U.S. patent application number 15/617622 was filed with the patent office on 2017-09-21 for transparent material cutting with ultrafast laser and beam optics.
The applicant listed for this patent is CORNING INCORPORATED. Invention is credited to Sasha Marjanovic, Garrett Andrew Piech, Sergio Tsuda, Robert Stephen Wagner.
Application Number | 20170266757 15/617622 |
Document ID | / |
Family ID | 53367591 |
Filed Date | 2017-09-21 |
United States Patent
Application |
20170266757 |
Kind Code |
A1 |
Marjanovic; Sasha ; et
al. |
September 21, 2017 |
TRANSPARENT MATERIAL CUTTING WITH ULTRAFAST LASER AND BEAM
OPTICS
Abstract
A system for laser drilling of a material includes a pulsed
laser configured to produce a pulsed laser beam having a wavelength
less than or equal to about 850 nm, the wavelength selected such
that the material is substantially transparent at this wavelength.
The system further includes an optical assembly positioned in the
beam path of the laser, configured to transform the laser beam into
a laser beam focal line oriented along the beam propagation
direction, on a beam emergence side of the optical assembly.
Inventors: |
Marjanovic; Sasha; (Painted
Post, NY) ; Piech; Garrett Andrew; (Corning, NY)
; Tsuda; Sergio; (Horseheads, NY) ; Wagner; Robert
Stephen; (Corning, NY) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
CORNING INCORPORATED |
CORNING |
NY |
US |
|
|
Family ID: |
53367591 |
Appl. No.: |
15/617622 |
Filed: |
June 8, 2017 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
14529801 |
Oct 31, 2014 |
9687936 |
|
|
15617622 |
|
|
|
|
62022888 |
Jul 10, 2014 |
|
|
|
61917140 |
Dec 17, 2013 |
|
|
|
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
B23K 26/0624 20151001;
Y02P 40/57 20151101; B23K 26/382 20151001; B23K 26/53 20151001;
C03B 33/0222 20130101 |
International
Class: |
B23K 26/00 20060101
B23K026/00; B23K 26/382 20060101 B23K026/382; C03B 33/02 20060101
C03B033/02; B23K 26/0622 20060101 B23K026/0622 |
Claims
1. A system for laser drilling of a material, the system
comprising: a pulsed laser configured to produce a pulsed laser
beam having a wavelength less than or equal to about 850 nm; and an
optical assembly positioned in the beam path of the laser,
configured to transform the laser beam into a laser beam focal line
oriented along a beam propagation direction, on the beam emergence
side of the optical assembly, the optical assembly including a
focusing optical element with spherical aberration configured to
generate the laser beam focal line, said laser beam focal line
adapted to generate an induced absorption within the material, the
induced absorption producing a defect line having a diameter less
than or equal to about 300 nm along the laser beam focal line
within the material.
2. The system of claim 1, wherein the laser beam has a wavelength
less than or equal to about 775 nm.
3. The system of claim 2, wherein the laser beam has a wavelength
less than or equal to about 600 nm.
4. The system of claim 3, wherein the laser beam has a wavelength
less than or equal to about 532 nm.
5. The system of claim 1, wherein the induced absorption produces
subsurface damage up to a depth less than or equal to about 75
.mu.m within the material.
6. The system of claim 1, wherein the induced absorption produces
an Ra surface roughness less than or equal to about 0.5 .mu.m.
7. The system of claim 1, wherein the optical assembly includes an
annular aperture positioned in the beam path of the laser before
the focusing optical element, the annular aperture configured to
block out one or more rays in the center of the laser beam so that
only marginal rays outside the center are incident on the focusing
optical element, and thereby only a single laser beam focal line,
viewed along the beam direction, is produced for each pulse of the
pulsed laser beam.
8. The system of claim 1, wherein the focusing optical element is a
spherically cut convex lens.
9. The system of claim 1, wherein the focusing optical element is a
conical prism having a non-spherical free surface.
10. The system of claim 9, wherein the conical prism is an
axicon.
11. The system of claim 1, wherein the optical assembly further
includes a defocusing optical element, the optical elements
positioned and aligned such that the laser beam focal line is
generated on the beam emergence side of the defocusing optical
element at a distance from the defocusing optical element.
12. The system of claim 1, wherein the optical assembly further
includes a second focusing optical element, the two focusing
optical elements positioned and aligned such that the laser beam
focal line is generated on the beam emergence side of the second
focusing optical element at a distance from the second focusing
optical element.
13. The system of claim 1, wherein the pulsed laser is configured
to emit pulses produced in bursts of at least two pulses separated
by a duration in a range of between about 1 nsec and about 50 nsec,
and the burst repetition frequency is in a range of between about 1
kHz and about 2 MHz.
14. The system of claim 13, wherein the pulses are separated by a
duration of about 20 nsec.
15. The system of claim 1, wherein the laser beam focal line has a
length in a range of between 0.1 mm and 100 mm.
16. The system of claim 1, wherein the laser beam focal line has an
average spot diameter in a range of between 0.1 .mu.m and 5 .mu.m.
Description
CROSS-REFERENCE TO RELATED APPLICATIONS
[0001] This application is a divisional of U.S. application Ser.
No. 14/529,801 filed on Oct. 31, 2014 entitled "Transparent
Material Cutting With Ultrafast Laser And Beam Optics" which also
claims the benefit of U.S. Provisional Application No. 62/022,888
filed on Jul. 10, 2014 and U.S. Provisional Application No.
61/917,140 Dec. 17, 2013. The entire teachings of these
applications are incorporated herein by reference.
BACKGROUND
[0002] In recent years, precision micromachining and its
improvement of process development to meet customer demand to
reduce the size, weight and material cost of leading-edge devices
has led to fast pace development in high-tech industries in flat
panel displays for touch screens, tablets, smartphones and TVs,
leading to ultrafast industrial lasers becoming important tools for
applications requiring high precision.
[0003] There are various known ways to cut glass. In a conventional
laser glass cutting process, the separation of glass relies on
laser scribing or perforation with separation by mechanical force
or thermal stress induced crack propagation. Nearly all current
laser cutting techniques exhibit one or more shortcomings: (1) they
are limited in their ability to perform a free form shaped cut of
thin glass on a carrier due to a large heat-affected zone (HAZ)
associated with long pulse lasers (nanosecond scale or longer), (2)
they produce a thermal stress that often results in cracking of the
surface near the laser illuminated region due to a shock wave and
uncontrolled material removal and, (3) the process creates
sub-surface damage extending tens of microns (or more) into the
body of the material resulting in defect sites which can become
crack sources.
[0004] Therefore, there is a need for an improved process of laser
drilling a material, such as glass that minimizes or eliminates one
or more of the above mentioned problems, that minimizes or
eliminates the above mentioned problems.
SUMMARY
[0005] The following embodiments relate to a method and an
apparatus to create small (micron and smaller) "holes" or defect
lines in transparent materials (e.g., glass, sapphire, etc.) for
the purpose of drilling and cutting.
[0006] More particularly, according to some embodiments, a pulsed
laser beam having a wavelength less than 1000 nm is focused into a
laser beam focal line, and the focal line is directed into the
material, where the laser beam focal line generates an induced
absorption within the material, the induced absorption producing a
defect line having a diameter less than or equal to about 300 nm
along the laser beam focal line within the material. For example,
an ultra-short (e.g., 10.sup.-10 to 10.sup.-15 second) pulse beam
(wavelength less than 1000 nanometers (nm) having a Gaussian
profile, is shaped and focused to create a linear focal region in
the body of the material. The resulting energy density is above the
threshold for material modification, creating a "defect line" or
"hole" in that region. By spacing these features close together,
the material may be separated (mechanically or thermally) along the
perforation line. In some embodiments the pulsed laser beam's
wavelength is less than or equal to 850 nm, in some embodiments
less than or equal to 800 nm, in some embodiments less than 620 nm,
and in some embodiments not greater than 552 nm.
[0007] For example, according to some embodiments, a pulsed laser
beam having a wavelength less than or equal to about 800 nm (.+-.50
nm, preferably .+-.20 nm, more preferably .+-.2 nm) such as the
laser beam produced by Ti: sapphire laser, less than or equal to
about 775 nm (frequency doubled Er-doped fiber laser), less than or
equal to about 600 nm (rhodamine based dye laser) and in some
embodiments less than or equal to about 532 nm (e.g., 532 nm.+-.20
nm, more preferably .+-.2 nm), is focused into a laser beam focal
line, and the focal line is directed into the material, where the
laser beam focal line generates an induced absorption within the
material, the induced absorption producing a defect line having a
diameter less than or equal to about 300 nm along the laser beam
focal line within the material. For example, an ultra-short (e.g.,
10.sup.-10 to 10.sup.-15 second) pulse beam (less than or equal to
about 800 nm, 775 nm, 600 nm, 532 nm 355 nm, or 266 nm) having a
Gaussian profile, is shaped and focused to create a linear focal
region in the body of the material. The resulting energy density is
above the threshold for material modification, creating a "defect
line" or "hole" in that region. By spacing these features close
together, the material may be separated (mechanically or thermally)
along the perforation line.
[0008] In one embodiment, a method of laser drilling a material
includes focusing a pulsed laser beam into a laser beam focal line
oriented along the beam propagation direction, the laser beam
having a wavelength less than or equal to about 850 nm, the
wavelength selected such that the material is substantially
transparent at this wavelength. The method also includes directing
the laser beam focal line into the material, the laser beam focal
line generating an induced absorption within the material, the
induced absorption producing a defect line having a diameter less
than or equal to about 300 nm along the laser beam focal line
within the material.
[0009] The induced absorption can produce subsurface damage up to a
depth less than or equal to about 100 .mu.m (for example less than
75 .mu.m) within the material, and an Ra surface roughness less
than or equal to about 0.5 .mu.m. A roughness of a surface can be
characterized, for example, by an Ra surface roughness statistic
(roughness arithmetic average of absolute values of the heights of
the sampled surface).
[0010] In some embodiments, the method further includes translating
the material and the laser beam relative to each other, thereby
drilling a plurality of defect lines within the material, the
defect lines spaced apart so as to separate the material into at
least two pieces. In certain embodiments, the laser is a pulse
burst laser and the repetition rate of the laser bursts (i.e.,
burst repetition rate) can be in a range of between about 10 kHz
and 2000 kHz such as 100 kHz, 200 kHz, 300 kHz, 400 kHz, 500 kHz,
1000 kHz, or 1500 kHz. In some embodiments, the laser beam has a
wavelength less than or equal to about 775 nm, less than or equal
to about 600 nm, or less than or equal to about 532 nm. In some
embodiments, the pulse duration of the individual pulses within a
pulse burst of the laser can be in a range of between about 5
picoseconds and about 100 picoseconds, for example 10, 20, 30, 40,
50, 60, 75, 80, 90 or 100 picoseconds, or therebetween.
[0011] The pulsed laser can be configured to emit pulses produced
in bursts of at least two pulses per burst, the adjacent pulses
being separated by a duration in a range of between about 1 nsec
and about 50 nsec (more preferably in the range of 15 to 30 nsec)
with the burst repetition frequency being in a range of between
about 1 kHz and about 500 kHz (and preferably 200 kHz). In some
embodiments, the individual pulses within a pulse burst can be
separated by a duration of about 20 nsec.
[0012] In certain embodiments, the laser beam focal line can have a
length L in a range of between about 0.1 mm and about 20 mm, in
some embodiments between 10 mm and 20 mm, for example a length in a
range of between about 0.1 mm and about 8 mm. The laser beam focal
line can have an average spot diameter in a range of between about
0.1 .mu.m and about 5 .mu.m.
[0013] In another embodiment, a system for laser drilling of a
material includes a pulsed laser configured to produce a pulsed
laser beam having a wavelength less than or equal to about 850 nm,
the wavelength selected such that the material is substantially
transparent at this wavelength. The system further includes an
optical assembly positioned in the beam path of the laser,
configured to transform the laser beam into a laser beam focal
line, oriented along the beam propagation direction, on the beam
emergence side of the optical assembly, the optical assembly
including a focusing optical element with spherical aberration
configured to generate the laser beam focal line. The laser beam
focal line can be adapted to generate an induced absorption within
the material, the induced absorption producing a defect line having
a diameter less than or equal to about 300 nm along the laser beam
focal line within the material.
[0014] In some embodiments, the laser beam has a wavelength less
than or equal to about 775 nm, less than or equal to about 600 nm,
or less than or equal to about 532 nm. The induced absorption can
produce subsurface damage up to a depth less than or equal to about
75 .mu.m within the material, such as less than or equal to about
40 .mu.m, an Ra surface roughness less than or equal to about 0.8
.mu.m, and a RMS surface roughness less than or equal to about 0.9
.mu.m. The optical assembly can include an annular aperture
positioned in the beam path of the laser before the focusing
optical element, the annular aperture configured to block out one
or more rays in the center of the laser beam so that only marginal
rays outside the center are incident upon on the focusing optical
element, and thereby only a single laser beam focal line, oriented
along the beam direction, is produced for each pulse of the pulsed
laser beam. The focusing optical element can be a spherically cut
convex lens. Alternatively, the focusing optical element can be a
conical prism having a non-spherical free surface, such as an
axicon.
[0015] In some embodiments, the optical assembly can further
include a defocusing optical element, the optical elements
positioned and aligned such that the laser beam focal line is
generated on the beam emergence side of the defocusing optical
element at a distance from the defocusing optical element.
Alternatively, the optical assembly can further include a second
focusing optical element, the two focusing optical elements
positioned and aligned such that the laser beam focal line is
generated on the beam emergence side of the second focusing optical
element at a distance from the second focusing optical element. The
pulsed laser can be configured to emit pulses produced in bursts of
at least two pulses (such as, for example, at least 3 pulses, at
least 4 pulses, at least 5 pulses, at least 10 pulses, at least 15
pulses, at least 20 pulses, or more). The pulses inside the burst
are separated by a duration in a range of between about 1 nsec and
about 50 nsec, for example 10 to 30 nsec, such as about 20 nsec,
and the burst repetition frequency can be in a range of between
about 1 kHz and about 2 MHz, such as a burst repetition frequency
of about 100 kHz, about 200 kHz, about 300 kHz, about 400 kHz,
about 500 kHz, about 1 MHz, or about 1.5 MHz. (Bursting or
producing pulse bursts is a type of laser operation where the
emission of pulses is not in a uniform and steady stream but rather
in tight clusters of pulses.) The glass is moved relative to the
laser beam (or the laser beam is translated relative to the glass)
to create perforated lines that trace out the shape of any desired
parts. The pulse burst laser beam can have a wavelength selected
such that the material is substantially transparent at this
wavelength. With a focal line about 1 mm in length, and a 532 nm
picosecond laser that produces output power of about 2 W or more at
a burst repetition rate of 20 kHz (about 100 microJoules/burst)
measured at the glass composition, the optical intensities in the
focal line region are high enough to create non-linear absorption
in the glass composition The average laser power per burst measured
at the material can be greater than 40 microJoules per mm thickness
of material, for example between 40 microJoules/mm and 2500
microJoules/mm, or between 500 and 2250 microJoules/mm. For
example, for 0.4 mm thick code 2320 glass (available from Corning
Incorporated, Corning, N.Y.) one may use 100 .mu.J pulse bursts to
cut and separate the glass, which gives an exemplary range of 250
.mu.J/mm. The laser beam focal line can have a length in a range of
between 0.1 mm and 20 mm, and an average spot diameter in a range
of between 0.1 .mu.m and 5 .mu.m.
[0016] In another embodiment, a method of laser drilling a material
includes focusing a pulsed laser beam into the laser beam focal
line oriented along the beam propagation direction, the laser beam
having a wavelength less than 850 nm. The method also includes
directing the laser beam focal line into the material, the laser
beam focal line generating an induced absorption within the
material, the induced absorption producing a defect line having an
internal diameter less than 0.5 .mu.m along the laser beam focal
line within the material. In some embodiments, producing the defect
line includes producing the defect line with internal diameter of
less than 0.4 .mu.m. In some embodiments, producing the defect line
further includes producing the defect line with internal diameter
of less than 0.3 .mu.m or 0.2 .mu.m.
[0017] These embodiments have many advantages, such as less
subsurface damage due to the laser wavelength being less than or
equal to about 850 nm, and preferably less or equal to about 532
nm, as compared to prior art laser drilling methods, producing less
surface debris, less adhered debris, and less thermal interaction.
While laser ablation cutting of thin glasses exhibits slow
processing speed due to low output power and pulse energy, it has
advantages which include no crack creation near the ablation
region, free form shaping, and controllable cutting thickness by
adjusting a focal length. It is important for flat panel displays
that edge cracking and residual edge stress are avoided in glass
substrates, because such substrates almost always break from the
edge, even when stress is applied to the center. The high peak
power of ultrafast lasers combined with tailored beam delivery can
avoid these problems by using cold ablation cutting without
measurable heat effect. Laser cutting by ultrafast lasers produces
essentially no residual stress in the glass.
BRIEF DESCRIPTION OF THE DRAWINGS
[0018] The foregoing will be apparent from the following more
particular description of example embodiments, as illustrated in
the accompanying drawings in which like reference characters refer
to the same parts throughout the different views. The drawings are
not necessarily to scale, emphasis instead being placed upon
illustrating embodiments.
[0019] FIG. 1 is a graph of ionization energy as a function of
atomic number for several atoms.
[0020] FIGS. 2A and 2B are illustrations of positioning of the
laser beam focal line, i.e., the processing of a material
transparent for the laser wavelength.
[0021] FIG. 3A is an illustration of an optical assembly for laser
drilling according to one embodiment.
[0022] FIG. 3B1-3B4 are illustrations of various possibilities to
process the substrate by differently positioning the laser beam
focal line relative to the substrate.
[0023] FIG. 4 is an illustration of a second optical assembly for
laser drilling according to some embodiments.
[0024] FIGS. 5A and 5B are illustrations of a third optical
assembly for laser drilling according to some embodiments.
[0025] FIG. 6 is a schematic illustration of a fourth optical
assembly for laser drilling according to some embodiments.
[0026] FIG. 7A is a graph of laser emission as a function of time
for a picosecond laser. Each emission is characterized by a pulse
"burst" which may contain one or more pulses.
[0027] FIG. 7B illustrates schematically the relative intensity of
laser pulses within an exemplary pulse burst vs. time, with each
exemplary pulse burst having 3 pulses;
[0028] FIG. 7C illustrates schematically relative intensity of
laser pulses vs. time within an exemplary pulse burst, with each
exemplary pulse burst containing 5 pulses.
[0029] FIG. 8 is a photograph of Corning 2320 Gorilla.RTM. glass
samples on a vacuum fixture.
[0030] FIG. 9 is a graph of failure probability as a function of
stress for Corning 2320 Gorilla.RTM. glass showing results of
stress tests with laser-in-compression (LIC) and laser-in-tension
(LIT).
[0031] FIG. 10 shows several photographs of various cut surfaces at
20.times. magnification.
[0032] FIGS. 11A-11D are screen shots of representative Zygo scans
(total of 5 scans collected for each condition). FIG. 11B shows a
screen shot of Zygo scans of ST2 strength part condition.
[0033] FIGS. 12A-12C are SEM micrographs comparing cross sectional
views of edges obtained with the 532 nm process with a cross
sectional view of a reference edge obtained with the 1064 nm
process (FIG. 12B).
[0034] FIGS. 13A-13B are SEM micrographs comparing a reference edge
made with the 1064 nm process (FIG. 13A) with an edge made with the
532 nm process (FIG. 13B).
[0035] FIGS. 14A-14B are graphs of failure probability as a
function of stress for Corning 2320 Gorilla.RTM. glass showing
results of stress tests with laser-in-compression (LIC) and
laser-in-tension (LIT) for edges separated with the 1064 nm process
(FIG. 14A) and for edges separated with the 532 nm process (FIG.
14B).
DETAILED DESCRIPTION
[0036] A description of example embodiments follows.
[0037] Disclosed herein is a method or process, and apparatus for
optically producing high precision through-cuts in transparent
materials with low sub-surface damage and low debris. In addition,
by judicious selection of optics, it is possible to selectively cut
individual layers of stacked transparent materials.
[0038] Micromachining and cutting thin glasses with minimal
sub-surface damage and surface debris is accomplished by selection
of an appropriate laser source and wavelength along with beam
delivery optics. The laser source consists of an ultrafast laser
system providing pulses of sub-nanosecond duration along with a
beam delivery that illuminates a "linear" focus region within the
body of the transparent material. The energy density along the
"linear" focus region needs to be greater than the energy necessary
to separate the material in that zone. This necessitates the use of
high energy pulsed laser sources.
[0039] In addition, the selection of wavelength is important.
Materials with stronger molecular bonds will exhibit "better"
separation using shorter wavelengths (i.e., less than 1000 nm, such
as 850 nm, 820 nm, 800 nm, 775 nm, 600 nm, 532 nm, 355 nm, or 266
nm). Also, the shorter wavelengths focus tighter, resulting in
higher volumetric energy densities in the focal region.
[0040] Thus, it is possible to create a microscopic (i.e., <0.5
.mu.m and >100 nm in diameter) elongated "hole (also called a
perforation or a defect line) in a transparent material using a
single high energy burst pulse. These individual perforations can
be created at rates of several hundred kilohertz (several hundred
thousand perforations per second, for example). Thus, with relative
motion between the source and the material, these perforations can
be placed adjacent to one another (spatial separation varying from
sub-micron to several microns as desired). This spatial separation
(pitch) is selected in order to facilitate cutting.
[0041] In some embodiments the defect line is a "through hole",
which is a hole or an open channel that extends from the top to the
bottom of the transparent material. In some embodiments the defect
line may not be a continuous channel, and may be blocked or
partially blocked by portions or sections of solid material (e.g.,
glass). As defined herein, the internal diameter of the defect line
is the internal diameter of the open channel or the air hole. For
example, in embodiments described herein the internal diameter of
the defect line is <500 nm, for example <400 nm, or <300
nm, or .ltoreq.200nm. The disrupted or modified area (e.g.,
compacted, melted, or otherwise changed) of the material
surrounding the holes in the embodiments disclosed herein,
preferably has diameter of <50 .mu.m (e.g., <0.10 .mu.m).
[0042] The selection of the laser source is predicated on the
ability to create multi-photon absorption (MPA) in transparent
materials. MPA is the simultaneous absorption of two or more
photons of identical or different frequencies in order to excite a
molecule from one state (usually the ground state) to a higher
energy electronic state (ionization). The energy difference between
the involved lower and upper states of the molecule is equal to the
sum of the energies of the two photons. MPA, also called induced
absorption, is a third-order process that is several orders of
magnitude weaker than linear absorption. It differs from linear
absorption in that the strength of induced absorption depends on
the square of the light intensity, and thus it is a nonlinear
optical process.
[0043] Thus, the laser needs to generate pulse energies sufficient
to stimulate MPA in transparent materials over a length of
interest. For this application, a laser capable of sourcing 532 nm
(or shorter wavelength) light pulses of about 50 .mu.J or higher
energy for each pulse is necessary. Optical elements are selected
to produce a laser beam focal line within the body of the
transparent material as described below and in US application No.
61/752,489 filed on Jan. 15, 2013, the entire contents of which are
incorporated by reference as if fully set forth herein. The pulse
energy is then shaped and focused into a linear focal region
creating a minimum energy/length of about 100 .mu.J/mm. Within the
focal region (e.g., about 0.5 mm) the energy density is
sufficiently high to result in ionization. A photon at a wavelength
of 532 nm has an energy of about 2.3 eV. At the atomic level, the
ionization of individual atoms has discrete energy requirements as
shown in FIG. 1. Several elements commonly used in glass (e.g., Si,
Na, K) have relatively low ionization energies (about 5 eV).
Without the phenomena of MPA, a wavelength of about 248 nm would be
required to create linear ionization at 5 eV. With MPA, these bonds
are selectively ionized in the focal region, resulting in
separation from the adjacent molecules. This "disruption" in the
molecular bonding can result in non-thermal ablation removing
material from that region (perforating and thereby creating a
defect line). This can be accomplished with a single "burst" of
high energy pico-second pulses (spaced close together in
time--measured in nano-seconds). These "bursts" may be repeated at
high burst repetition rates (e.g., several hundred kHz). The
perforations, holes, or defect lines (these three terms are used
interchangeably herein) can be spaced apart by controlling the
relative velocity of a substrate. For example, the perforations are
generally spaced from 0.5 to 15 microns apart (for example, 2-12
microns, or, 5-10 microns). As an example, in a thin transparent
substrate moving at 200 mm/sec exposed to a 100 kHz series of
pulses, the perforations would be spaced 2 microns apart. This
spacing pitch is sufficient to allow for mechanical or thermal
separation. It has been noted that resulting debris is deposited in
a region local to the "cut" of about 50 microns in length, and the
debris is lightly adhered to the surface when the laser wavelength
is 532 nm. Particle size of the debris is typically less than about
500 nm.
[0044] The interior diameter (open air hole diameter) of the defect
line (typically less than about 300 nm) is consistent with the Abbe
diffraction limit described below.
[0045] Turning to FIGS. 2A and 2B, a method of laser drilling a
material includes focusing a pulsed laser beam 2 into a laser beam
focal line 2b, oriented along the beam propagation direction. As
shown in FIG. 3, laser 3 (not shown) emits laser beam 2, at the
beam incidence side of the optical assembly 6 referred to as 2a,
which is incident onto the optical assembly 6. The optical assembly
6 turns the incident laser beam into an extensive laser beam focal
line 2b on the output side over a defined expansion range along the
beam direction (length 1 of the focal line). The planar substrate 1
to be processed is positioned in the beam path after the optical
assembly overlapping at least partially the laser beam focal line
2b of laser beam 2. Reference la designates the surface of the
planar substrate facing the optical assembly 6 or the laser,
respectively, reference 1 b designates the reverse surface of
substrate 1 usually spaced in parallel. The substrate thickness
(measured perpendicularly to the planes la and lb, i.e., to the
substrate plane) is labeled with d.
[0046] As FIG. 2A depicts, substrate 1 is aligned perpendicularly
to the longitudinal beam axis and thus behind the same focal line
2b produced by the optical assembly 6 (the substrate is
perpendicular to the drawing plane) and oriented along the beam
direction it is positioned relative to the focal line 2b in such a
way that the focal line 2b viewed in beam direction starts before
the surface la of the substrate and stops before the surface lb of
the substrate, i.e. still within the substrate. In the overlapping
area of the laser beam focal line 2b with substrate 1, i.e. in the
substrate material covered by focal line 2b, the extensive laser
beam focal line 2b thus generates (in case of a suitable laser
intensity along the laser beam focal line 2b which is ensured due
to the focusing of laser beam 2 on a section of length 1, i.e. a
line focus of length 1) an extensive section 2c viewed along the
longitudinal beam direction, along which an induced absorption is
generated in the substrate material which induces a defect line or
crack formation in the substrate material along section 2c. The
crack formation is not only local, but over the entire length of
the extensive section 2c of the induced absorption. The length of
section 2c (i.e., after all, the length of the overlapping of laser
beam focal line 2b with substrate 1) is labeled with reference L.
The average diameter or the average extension of the section of the
induced absorption (or the sections in the material of substrate 1
undergoing the crack formation) is labeled with reference D. This
average extension D basically corresponds to the average diameter
.delta. of the laser beam focal line 2b, that is, an average spot
diameter in a range of between about 0.1 .mu.m and about 5
.mu.m.
[0047] As FIG. 2A shows, substrate material transparent for the
wavelength .lamda. of laser beam 2 is heated due to the induced
absorption along the focal line 2b. FIG. 2B outlines that the
warming material will eventually expand so that a correspondingly
induced tension leads to micro-crack formation, with the tension
being the highest at surface 1a.
[0048] Concrete optical assemblies 6, which can be applied to
generate the focal line 2b, as well as a concrete optical setup, in
which these optical assemblies can be applied, are described below.
All assemblies or setups are based on the description above so that
identical references are used for identical components or features
or those which are equal in their function. Therefore only the
differences are described below.
[0049] As the parting face eventually resulting in the separation
is or must be of high quality (regarding breaking strength,
geometric precision, roughness and avoidance of re-machining
requirements), the individual focal lines to be positioned on the
substrate surface along parting line 5 should be generated using
the optical assembly described below (hereinafter, the optical
assembly is alternatively also referred to as laser optics). The
roughness results particularly from the spot size or the spot
diameter of the focal line. In order to achieve a low spot size of,
for example, 0.5 .mu.m to 2 .mu.m in case of a given wavelength
.lamda. of laser 3 (interaction with the material of substrate 1),
certain requirements must usually be imposed on the numerical
aperture of laser optics 6. These requirements are met by laser
optics 6 described below.
[0050] In order to achieve the required numerical aperture, the
optics must, on the one hand, dispose of the required opening for a
given focal length, according to the known Abbe formulae (N.A.=n
sin (theta), n: refractive index of the glass to be processes,
theta: half the aperture angle; and theta=arctan (D/2f); D:
aperture, f: focal length). On the other hand, the laser beam must
illuminate the optics up to the required aperture, which is
typically achieved by means of beam widening using widening
telescopes between laser and focusing optics.
[0051] The spot size should not vary too strongly for the purpose
of a uniform interaction along the focal line. This can, for
example, be ensured (see the embodiment below) by illuminating the
focusing optics only in a small, circular area so that the beam
opening and thus the percentage of the numerical aperture only vary
slightly.
[0052] According to FIG. 3A (section perpendicular to the substrate
plane at the level of the central beam in the laser beam bundle of
laser radiation 2; here, too, laser beam 2 is incident
perpendicularly to the substrate plane, i.e. angle .beta. is
0.degree. so that the focal line 2b or the extensive section of the
induced absorption 2c is parallel to the substrate normal), the
laser radiation 2a emitted by laser 3 is first directed onto a
circular aperture 8 which is completely opaque for the laser
radiation used. Aperture 8 is oriented perpendicular to the
longitudinal beam axis and is centered on the central beam of the
depicted beam bundle 2a. The diameter of aperture 8 is selected in
such a way that the beam bundles near the center of beam bundle 2a
or the central beam (here labeled with 2aZ) hit the aperture and
are completely absorbed by it. Only the beams in the outer
perimeter range of beam bundle 2a (marginal rays, here labeled with
2aR) are not absorbed due to the reduced aperture size compared to
the beam diameter, but pass aperture 8 laterally and hit the
marginal areas of the focusing optic elements of the optical
assembly 6, which is designed as a spherically cut, bi-convex lens
7 here.
[0053] Lens 7 centered on the central beam is deliberately designed
as a non-corrected, bi-convex focusing lens in the form of a
common, spherically cut lens. Put another way, the spherical
aberration of such a lens is deliberately used. As an alternative,
aspheres or multi-lens systems deviating from ideally corrected
systems, which do not form an ideal focal point but a distinct,
elongated focal line of a defined length, can also be used (i.e.,
lenses or systems which do not have a single focal point). The
zones of the lens thus focus along a focal line 2b, subject to the
distance from the lens center. The diameter of aperture 8 across
the beam direction is approximately 90% of the diameter of the beam
bundle (beam bundle diameter defined by the extension to the
decrease to 1/e) and approximately 75% of the diameter of the lens
of the optical assembly 6. The focal line 2b of a not
aberration-corrected spherical lens 7 generated by blocking out the
beam bundles in the center is thus used. FIG. 3A shows the section
in one plane through the central beam, the complete
three-dimensional bundle can be seen when the depicted beams are
rotated around the focal line 2b.
[0054] One disadvantage of this focal line is that the conditions
(spot size, laser intensity) along the focal line, and thus along
the desired depth in the material, vary and that therefore the
desired type of interaction (no melting, induced absorption,
thermal-plastic deformation up to crack formation) may possibly
only be selected in a part of the focal line. This means in turn
that possibly only a part of the incident laser light is absorbed
in the desired way. In this way, the efficiency of the process
(required average laser power for the desired separation speed) is
impaired on the one hand, and on the other hand the laser light
might be transmitted into undesired deeper places (parts or layers
adherent to the substrate or the substrate holding fixture) and
interact there in an undesirable way (heating, diffusion,
absorption, unwanted modification).
[0055] FIG. 3B-1-4 show (not only for the optical assembly in FIG.
3A, but basically also for any other applicable optical assembly 6)
that the laser beam focal line 2b can be positioned differently by
suitably positioning and/or aligning the optical assembly 6
relative to substrate 1 as well as by suitably selecting the
parameters of the optical assembly 6: As FIG. 3B-1 outlines, the
length 1 of the focal line 2b can be adjusted in such a way that it
exceeds the substrate thickness d (here by factor 2). The laser
beam focal line 2b can have a length 1 in a range of between about
0.1 mm and about 100 mm or in a range of between about 0.1 mm and
about 10 mm, for example. Various embodiments can be configured to
have length 1 of about 0.1 mm, 0.2 mm, 0.3 mm, 0.4 mm, 0.5 mm, 0.7
mm, 1 mm, 2 mm, 3 mm or 5 mm, for example. If substrate 1 is placed
(viewed in longitudinal beam direction) centrally to focal line 2b,
an extensive section of induced absorption 2c is generated over the
entire substrate thickness.
[0056] In the case shown in FIG. 3B-2, a focal line 2b of length 1
is generated which corresponds more or less to the substrate
extension d. As substrate 1 relative to line 2 is positioned in
such a way that line 2b starts in a point before, i.e. outside the
substrate, the length L of the extensive section of induced
absorption 2c (which extends here from the substrate surface to a
defined substrate depth, but not to the reverse surface 1b) is
smaller than the length 1 of focal line 2b. FIG. 3B-3 shows the
case in which the substrate 1 (viewed along the beam direction) is
partially positioned before the starting point of focal line 2b so
that, here too, it applies to the length 1 of line 2b 1>L
(L=extension of the section of induced absorption 2c in substrate
1). The focal line thus starts within the substrate and extends
over the reverse surface 1 b to beyond the substrate. FIG. 3B-4
finally shows the case in which the generated focal line length 1
is smaller than the substrate thickness d so that--in case of a
central positioning of the substrate relative to the focal line
viewed in the direction of incidence--the focal line starts near
the surface 1a within the substrate and ends near the surface 1b
within the substrate (1=0.75d).
[0057] It is particularly advantageous to realize the focal line
positioning in such a way that at least one surface 1a, 1b is
covered by the focal line, i.e. that the section of induced
absorption 2c starts at least on one surface. In this way it is
possible to achieve virtually ideal cuts avoiding ablation,
feathering and particulation at the surface.
[0058] FIG. 4 depicts another applicable optical assembly 6. The
basic construction follows the one described in FIG. 3A so that
only the differences are described below. The depicted optical
assembly is based the use of optics with a non-spherical free
surface in order to generate the focal line 2b, which is shaped in
such a way that a focal line of defined length 1 is formed. For
this purpose, aspheres can be used as optic elements of the optical
assembly 6. In FIG. 4, for example, a so-called conical prism, also
often referred to as axicon, is used. An axicon is a special,
conically cut lens which forms a spot source on a line along the
optical axis (or transforms a laser beam into a ring). The layout
of such an axicon is generally known to one of skill in the art;
the cone angle in the example is 10.degree.. The apex of the axicon
labeled here with reference 9 is directed towards the incidence
direction and centered on the beam center. As the focal line 2b of
the axicon 9 already starts in its interior, substrate 1 (here
aligned perpendicularly to the main beam axis) can be positioned in
the beam path directly behind axicon 9. As FIG. 4 shows, it is also
possible to shift substrate 1 along the beam direction due to the
optical characteristics of the axicon without leaving the range of
focal line 2b. The extensive section of the induced absorption 2c
in the material of substrate 1 therefore extends over the entire
substrate depth d.
[0059] However, the depicted layout is subject to the following
restrictions: As the focal line of axicon 9 already starts within
the lens, a significant part of the laser energy is not focused
into part 2c of focal line 2b, which is located within the
material, in case of a finite distance between lens and material.
Furthermore, length 1 of focal line 2b is related to the beam
diameter for the available refraction indices and cone angles of
axicon 9, which is why, in case of relatively thin materials
(several millimeters), the total focal line is too long, having the
effect that the laser energy is again not specifically focused into
the material.
[0060] This is the reason for an enhanced optical assembly 6 which
comprises both an axicon and a focusing lens. FIG. 5A depicts such
an optical assembly 6 in which a first optical element (viewed
along the beam direction) with a non-spherical free surface
designed to form an extensive laser beam focal line 2b is
positioned in the beam path of laser 3. In the case shown in FIG.
5A, this first optical element is an axicon 10 with a cone angle of
5.degree., which is positioned perpendicularly to the beam
direction and centered on laser beam 3. The apex of the axicon is
oriented towards the beam direction. A second, focusing optical
element, here the plano-convex lens 11 (the curvature of which is
oriented towards the axicon), is positioned in beam direction at a
distance z1 from the axicon 10. The distance z1, in this case
approximately 300 mm, is selected in such a way that the laser
radiation formed by axicon 10 circularly is incident on the
marginal area of lens 11. Lens 11 focuses the circular radiation on
the output side at a distance z2, in this case approximately 20 mm
from lens 11, on a focal line 2b of a defined length, in this case
1.5 mm. The effective focal length of lens 11 is 25 mm here. The
circular transformation of the laser beam by axicon 10 is labeled
with the reference SR.
[0061] FIG. 5B depicts the formation of the focal line 2b or the
induced absorption 2c in the material of substrate 1 according to
FIG. 5A in detail. The optical characteristics of both elements 10,
11 as well as the positioning of them is selected in such a way
that the extension 1 of the focal line 2b in beam direction is
exactly identical with the thickness d of substrate 1.
Consequently, an exact positioning of substrate 1 along the beam
direction is required in order to position the focal line 2b
exactly between the two surfaces la and lb of substrate 1, as shown
in FIG. 5B.
[0062] It is therefore advantageous if the focal line is formed at
a certain distance from the laser optics, and if the greater part
of the laser radiation is focused up to a desired end of the focal
line. As described, this can be achieved by illuminating a
primarily focusing element 11 (lens) only circularly on a required
zone, which, on the one hand, serves to realize the required
numerical aperture and thus the required spot size, on the other
hand, however, the circle of diffusion diminishes in intensity
after the required focal line 2b over a very short distance in the
center of the spot, as a basically circular spot is formed. In this
way, the crack formation is stopped within a short distance in the
required substrate depth. A combination of axicon 10 and focusing
lens 11 meets this requirement. The axicon acts in two different
ways: due to the axicon 10, a usually round laser spot is sent to
the focusing lens 11 in the form of a ring, and the asphericity of
axicon 10 has the effect that a focal line is formed beyond the
focal plane of the lens instead of a focal point in the focal
plane. The length 1 of focal line 2b can be adjusted via the beam
diameter on the axicon. The numerical aperture along the focal
line, on the other hand, can be adjusted via the distance z1
axicon-lens and via the cone angle of the axicon. In this way, the
entire laser energy can be concentrated in the focal line.
[0063] If the crack (i.e., defect line) formation is supposed to
continue to the emergence side of the substrate, the circular
illumination still has the advantage that, on the one hand, the
laser power is used in the best possible way as a large part of the
laser light remains concentrated in the required length of the
focal line, on the other hand, it is possible to achieve a uniform
spot size along the focal line--and thus a uniform separation
process along the focal line--due to the circularly illuminated
zone in conjunction with the desired aberration set by means of the
other optical functions. The defect lines 120 extend, for example,
through the thickness of the glass sheet, and in the exemplary
embodiments described herein are orthogonal to the major (flat)
surfaces of the glass sheet.
[0064] Instead of the plano-convex lens depicted in FIG. 5A, it is
also possible to use a focusing meniscus lens or another higher
corrected focusing lens (asphere, multi-lens system).
[0065] In order to generate very short focal lines 2b using the
combination of an axicon and a lens depicted in FIG. 5A, it would
be necessary to select a very small beam diameter of the laser beam
inciding on the axicon. This has the practical disadvantage that
the centering of the beam onto the apex of the axicon must be very
precise and that therefore the result is very sensitive to
direction variations of the laser (beam drift stability).
Furthermore, a tightly collimated laser beam is very divergent,
i.e. due to the light deflection the beam bundle becomes blurred
over short distances.
[0066] As shown in FIG. 6, both effects can be avoided by inserting
another lens, a collimating lens 12: this further, positive lens 12
serves to adjust the circular illumination of focusing lens 11 very
tightly. The focal length f of collimating lens 12 is selected in
such a way that the desired circle diameter dr results from
distance z1a from the axicon to the collimating lens 12, which is
equal to f. The desired width br of the ring can be adjusted via
the distance z1b (collimating lens 12 to focusing lens 11). As a
matter of pure geometry, the small width of the circular
illumination leads to a short focal line. A minimum can be achieved
at distance f.
[0067] The optical assembly 6 depicted in FIG. 6 is thus based on
the one depicted in FIG. 5A so that only the differences are
described below. The collimating lens 12, here also designed as a
plano-convex lens (with its curvature towards the beam direction)
is additionally placed centrally in the beam path between axicon 10
(with its apex towards the beam direction), on the one side, and
the plano-convex lens 11, on the other side. The distance of
collimating lens 12 from axicon 10 is referred to as z1a, the
distance of focusing lens 11 from collimating lens 12 as z1b, and
the distance of the generated focal line 2b from the focusing lens
11 as z2 (always viewed in beam direction). As shown in FIG. 6, the
circular radiation SR formed by axicon 10, which is incident
divergently and under the circle diameter dr on the collimating
lens 12, is adjusted to the required circle width br along the
distance z1b for an at least approximately constant circle diameter
dr at the focusing lens 11. In the case shown, a very short focal
line 2b is supposed to be generated so that the circle width br of
approximately 4 mm at lens 12 is reduced to approximately 0.5 mm at
lens 11 due to the focusing properties of lens 12 (circle diameter
dr is 22 mm in the example).
[0068] In the depicted example it is possible to achieve a length
of the focal line 1 of less than 0.5 mm using a typical laser beam
diameter of 2 mm, a focusing lens 11 with a focal length f=25 mm,
and a collimating lens with a focal length f=150 mm, and choosing
Z1a=Z1b=140 mm and Z2=15 mm.
[0069] Note that, as shown in FIGS. 7A-7C, according to at least
some embodiments, the typical operation of such a picosecond laser
creates a "burst" 710 of pulses 720 (also referred to as a pulse
burst herein). Each "burst" 710 may contain multiple pulses 720
(such as at least 2 pulses, at least 3 pulses as shown in FIGS.
7A-7B, at least 4 pulses, at least 5 pulses as shown in FIG. 7C, at
least 10 pulses, at least 15 pulses, at least 20 pulses, or more)
of very short duration (e.g., .about.10 psec). Each pulse 720
within a burst is separated from an adjacent pulse in time by a
duration in a range of between about 1 nsec and about 50 nsec, such
as approximately 20 nsec (50 MHz), with the time often governed by
the laser cavity design. The time between each "burst" 710 will be
much longer, often about 10 .mu.sec, for a laser burst repetition
rate of about 100 kHz. That is, a pulse burst is a "pocket" of
pulses, and the bursts are separated from one another by a longer
duration than the separation of individual adjacent pulses within
each burst. The exact timings, pulse durations, and burst
repetition rates can vary depending on the laser design, but short
pulses (i.e., less than about 15 psec) of high intensity have been
shown to work well with this technique.
[0070] More specifically, in these embodiments pulses 720 typically
have pulse duration T.sub.d of up to 100 psec (for example, 0.1
psec, 5 psec, 10 psec, 15 psec, 18 psec, 20 psec, 22 psec, 25 psec,
30 psec, 50 psec, 75 psec, or therebetween). The energy or
intensity of each individual pulse 720 within the burst may not be
equal to that of other pulses within the burst, and the intensity
distribution of the multiple pulses within a burst 710 often
follows an exponential decay in time governed by the laser design.
Preferably, each pulse 720 within the burst 710 of the exemplary
embodiments described herein are separated in time from the
subsequent pulse in the burst by a duration T.sub.p from 1 nsec to
50 nsec (e.g. 10-50 nsec, or 10-30 nsec, with the time often
governed by the laser cavity design). For a given laser, the time
separation T.sub.p between adjacent pulses (pulse-to-pulse
separation) within a burst 710 is relatively uniform (.+-.10%). For
example, in some embodiments, each pulse within a burst is
separated in time from the subsequent pulse by approximately 20
nsec (50 MHz). For example, for a laser that produces pulse
separation T.sub.p of about 20 nsec, the pulse to pulse separation
T.sub.p within a burst is maintained within about .+-.10%, or is
about .+-.2 nsec. The time between each "burst" of pulses (i.e.,
time separation T.sub.b between bursts) will be much longer (e.g.,
0.25.ltoreq.T.sub.b.ltoreq.1000 microseconds, for example 1-10
microseconds, or 3-8 microseconds). In some of the exemplary
embodiments of the laser described herein the time separation
T.sub.b is around 5 microseconds for a laser with burst repetition
rate or frequency of about 200 kHz. The laser burst repetition rate
is defined as the time between the first pulse in a burst to the
first pulse in the subsequent burst. In some embodiments, the burst
repetition frequency may be in a range of between about 1 kHz and
about 4 MHz. More preferably, the laser burst repetition rates can
be, for example, in a range of between about 10 kHz and 650 kHz.
The time T.sub.b between the first pulse in each burst to the first
pulse in the subsequent burst may be 0.25 microsecond (4 MHz burst
repetition rate) to 1000 microseconds (1 kHz burst repetition
rate), for example 0.5 microseconds (2 MHz burst repetition rate)
to 40 microseconds (25 kHz burst repetition rate), or 2
microseconds (500 kHz burst repetition rate) to 20 microseconds (50
k Hz burst repetition rate). The exact timings, pulse durations,
and burst repetition rates can vary depending on the laser design,
but short pulses (T.sub.d<20 psec and preferably
T.sub.d.ltoreq.15 psec) of high intensity have been shown to work
particularly well.
[0071] The required energy to modify the material can be described
in terms of the burst energy--the energy contained within a burst
(each burst 710 contains a series of pulses 720), or in terms of
the energy contained within a single laser pulse (many of which may
comprise a burst). For these applications, the energy per burst can
be from 25-750 .mu.J, more preferably 50-500 .mu.J, or 50-250
.mu.J. In some embodiments the energy per burst is 100-250 .mu.J.
The energy of an individual pulse within the pulse burst will be
less, and the exact individual laser pulse energy will depend on
the number of pulses 720 within the pulse burst 710 and the rate of
decay (e.g., exponential decay rate) of the laser pulses with time
as shown in FIGS. 7B and 7C. For example, for a constant
energy/burst, if a pulse burst contains 10 individual laser pulses
720, then each individual laser pulse 720 will contain less energy
than if the same pulse burst 710 had only 2 individual laser
pulses.
[0072] The use of laser capable of generating such pulse bursts is
advantageous for cutting or modifying transparent materials, for
example glass. In contrast with the use of single pulses spaced
apart in time by the repetition rate of the single- pulsed laser,
the use of a pulse burst sequence that spreads the laser energy
over a rapid sequence of pulses within the burst 710 allows access
to larger timescales of high intensity interaction with the
material than is possible with single-pulse lasers. While a
single-pulse can be expanded in time, as this is done the intensity
within the pulse must drop as roughly one over the pulse width.
Hence if a 10 psec single pulse is expanded to a 10 nsec pulse, the
intensity drops by roughly three orders of magnitude. Such a
reduction can reduce the optical intensity to the point where
non-linear absorption is no longer significant, and light material
interaction is no longer strong enough to allow for cutting. In
contrast, with a pulse burst laser, the intensity during each pulse
720 within the burst 710 can remain very high--for example three 10
psec pulses 720 spaced apart in time by approximately 10 nsec still
allows the intensity within each pulse to be approximately three
times higher than that of a single 10 psec pulse, while the laser
is allowed to interact with the material over a timescale that is
now three orders of magnitude larger. This adjustment of multiple
pulses 720 within a burst thus allows manipulation of time-scale of
the laser-material interaction in ways that can facilitate greater
or lesser light interaction with a pre-existing plasma plume,
greater or lesser light-material interaction with atoms and
molecules that have been pre-excited by an initial or previous
laser pulse, and greater or lesser heating effects within the
material that can promote the controlled growth of micro cracks.
The required amount of burst energy to modify the material will
depend on the substrate material composition and the length of the
line focus used to interact with the substrate. The longer the
interaction region, the more the energy is spread out, and higher
burst energy will be required. The exact timings, pulse durations,
and burst repetition rates can vary depending on the laser design,
but short pulses (<15 psec, or <10 psec) of high intensity
have been shown to work well with this technique. The defect line
or a hole is formed in the material when a single burst of laser
pulses strikes essentially the same location on the glass. That is,
multiple laser pulses within a single burst correspond to a single
defect line or a hole location in the glass. Of course, since the
glass is translated (for example by a constantly moving stage) (or
the beam is moved relative to the glass, the individual pulses
within the burst cannot be at exactly the same spatial location on
the glass. However, they are well within 1 .mu.m of one
another--i.e., they strike the glass at essentially the same
location. For example, they may strike the glass at a spacing sp
where 0<sp.ltoreq.500 nm from one another. For example, when a
glass location is hit with a burst of 20 pulses the individual
pulses within the burst strike the glass within 250 nm of each
other. Thus, in some embodiments 1 nm<sp<250 nm. In some
embodiments 1 nm<sp<100 nm.
[0073] The laser beam has a wavelength less than or equal to about
850 nm, the wavelength selected such that the material is
substantially transparent (i.e., absorption less than about 10%,
preferably less than about 1% per mm of material depth) at this
wavelength, the laser beam having an average laser energy measured
at the material greater than about 50 .mu.J per mm of material
thickness, and pulses having a duration in a range of between
greater than about 1 picosecond and less than about 100
picoseconds, and a pulse burst repetition rate in a range of
between about 1 kHz and about 2 MHz. The method then includes
directing the laser beam focal line into the material, the laser
beam focal line generating an induced absorption within the
material, the induced absorption producing a defect line along the
laser beam focal line within the material, and producing subsurface
damage up to a depth less than or equal to 100 .mu.m within the
material, for example than or equal to about 75 .mu.m within the
material, and in some embodiments .ltoreq.50 .mu.m, for example,
<40 .mu.m.
[0074] The depth of subsurface damage can be measured by using a
confocal microscope to look at the cut surface, the microscope
having an optical resolution of a few nm. Surface reflections are
ignored while cracks are sought out down into the material, the
cracks showing up as bright lines. One then steps into the material
until there are no more "sparks", collecting images at regular
intervals. The images are then manually processed by looking for
cracks and tracing them through the depth of the glass to get a
maximum depth (typically measured in microns (.mu.m)) of subsurface
damage. There are typically thousands and thousands of cracks, so
one typically just tracks the largest ones. One typically repeats
this process on about 5 locations of a cut edge. Any cracks that
are directly perpendicular to the edge of the glass will not be
detected by this method.
[0075] In some embodiments, the method further includes translating
the material and the laser beam relative to each other, thereby
drilling a plurality of defect lines within the material, the
defect lines spaced apart so as to separate the material into at
least two pieces. For cutting operations, the laser triggering
generally is synchronized with the stage driven motion of the
material beneath the beam, so laser pulse bursts are triggered at a
fixed interval, such as for example every 1 .mu.m, or every 5
.mu.m. The exact spacing between adjacent perforations or defect
lines is determined by the material properties that facilitate
crack propagation from perforated hole (i.e., defect line) to
perforated hole, given the stress level in the substrate. However,
in contrast to cutting a substrate, it is also possible to use the
same method to only perforate the material, such as for creating
holes for conducting electrical signals from one part to another,
thereby creating components called interposers. In the case of
interposers, the defect lines are generally separated by much
greater distance than required for cutting--instead of a pitch of
about 10 .mu.m or less, the spacing between defect lines can be
hundreds of microns. The exact locations of the defect lines need
not be at regular intervals--the location simply is determined by
when the laser is triggered to fire, and may be at any location
within the part.
[0076] The embodiments of the process described herein can cut
glass at a cutting speed of 0.25 m/sec, or faster. A cut speed (or
cutting speed) is the rate the laser beam moves relative to the
surface of the transparent material (e.g., glass) while creating
multiple holes or modified regions.) High cut speeds, such as, for
example 400 mm/sec, 500 mm/sec, 750 mm/sec, 1 m/sec, 1.2 m/sec, 1.5
m/sec, or 2 m/sec, or even 3 m/sec to 4 m/sec are often desired in
order to minimize capital investment for manufacturing, and to
optimize equipment utilization rate. The laser power is equal to
the burst energy multiplied by the burst repetition frequency
(rate) of the laser. In general, to cut such glass materials at
high cutting speeds, the damage tracks are typically spaced apart
by 1-25 microns, in some embodiments the spacing is preferably 2
microns or larger--for example 2-12 microns, or for example 3-10
microns.
[0077] For example, to achieve a linear cutting speed of 300
mm/sec, 3 micron hole pitch corresponds to a pulse burst laser with
at least 100 kHz burst repetition rate. For a 600 mm/sec cutting
speed, a 3 micron pitch corresponds to a burst-pulsed laser with at
least 200 kHz burst repetition rate. A pulse burst laser that
produces at least 40 .mu.J/burst at 200 kHz, and cuts at a 600 mm/s
cutting speed needs to have laser power of at least 8 Watts. Higher
cut speeds therefore require even higher laser powers.
[0078] For example 0.4 m/sec cut speed at 3 .mu.m pitch and 4
.mu.J/burst would require at least 5 Watt laser power delivered, a
0.5 m/sec cut speed at 3 .mu.m pitch and 40 .mu.J/burst would
require at least 6 Watt laser delivered. Thus, preferably the laser
power of the pulse burst picosecond laser is 6 watts or higher,
more preferably at least 8 Watts or higher, and even more
preferably at least 10W or higher. For example in order to achieve
a 0.4 m/sec cut speed at 4 .mu.m pitch (defect lines pacing, or
between damage tracks spacing) and 100 .mu.J/burst one would
require at least a 10 Watt laser, and to achieve a 0.5 m/sec cut
speed at 4 .mu.m pitch and 100 .mu.J/burst one would require at
least a 12 Watt laser. For example, to achieve a cut speed of 1
m/sec at 3 .mu.m pitch and 40 .mu.J/burst one would require at
least a 13 Watt laser. Also for example 1 m/sec cut speed at 4
.mu.m pitch and 400 .mu.J/burst would require at least a 100 Watt
laser. The optimal pitch between damage tracks and the exact burst
energy is material dependent, and can be determined empirically.
However, it should be noted that raising the laser pulse energy or
making the damage tracks at a closer pitch are not conditions that
always make the substrate material separate better or with improved
edge quality. Too dense a pitch (for example <0.1 micron, in
some exemplary embodiments <1 .mu.m, or in some embodiments
<2 .mu.m) between damage tracks can sometimes inhibit the
formation of nearby subsequent damage tracks, and often can inhibit
the separation of the material around the perforated contour, and
may also result in increased unwanted micro cracking within the
glass. Too long a pitch (>50 .mu.m, and in some glasses >25
.mu.m or even >20 .mu.m) may result in "uncontrolled
microcracking"--i.e., where instead of propagating from hole to
hole the microcracks propagate along a different path, and cause
the glass to crack in a different (undesirable) direction. This may
ultimately lower the strength of the separated glass part, since
the residual microcracks will acts as flaws which weaken the glass.
Too high a burst energy (e.g., >2500 .mu.J/burst, and in some
embodiments >500 .mu.J/burst) used to form each damage track can
cause "healing" or re-melting of already formed microcracks of
adjacent damage tracks, which will inhibit separation of the glass.
Accordingly, it is preferred that burst energy be <2500
.mu.J/burst, for example, .ltoreq.500 .mu.J/burst. Also, using a
burst energy that is too high can cause formation of microcracks
that are extremely large and create flaws which reduce the edge
strength of the parts after separation. Too low a burst energy
(<40 .mu.J/burst) may result in no appreciable damage track
formed within the glass, and hence very high separation strength or
complete inability to separate along the perforated contour.
[0079] Typical exemplary cutting rates (speeds) enabled by this
process are, for example, 0.250 m/sec and higher. In some
embodiments the cutting rates are at least 300 mm/sec. In some
embodiments described herein the cutting rates are at least 400
mm/sec, for example 500 mm/sec to 2000 mm/sec, or higher. In some
embodiments the picosecond laser utilizes pulse bursts to produce
defect lines with periodicity between 0.5 microns and 13 microns,
e.g. 0.5 and 3 microns. In some embodiments the pulsed laser has
laser power of 10 W-100 W and the material and/or the laser beam
are translated relative to one another at a rate of at least 0.25
m/sec, for example at the rate of 0.25 to 0.35 m/sec, or 0.4 m/sec
to 5 m/sec. Preferably, each pulse burst of the pulsed laser beam
has an average laser energy measured at the workpiece greater than
40 microJoules per burst mm thickness of workpiece. Preferably,
each pulse burst of the pulsed laser beam has an average laser
energy measured at the workpiece greater of less than 2500
microJoules per burst per mm thickness of workpiece, and preferably
less than about 2000 microJoules per burst per mm, and in some
embodiments less than 1500 microJoules per burst per mm thickness
of workpiece, for example not more than 500 microJoules per burst
per mm thickness of workpiece
[0080] Accordingly, it is preferable that the laser produces pulse
bursts with at least 2 pulses per burst. For example, in some
embodiments the pulsed laser has laser power of 10W-150W (e.g.,
10-100W) and produces pulse bursts with at least 2 pulses per burst
(e.g., 2 to 25 pulses per burst). In some embodiments the pulsed
laser has the power of 25 W to 60 W, and produces pulse bursts with
at least 2 to 25 pulses per burst, and the distance or periodicity
between the adjacent defect lines produced by the laser bursts is
2-10 microns. In some embodiments the pulsed laser has laser power
of 10W to 100W, produces pulse bursts with at least 2 pulses per
burst, and the workpiece and the laser beam are translated relative
to one another at a rate of at least 0.25 m/sec. In some
embodiments the workpiece and/or the laser beam are translated
relative to one another at a rate of at least 0.4 m/sec
[0081] For cutting speeds of 0.4 m to 5 m/sec, laser powers should
preferably be 10 W-150W, with burst energy of 40-750 .mu.J/burst,
2-25 bursts per pulse (depending on the material that is cut), and
hole separation (or pitch) of 3 to 15 .mu.m, or 3-10 .mu.m. The use
of picosecond pulse burst lasers would be preferable for these
cutting speeds because they generate high power and the required
number of pulses per burst. Thus, according to some exemplary
embodiments, the pulsed laser produces 10-100W of power, for
example 25 W to 60 Watts, and produces pulse bursts at least 2-25
pulses per burst and the distance between the defect lines is 2-15
microns; and the laser beam and/or the workpiece are translated
relative to one another at a rate of at least 0.25 m/sec, in some
embodiments at least 0.4 m/sec, for example 0.5 m/sec to 5 m/sec,
or faster.
Exemplification
[0082] Samples of Corning 2320 Gorilla.RTM. (ion-exchanged, also
called "Full Gorilla" (FG)) glass, shown in FIG. 8, were tested for
strength, sub-surface damage, and surface roughness. The edge
strength of samples cut using the 532 nm process described above is
shown in FIG. 9. As shown in Tables 1 and 2 below, samples cut
using the 532 nm process had an average subsurface damage of about
23 .mu.m, while samples cut using a 1064 nm process described in
U.S. Application No. 61/752,489 filed on Jan. 15, 2013 had an
average subsurface damage of about 74 .mu.m.
TABLE-US-00001 TABLE 1 RMS roughness and subsurface damage results
SAMPLE 203- 0.545 mm Full GG 2320 SSD .mu.m RMS PV 1064 nm (nm)
(nm) Results Spot A 1061 11434 69 Spot B 1220 14745 77 Spot C 1164
11350 68 Spot D 1165 15476 78 Spot E 994 10136 80 Mean 1120.8
12628.2 74.4 Range 226 5340 12 Std Dev 91.3 2337.8 5.5
TABLE-US-00002 TABLE 2 Subsurface damage at different burst
repetition rates and stage speeds Max No. of SSD Areas Each Area
Sample ID (.mu.m) Examined Max SSD 1.sup.st Sample 29 6 22, 24, 26,
26, 23, 29 20 KHz - 20 mm/s - 58 1 58 Side of Sample 20 KHz - 20
mm/s 38 3 28, 38, 32 20 KHz - 40 mm/s 28 5 21, 20, 23, 28, 23 10
KHz - 30 mm/s 37 5 29, 27, 24, 27, 37 10 KHz - 40 mm/s 54 5 54, 38,
35, 33, 31
[0083] FIG. 10 includes several SEM micrographs of several cut
surfaces, showing laser confocal cut surface scans at 20.times.
magnification. Table 3 below shows measurements of surface
roughness as a function of pulse spacing, measured with a Zygo
optical surface profiler, showing that Ra and RMS surface roughness
both appear to increase with pulse spacing. Zygo Corporation,
Middlefield, Conn.
TABLE-US-00003 TABLE 3 Zygo Surface Roughness vs. Pulse Spacing
Sample ID Ra (.mu.m) RMS (.mu.m) Pulse Spacing AVG AVG Varied 5
Points 5 Points 1 .31 .4 2 .43 .56 3 .54 .7 4 .66 .83
[0084] Zygo representative scans are shown in FIGS. 11A-11D. FIGS.
12A-12C show photographs of edges of samples cut using the 532 nm
process (FIGS. 12A and 12C) and a reference edge of a sample cut
using the 1064 nm process. FIG. 13A shows a higher magnification
photograph of features made with the 1064 nm process, where the
diameter of the hole was measured at 347 nm, as compared to the
diameter of a hole made with the 532 nm process (FIG. 13B) that was
190 nm.
[0085] Table 4 below shows a comparison of the RMS surface
roughness and subsurface damage (SSD) between samples cut using the
1064 nm process and samples cut using the 532 nm described
above.
TABLE-US-00004 TABLE 4 1064 nm v. 532 nm Direct Comparison 1064 nm
532 nm 1064 nm 532 nm RMS RMS SSD SSD (nm).sup.1 (nm).sup.2
(.mu.m).sup.1 (.mu.m).sup.2 Avg 1120.8 556 74.4 23 Range 226 20 12
8 Std Dev 91.3 9 5.5 3 Notes: .sup.10.55 mm 2320 FG glass
.sup.20.40 mm 2320 FG glass (2 .mu.m spacing cut-strength
sample)
[0086] FIGS. 14A and 14B show a comparison of edge strength of
samples cut using the 1064 nm process (FIG. 14A) and the 532 nm
process (FIG. 14B), showing that the edge strengths of the samples
produced by the two processes are relatively similar.
[0087] The relevant teachings of all patents, published
applications and references cited herein are incorporated by
reference in their entirety.
[0088] While exemplary embodiments have been disclosed herein, it
will be understood by those skilled in the art that various changes
in form and details may be made therein without departing from the
scope encompassed by the appended claims.
* * * * *