roughness (Ra) is less than 0.3 μm.   The method according to claim 1, wherein the average surface roughness (Ra) is less than 0.25 μm. further,
Generating a second laser pulse beam;
Focusing a laser pulse in the second laser pulse beam to generate a beam waist;
The focused second laser pulse beam is along a beam axis intersecting the processed workpiece surface such that the beam waist is located in the workpiece or on the second surface of the workpiece. Irradiate
Machining the workpiece with the beam waist;
The method according to claim 1.   15. The workpiece of claim 14, wherein the workpiece is more transparent to the wavelength of the laser pulse in the second laser pulse beam than to the wavelength of the laser pulse in the first laser pulse beam. Method.