KR100829009B1 - Energy-efficient, laser-based method and system for processing target material - Google Patents

Abstract

There is provided a method and system for treating a target material, such as a microstructure, in a microscopic region without causing undesirable changes in the electrical and / or physical properties of the material around the target material. The system includes a controller for generating a control signal and a signal generator for generating a modulated drive waveform based on the control signal. The system also includes a gain switched pulse seed laser having a wavelength that generates a laser pulse train at a predetermined repetition rate. The drive waveforms are pumped so that each pulse of the pulse train has a predetermined shape. Further, the system includes a laser amplifier that optically amplifies the pulse train to obtain an amplified pulse train such that the predetermined shape of the pulse is not significantly changed. Each amplified pulse has a temporary quadratic power density distribution, a steep rise time, a pulse duration and a fall time. The system also includes a beam delivery and focusing subsystem that focuses at least a portion of the amplified pulse train to a spot of target material material.

Energy efficient laser

Description

[0001] ENERGY-EFFICIENT, LASER-BASED METHOD AND SYSTEM FOR PROCESSING TARGET MATERIAL [0002]

The present invention relates to an energy efficient, laser-based target material processing method and system. More particularly, the present invention relates to the use of pulsed laser beams to ablate or modify portions of circuit elements on a semiconductor substrate, and more particularly to evaporation of metal, polysilicide, and polysilicon links for memory repair . Additional applications can be found in laser-based micromachining and other repair work, particularly when removing microstructures without damaging surrounding areas and structures with often non-homogeneous optical and thermal properties Or if it is desired to change it, additional embodiments may be found. Likewise, material handling operations can be applied to other micro-semiconductor devices, such as microelectromechanical machines. There may also be applications for medical applications, such as microfibers or ablation by microfibre optical probes.

Semiconductor devices, such as memories, typically have a conductive link attached to a transparent insulating layer, such as silicon oxide, supported by a main silicon substrate. During laser processing of such a semiconductor device, when the beam is incident on a link or circuit, some of its energy reaches the substrate and other structures as well. Depending on the power of the beam, the length of time it takes to apply the beam, and other operating parameters, the silicon substrate and / or its vicinity may be overheated and damaged.

The prior art of various documents describes the importance of wavelength selection as a critical parameter for controlling substrate damage. U.S. Patent Nos. 5,265,114, 5,473,624, 5,998,759 disclose advantages over choosing wavelengths in the range of 1.2 袖 m or more to prevent silicon substrate damage.

The '759 patent further describes the wavelength characteristics of silicon in detail. The absorption in silicon has dropped abruptly after about 1 micron, and has an absorption edge of about 1.12 microns at room temperature. At wavelengths greater than 1.12 microns, silicon begins to transfer more and more easily, and therefore it is possible to obtain better part yields as soon as material is removed from the silicon. In the range of about 1 micron, the absorption coefficient is lowered by 10 ^ 4 times while going from 0.9 micron to 1.2 micron. The curve shows a decline of 10 ^ 2 while going from 1.047 micron to 1.2 micron, which is the standard laser wavelength. This indicates that there is a drastic change in absorption for very small changes in wavelength. Thus avoiding damage to the substrate by operating the laser at wavelengths above the absorption edge of the substrate. This is especially important when the laser beam has some misalignment with respect to the link or when the focused spot extends beyond the link structure. Furthermore, as the substrate temperature rises during processing, more absorption curve shifts occur in the infrared direction, which can lead to thermal runaway conditions and very severe damage.

The problem of repairing liquid crystals is similar to the problem of removing metal links. The principle of wavelength selection for maximizing absorption contrast has been beneficially applied in the green wavelength range in a manner similar to that described in the above document, for the same purpose, i.e. to remove metal without damaging the substrate. Systems manufactured by Florod are described in the publication "Xenon Laser Repairs Liquid Crystal Displays," LASERS AND OPTRONICS, April 1988, pages 39-41.

It has been determined that other parameters can be adjusted to improve the laser processing window, as wavelength selection has proven useful. For example, LM Scarfone and JD Chlipala, "Computer Simulation of Target Link Explosion in Laser Programmable Redundancy for Silicon Memory, 1986, page 371" It is better to choose the wavelength of the laser and the thickness of the material to increase absorption and reduce absorption for other purposes. "Generally, the usefulness of placing a thicker insulating layer under a link or circuit element, The usefulness of limiting the duration has also been identified, as shown in the co-authored article "Laser Adjustment of Linear Monolithic Circuits," Litwin and Smart, 100 / LIA, Vol. 38, ICAELO (1983).

The '759 patent discloses a tradeoff with the choice of longer wavelengths, more specifically a compromise on available spot size, depth of focus, and pulse width from an Nd: YAG laser Lt; / RTI > These parameters are extremely important in laser processing at finer dimensions and in the presence of the possibility of incidental damage to the surrounding structures.

In practice, in industry, any improvement that widens the processing window is not possible because the depth or lateral dimension seeks higher density microstructures and related geometries that are a fraction of a micron. useful. In this scale, the tolerance in energy control and target absorption is large compared to the energy required to treat the microstructure. It should be noted that, in the above description, the laser processing parameters are not necessarily independent of each other in micromachining applications where small laser spots of about 1 mu m are required. For example, the spot size and pulse width are generally minimized at short wavelengths of about 1.2 mu m or less, but the absorption contrast is not maximized. It is common for semiconductor device manufacturers to continue to produce previously developed products during the development and start of production of a higher version of the product, typically using different structures and processes. Many recent memory products use polysilicon or polysilicon links, with smaller metal link structures being used in more advanced products such as 256 megabit memory. In such large memory memories, links of 1 micron width and 1/3 micron depth, lying on a thin silicon oxide layer of 0.3 to 0.5 micron, are used. As production equipment, Q-switching diode pump YAG lasers and related equipment capable of operating at existing wavelengths of 1.047 μm-1.32 μm and related equipment capable of operating within the wavelength range determined by low absorption by silicon have been utilized. However, the user also recognizes the need for equipment improvements that can cleanly cut the link structure without the risk of subsequent chip breakage due to contamination or conductive residues around the site of removal.

Other degrees of freedom include laser pulse energy density (delivered to the target material) and pulse duration. The prior art describes that the pulse width must be limited to prevent damage in micromachining applications. For example, U.S. Patent No. 5,059,764 discloses a laser processing workstation wherein a q-switching laser system is utilized to produce relatively short pulses, among others, in particular 10-50 ns. Here, the output pulse width must be relatively short for material processing applications (such as semiconductor memory repair through link blowing and precision engraving), and for pulse widths less than 50 ns for many applications, For example, a pulse width of 30 ns is required. Elimination (evaporation without melting) is possible by appropriately selecting the pulse width.

The high-speed pulse laser design can utilize Q-switching, gain-switched, or mode-locked operation. The pulse duration and shape of standard Q-switching and other pulsed lasers can be determined by integrating the combined ratio equation describing the photon number density and the population inversion related to the laser operating threshold at the start of the pulse It can be estimated at the basic level. In the case of Q-switching, in a normalized scale, the pulse rise time is faster as the number of atoms is higher in the inverted probability density with respect to the threshold value, and the peak energy is higher as the width is narrower. As this ratio decreases, the pulse shape becomes wider and has lower energy density.

Often the Q-switching laser pulses resemble a Gaussian time distribution, a mixed form of Gaussian with an exponential decaying tail. As disclosed in the '759 patent, a diode pump system of shorter wavelength is measured at half power points (i.e., the standard definition of pulse duration), and when operating in the preferred wavelength range, A relatively short pulse of 10 ns can be generated. Despite successful operation, the Applicant has found several limitations associated with the temporal pulse shape characteristics of a standard diode pump Q-switching laser system. This includes limitations of actual rise time, power distribution between half maximum points, and pulse decay characteristics. Improved metal link blowing applications using the method and system according to the present invention provide significantly better results.

Throughout the remainder of the description, "pulse shaping" refers to the generation of laser pulses detected by an electromagnetic radiation detector, where the "shape" is the power appearing at the detector as a function of time Quot; Further, "pulse width" or "pulse duration" refers to the full width at half maximum (FWHM) unless otherwise stated. Q-switched pulses may also be used to generate a Q-switched pulses, for example, a Q-switched pulses, for example, which resemble a composite of a very slowly decaying exponential tail and a Gaussian central lobe The temporal distribution of the pulses obtained in the system is referred to collectively. Such a wave shape is formally referred to as a "Q-switched pulse envelope" in the laser related literature. Figure 1C shows such a pulse.

U.S. Patent No. 5,208,437 (i.e., the '437 patent) specifies a pulse width specification of 1 ns or less for application to memory repair. &Quot; Laser Cutting of Aluminum Thin Film with No Damage to Under Layers ", ANNALS OF THE CIRP, Vol. 28/1, 1979, As well as experimental results of relatively short laser pulses having a "Gaussian" shape. The result indicates that the "preferred interconnection pattern portion" made of aluminum or the like can be "cut without damaging the layer underlying the interconnection pattern". Specifications for a pulse width of substantially less than 1 ns with an energy density of substantially 10 < ⁶ > W / cm < ² > However, although the beam is spatially formed to correspond to the interconnection pattern, a temporal pulse forming method has not been disclosed. Furthermore, in the ultrafast range approaching 100-300 ps as used in the '437 patent, applicant's analysis of high density memory devices with multiple layers with specified pulse widths was unsatisfactory. To overcome these limitations, ultrafast laser systems are needed to generate multiple pulses to process each target location, which can slow the laser processing speed to an unacceptable level.

Following the very high-speed scale, experimental results on micromachining work have been disclosed. Ultrafast pulses exhibit atomic and molecular level material properties that are fundamentally different from those found in the range of several hundreds ps to ns at reduced scales with durations of fs (10-15 sec) to ps (10-12) .

In US Pat. No. 5,656,186 and in the publication "Ultrashort Laser Pulses tackle precision Machining", Laser Focus World, August 1997, pages 101-118, machining operations at various wavelengths were analyzed, It has been suggested that the size of a feature is significantly smaller than the diffraction limited spot size of the focused beam.

Laser systems for ultrafast pulse generation vary in complexity and exemplary embodiments are described in U.S. Patent Nos. 5,920,668 and 5,400,350 and in PHOTONICS SPECTRA, July 1998, 157-161, "High Speed Laser Ultrafast Lasers Escape The Lab ". This embodiment generally involves a method of pulse stretching a mode lock ultrafast pulse before amplifying to prevent amplifier saturation followed by compression to an extremely narrow width. These techniques presuppose some kind of micromachining and possibly some finer-scale "nanomachining" work, where the advantages of nanomechanical machining are due to machining below the diffraction limit . However, the Applicant has found practical limitations at the present time with the available power within each pulse in applications such as metal link blowing and similar micromachining applications, which limit is not acceptable for multiple pulses Leads.

Applicants will elaborate on the theoretical interpretation of the use of short pulses, and since the fast rise time pulses have a variety of reasons, and many theoretical and authoritative papers and books have been written, this will be mentioned in the next paragraph . Although removal of the metal link is used as an example, the principle can be applied in a number of laser treatments in which the target material is surrounded by materials having substantially different optical and thermal properties. The following are examples of References 1-3.

1. John F. Ready, Effects of High Power Laser Radiation, ACADEMIC PRESS, New York 1971, pages 115-116.

2. Sidney S. Charschan, Guide for Material Processing by Lasers, Laser Institute of America, The Paul M. Harrod Company, Baltimore MD, 1977, pages 5-13.

3. Joseph Bernstein, J. H. Lee, Gang Yang, and Tariq A. Dahmas, Analysis of Laser Metal-Cut Energy Process Window.

Metal reflectivity

The reflectivity of the metal becomes lower as the power density of the laser pulse increases (Reference 1). The reflectivity of the metal is directly proportional to the free electron conductivity in the material. At high electric field densities, such as when delivered by a high intensity laser, the collision time between the electrons and the lattice is reduced. This shortening of the collision time reduces the conductivity and thus reduces the reflectivity. For example, as the laser power density increases to a range of 10 ⁹ watts / cm ² , the reflectivity of aluminum is reduced from 92% to less than 25%. It is therefore advantageous to obtain a high power density in the workpiece as short as possible in order to prevent the laser energy from being lost to reflection.

Thermal diffusivity

The distance D during which the heat travels during the laser pulse is proportional to the laser pulse width as follows.

here,

K is the thermal diffusivity of the material; And

t is the length of the laser pulse.

Therefore, it can be seen that heat dissipation to the substrate under the link where the short laser pulse is melted is prevented, and heat is prevented from being conducted in the lateral direction by the material in contact with the link. However, the pulse must be long enough to heat the link material to the end.

Remove heat stress and link

Through the absorption of laser energy, the target metal link is heated and tends to expand. However, it contains a substance that expands the oxide surrounding the link. Therefore, stress accumulates in the oxide. At some point, the pressure of the expanding metal exceeds the yield point of the oxide, causing the oxide to crack and the metal link to explode into fine particle vapors. The main cracking point of the metal link is the highest stress point, which is the edge of the link at the top and bottom, as shown in FIG.

If the oxide on the link is somewhat thinner, cracking of the oxide will occur only at the top of the link, and the link will be cleanly removed as shown in FIG. 1A. However, if the oxide is somewhat thick, cracking can occur at the bottom as well as at the top of the link, and cracks will propagate down the substrate, as shown in FIG. This is a very undesirable situation.

The Q-switching laser system can be modified to provide short pulses of various shapes. Conventional conventional lasers that produce high peak power, short pulse lasers are standard Q-switching lasers. These lasers generate temporal pulses with moderate pulse rise times. This temporal shape can be changed using a Pockels Cell pulse slicer that switches out the sections of the laser beam. U.S. Patent 4,483,005 (i.e., the '005 patent), invented by the applicant of the present invention and having the same assignee, discloses various methods of affecting (i.e., reducing) the laser beam pulse width. As described in the '005 patent, laser pulses can be formed to truncate the energy out of the central lobe to provide a "non-Gaussian" shaped beam. It should be noted that if a relatively wide Q-switching waveform is converted to a narrow and uniform shape, only a small portion of the pulse energy will be used. For example, the pulse energy is reduced by about 65% by providing a narrow pulse with truncation sharp rise time and flatness within 10%.

Similarly, U.S. Patent No. 4,114,018 (the '018 patent) describes the temporal pulse formation for generating a square pulse. Figure 7 shows the time interval for a very flat laser power output. In the patented method of '018, it is necessary to remove the time segment of the beam intensity to generate the desired pulse.

A preferred improvement over the prior art is to provide an efficient method for generating short pulses with a high collapse tail with high energy enclosure within the pulse duration. To achieve this, a laser technique for generating a pulse shape that is different from the pulse shape of the Q-switching pulse envelope is desirable. Such pulses have a fast rise time, uniform energy in the central lobe, and fast decay.

Fast rise times, high power density pulses, such as those provided by lasers other than standard Q-switched Nd: YAG, will best accomplish this task.

This advantage is realized in a preferred manner in a system using laser technology that is significantly different from conventional Q-switched, solid state diodes or lamp pumped YAG technologies.

A method and system for generating a pulse having a shape that is different from a standard Q-switching pulse with a faster rise time, a relatively uniform, higher energy concentration in the central lobe, and a faster fall time Is required to improve the prior art.

Applicants have determined that improved results in metal link blowing applications can be obtained. For example, pulse shapes that are non-Gaussian and substantially rectangular are particularly desirable for metal link processing in the presence of an overlying insulator. Applicants' results show that a fast rise time of 1 ns, preferably about 0.5 ns, provides a thermal shock to the top oxide layer to facilitate the link blowing procedure. In addition, the reflectivity decreases in the case of short pulses rising fast at high power densities. A pulse duration of about 5 ns with a substantially uniform pulse shape allows more energy to couple to the link, thereby reducing the energy required to remove the link. A sharp fall time of about 2 ns is important to rule out the possibility of substrate damage. Furthermore, the advantage of timely, nearly square power density pulses is that they are highest when power density is needed and off when not needed.

A short, fast rising pulse will cause the top of the link to melt and expand before the heat can diffuse down to the bottom of the link. Hence, stress is accumulated on the top of the link, which stresses prevent cracking down to the substrate and promote cracking of the upper layer.

It is an object of the present invention to provide a compact gain switching laser system with the ability to generate rise time pulses of sub-nanoseconds with short duration and abrupt fall times of several nanoseconds. The state-of-the-art high-speed pulse system includes a gain switching technique in which a low power semiconductor seed laser is rapidly and directly modulated to produce a controlled pulse shape which is then used as a high power laser diode or diode array such as a cladding pumped fiber optic system with an array of light emitting diodes (LEDs). Such a laser system is described in U.S. Patent No. 5,694,408 and PCT Application No. PCT / US98 / 42050, which is incorporated herein by reference in its entirety, for example, in the "building blocks" section of some of the ultrafast chirped pulse amplification systems described in U.S. Patent No. 5,400,350. )"to be.

One general purpose of the present invention is to improve the prior art laser processing methods and systems, particularly to improve optical and / or thermal properties of regions surrounding the target material when they are substantially different.

One general object of the present invention is to provide a method and apparatus for micromachining and laser material processing applications, such as laser removal, trimming, drilling, marking, and the like of a link or other interconnects in a semiconductor memory, And a laser pulse forming function for micromachining. A predetermined waveform shape is generated from the gain switching laser, which differs from the waveform of a standard Q-switching system.

It is an object of the present invention to provide an improvement and margin for semiconductor processing, e.g., 16-256 megabit semiconductor repair, resulting in subsequent device damage It is to clean the microstructure without any danger.

One purpose of the present invention is to provide a pulse waveform rise time of up to several hundreds of picoseconds, with a pulse duration of typically less than about 10 nanoseconds and having a sharp pulse decay, Thereby minimizing damage to the surrounding area due to thermal shock and diffusion.

It is an object of the present invention to achieve a high power density in a workpiece in a very short time with high power, fast rise time pulses at any wavelength suitable for the laser ablation process, To prevent damage to the structure.

One object of the present invention is to process a target site with a single laser processing pulse with a sufficiently fast rise time and sufficient power density to reduce the reflectivity of a metal target structure, such as a single metal link on a semiconductor memory, To provide efficient coupling. Rapidly rising laser pulses have a pulse duration sufficient to efficiently heat and vaporize each metallic target structure and have a relatively uniform power density during the removal duration but also a sharp pulse fall time after the target material is vaporized Prevent damage to the surrounding and underlying structures.

It is an object of the present invention to provide a method and apparatus for semiconductor metal link blowing that is superior in semiconductor metal link-blowing application, as compared to a system utilizing a standard Q-switching laser, typically a laser with a pulse rise time of several nanoseconds and represented by a Q- Performance. A laser pulse having a substantially rectangular pulse shape and having a pulse duration in the range of about 2 to 10 nanoseconds and a rise time of about 1 ns and preferably 0.4 ns is generated. In addition, the pulse collapse at the time of switch-off is abrupt, leaving only a very small fraction of the pulse energy after a predetermined pulse duration, and the pulse "tail" is rapidly attenuated to prevent the possibility of damaging the underlying substrate or other non- The level becomes sufficiently low. A comparison of these pulses is shown in FIG.

One object of the present invention is to enlarge the processing window of the semiconductor laser removal process to quickly and efficiently remove microstructures surrounded by materials with different optical and thermal properties. Such structures are typically arranged in a manner such that the width and spacing between the structures are less than about 1 micron and overlap in the depth direction. By applying a short laser pulse, the target material is cleanly removed, while damage to the surrounding material due to heat dissipation in the lateral direction or damage to the underlying substrate under the target material is prevented.

One object of the present invention is to controllably machine materials with substantially homogeneous optical and thermal properties by applying short pulses with high energy densities, wherein the pulse duration is such that the fluence threshold is < RTI ID = 0.0 > It is several nanoseconds in the material processing range which is roughly proportional to the square root of the pulse width.

In order to accomplish the foregoing and other objects of the invention, there is provided a method of treating a target material having a specified size in a microscopic region without causing undesirable changes in the electrical or physical properties of the material around the target material An energy efficient laser based method is provided. The method includes generating a laser pulse train using a laser having a predetermined wavelength at a predetermined repetition rate, wherein each pulse of the pulse train has a predetermined shape. The method also includes optically amplifying the pulse train to obtain an amplified pulse train such that the predetermined shape of the pulse is not significantly changed. Each amplified pulse has a substantially quadratic temporal power density distribution, a sharp rise time, a pulse duration and a fall time. The method also includes transferring and focusing at least a portion of the amplified pulse train to a spot of target material, wherein the ramp-up time is sufficient to efficiently couple the laser energy to the target material Fast, the pulse duration is sufficient to treat the target material, and the fall time is fast enough to prevent undesirable changes to the material around the target material.

The target material may include microstructures, such as conductive lines or links, where the link is a common circuit element of a redundant semiconductor memory. The conductive line may be a metal line, wherein the pulse duration is sufficient to effectively heat and vaporize the metal line, a particular portion of the metal line.

The target material may be part of a semiconductor device, such as, for example, a semiconductor memory having 16-256 megabits.

At least a portion of the material around the target material may be a substrate such as a semiconductor substrate.

The target material may be part of a microelectronic device.

The substantially quadratic temporal power density distribution is sufficient to substantially completely eliminate the target material.

The rise time is preferably 1 nanosecond or less, more preferably 0.5 nanosecond or less.

The pulse duration is preferably 10 nanoseconds or less, more preferably 5 nanoseconds or less.

The falling time is preferably 2 nanoseconds or less.

Typically, a single amplification pulse is sufficient to process the target material.

The target material may be reflective to the amplified pulse, wherein the power density of the amplified pulse is high enough to reduce the reflectivity of the target material to the amplified pulse and effectively connect the laser energy to the target material.

Each amplified pulse preferably has a very uniform power density distribution over the pulse duration.

Each pulse preferably has a uniform temporal power density distribution within 10 percent of the pulse duration.

Materials around the target material may have optical properties including thermal diffusivity and absorption and polarization sensitivity that are different from the corresponding properties of the target material.

Preferably, the repetition rate is at least 1000 pulses per second, and each amplified pulse has an energy of at least 0.1 to 3 microns.

The step of optically amplifying preferably provides a gain of at least 20 db.

Also, both the rise time and the fall time are less than half of the pulse duration, and the peak power of each amplified pulse is preferably substantially constant between the rise and fall times.

Each of the amplified pulses has a tail and the method further comprises attenuating the laser energy in the tail of the amplified pulse to reduce the fall time of the amplified pulse while substantially maintaining a power amount of the pulse .

To accomplish the foregoing and other objects of the invention, there is also provided a method of treating a target material having a specified size in a microscopic region without undesirable changes in the electrical or physical properties of the material around the target material An energy efficient system is provided. The system includes a controller for generating a process control signal and a signal generator for generating a drive waveform modulated based on the process control signal. The waveform has a rise time of sub-nanosecond. The system also includes a gain switching pulse seed laser having a predetermined wavelength to generate a laser pulse train at a predetermined repetition rate. The drive waveform pumps the laser so that each pulse of the pulse train has a predetermined shape. Further, the system includes a laser amplifier that optically amplifies the pulse train to obtain an amplified pulse train such that the predetermined shape of the pulse is not significantly changed. Each amplified pulse has a substantially rectangular temporal power density distribution, a steep rise time, a pulse duration and a fall time. The system further includes a beam delivery and focusing subsystem for focusing and transferring at least a portion of the amplified pulse train to a target material. The rise time is fast enough to efficiently connect the laser energy to the target material, the pulse duration is sufficient to treat the target material, and the fall time is fast enough to prevent undesirable changes in the material around the target material.

The laser amplifier preferably includes an optical fiber and a pump, such as a laser diode, that pumps the optical fiber, wherein the pump is distinct from the seed laser.

The laser diode pump source can also be diode switched by switching to an "off" state during an extended period in which the gain is switched (pulsed and directly modulated) and no laser treatment occurs.

The seed laser preferably includes a laser diode.

The system may include an attenuator that attenuates the laser energy at the tail of the amplified pulse to reduce the fall time of the amplified pulse while substantially maintaining the energy amount of the pulse.

The pulse duration can be selected as a function of the specified target material size. The specified target material size can be below the laser wavelength. A preferred laser is a high speed semiconductor laser having a wavelength of about 2 탆 or less. Fiber materials and semiconductor laser diode technology can provide further developed materials for operation in the visible range as well as at longer infrared wavelengths.

A seed laser diode is a single frequency (single mode) laser or a multimode laser that utilizes a distributed Bragg reflector (DBR), distributed feedback (DFB), an external cavity design, ) Diode laser.

The spot size typically has a size in the range of about 1 [mu] m - 4 [mu] m.

The density of the memory may be at least 16-256 megabits.

The semiconductor device may be a microelectromechanical device.

Preferably, the laser energy attenuated in the pulse tail is attenuated by at least 10 dB within 1.5 times the pulse duration.

In order to accomplish the foregoing and other objects of the invention, there is also provided a method of fabricating a semiconductor device, comprising the steps of: providing at least one passivation layer on at least one passivation layer, without undesirably altering the electrical or physical properties of the at least one passivation layer, An energy efficient laser based method is provided for ablating a metal link with a built-in specified size. The method includes generating a laser pulse train that utilizes a laser having a predetermined wavelength at a predetermined repetition rate. Each pulse of the pulse train has a predetermined shape. The method also includes optically amplifying the pulse train without a significant change in the predetermined shape of the pulse to obtain an amplified pulse train. Each amplified pulse has a substantially quadratic temporal power density distribution, a sharp rise time, a pulse duration and a fall time. The method also includes transmitting at least a portion of the amplified pulse train to a spot of the metal link for focusing. The rise time is fast enough to efficiently connect the laser energy to the metal link. The pulse duration is sufficient to remove the metal link and the fall time is fast enough to prevent undesirable changes in at least one passivation layer around the metal link.

In order to accomplish the foregoing and other objects of the invention, there is also provided, in accordance with the present invention, a method of fabricating a passivation layer that is embedded in at least one passivation layer without undesirably altering the electrical or physical properties of the at least one passivation layer around the metal link. An energy efficient system for ablating metal links of a certain size is provided. The system includes a controller for generating a process control signal and a signal generator for generating a drive waveform modulated based on the process control signal. The waveform has a rise time of sub-nanosecond. The system also includes a gain switching pulse seed laser having a predetermined wavelength that generates a laser pulse train at a predetermined repetition rate. The drive waveform pumps the laser so that each pulse of the pulse train has a predetermined shape. Further, the system includes a laser amplifier for optically amplifying the pulse train without a significant change in the predetermined shape of the pulse to obtain an amplified pulse train. Each amplified pulse has a substantially rectangular temporal power density distribution, a steep rise time, a pulse duration and a fall time. The system further includes a beam delivery and focusing subsystem for focusing at least a portion of the amplified pulse train to a spot of the metal link. The rise time is fast enough to efficiently couple the laser energy to the metal link. The pulse duration is sufficient to remove the metal link and the fall time is fast enough to prevent undesirable changes in at least one passivation layer around the metal link.

The metal link may be embedded in the upper passivation layer thereon and in the lower passivation layer thereunder. The pulse duration is sufficient to cause cracking in the upper passivation layer, but not enough to cause cracking in the lower passivation layer.

In order to accomplish the foregoing and other objects of the present invention, there is provided a method for ablating a target material while preventing damage to surrounding materials using a laser having a wavelength suitable for treating the laser material. The method includes modulating a laser beam to produce a predetermined gain switching pulse and focusing the laser beam onto a target material region. The predetermined gain switching pulse shape includes a rise time of the laser pulse that is fast enough to efficiently couple the laser energy to the target structure and includes a pulse duration of sufficient length to efficiently heat and vaporize the target material, Lt; RTI ID = 0.0 > collapse < / RTI >

In order to accomplish the foregoing and other objects of the invention, there is also provided a system for ablating a target material while avoiding damage to surrounding materials using a laser having a wavelength suitable for laser treatment. The system includes a laser source, a component for modulating the laser source to generate a laser pulse having a predetermined gain switching pulse shape, and an optical component for focusing the laser beam onto the target material region. The predetermined pulse shape is used to determine the optical rise time of the laser pulse fast enough to efficiently couple the laser energy to the target material and the pulse duration and the damage of the target material surrounding structure to a length sufficient to efficiently heat and vaporize the target material And has a pulse collapse time fast enough to prevent it.

In one preferred embodiment of the invention, the gain switching pulse shape comprises a fast rise time pulse, the top of which is substantially flat and has a fast pulse fall time. A "seed" laser diode is directly modulated to generate a predetermined pulse shape. The amplification by the fiber laser amplifier increases the optical power and outputs a power level sufficient for laser processing. Whereby the gain switching pulses at the fiber laser amplifier output are focused on the target area.

In one configuration of the invention, a "seed" diode is directly modulated to produce a predetermined gain switching square pulse and a low distortion amplification using a fiber laser amplifier to provide an output pulse level sufficient for material processing May be desirable.

In another configuration, the pulse temporal power distribution of the directly modulated seed diode can be used for non-uniformity or distortion of the fiber amplifier or other components, e.g., a "smooth" Lt; / RTI > Thus, the laser processing pulses focused on the target material area will have the desired shape: fast rise time, relatively flatness over the pulse duration, and fast collapse.

In one configuration of the present invention, the laser energy remaining at the output of the laser processing system when the "seed" laser pulse is terminated is used to prevent heating of the delicate structure that is not designated as a target material after processing is complete It may be useful to improve the performance of the laser processing system by providing a "pulse slicing" module. The "pulse slicing" technique is useful for attenuating the tail of a modified pulse or a standard Q-switching pulse. This is illustrated in Figures 4a and 4b, where the logarithmic scale is provided on the vertical axis of Figure 4b.

It is desirable to perform a laser treatment operation, particularly a metal link blowing, at a pulse rate of at least 1 KHz (1000 pulses per second) with a laser pulse energy of at least 0.1 microns per pulse, wherein 0.1 microns of the fiber amplifier Output, where the fiber optic amplifier gain is at least 20DB (1000: 1).

In one configuration of the present invention, a laser pulse having a substantially constant peak power between rising and falling times is formed with rise and fall times shorter than about half of the pulse duration.

In one configuration of the invention, it is possible to generate a series of closely spaced short pulses, which, when combined, produce the desired pulse shape, as shown in Figures 3A and 3B.

In one configuration of the system using the present invention, the beam position adjustment used to position the focused laser beam for material processing, to actuate the laser with a pulse repetition rate that exceeds the material processing rate, It may be useful to use a computer controlled optical switch that is operatively connected to the system.

The foregoing and other objects, features, and advantages of the present invention will become readily apparent from the following detailed description of the best mode for carrying out the invention when taken in conjunction with the accompanying drawings.

According to the present invention, due to the appropriate choice of wavelengths and pulse duration limitations, the silicon substrate is kept relatively cool with the corresponding square pulse with fast decay. The laser wavelength in this example is slightly smaller than the room temperature absorption edge of silicon (about 1.1 mu m). It should be noted that the results presented here did not show any substrate damage, but improved margins are possible if desired. For example, a Raman shifter can be used to move the output wavelength beyond the absorption edge. Alternatively, another diode laser wavelength may be potentially commercialized for MOPA configuration. Such wavelength selection and shifting techniques may be useful in other laser processing and micromachining applications. In any case, by limiting the heating, it can be ensured that the silicon moves its absorption edge into the infrared and does not reach a heat-dissipating state where silicon damage can occur.

A specific embodiment of a MOPA configuration for fast pulse generation to cleanly blow metal links is used as an example of pulse formation and is provided for illustrative purposes rather than limitation. Other applications, such as micromachining, marking, scribing, etc., will also benefit from precise and fast pulse control. For example, the seed diode may be easily modulated with a "sawtooth" waveform or other non-Q-switching waveform for the purpose of making or removing a specific feature on or on any surface. Similarly, because of the fast response of the laser diode, it is possible to generate a short pulse sequence of variable widths in rapid succession. Those skilled in the art of laser processing will recognize a wide variety of applications of the laser systems described herein. The scope of the invention is indicated by the following claims, and should not be construed in any way.

Laser processing system architecture

It will be appreciated by those skilled in the art that the following embodiments can be applied to various applications in micromachining and laser processing, with appropriate adjustment of parameters such as laser power, energy density, spot size, wavelength, pulse width, polarization and repetition rate There will be. A specific application for metal link blowing is illustrated for illustrative purposes.

In the preferred embodiment of FIG. 7, the seed laser 10 and the fiber amplifier are mounted on a moving platform 20 and a stable platform attached to the workpiece. It is very important that the beam be positioned to an accuracy of 3/10 microns or less in removing the link. Due to the continuous transfer required to achieve high processing speed, the timing of the laser pulse corresponding to the relative position of the target and the optical system is important.

The seed laser 10 is externally controlled by a computer 33 and a signal generator 11 and the seed laser 10 includes high numerical aperture optics and includes a beam deflector deflector, for example, a focusing subsystem 12, which may include a galvanometer mirror controlled by a scanner control via a computer 33. [ The computer 33 is also operatively connected to the signal generator 11 and the mobile system 20 for the subsystem to suitably accommodate the pulse generation time. The laser beam must be precisely controlled to produce a sharply focused beam and to have the spot size in the range of about 1.5-4 microns at the correcting positions of X, Y, and Z. [

As such, those skilled in the field of beam positioning and focusing will appreciate the importance of the modified optics to provide finite capabilities and near diffraction and to provide precise movement control of the laser head or target substrate will be. Depending on the specific laser processing application requirements, it may be useful to provide an optical system with a relatively narrow field of view for diffraction limited focusing and precision X, Y motion stages for beam positioning . Furthermore, various mirror movement combinations are feasible for rapid deflection combined with translation stages.

A substrate position adjustment mechanism 34, such as a step and repeat table, may also be used to move the wafer 22 into position and process each memory die 24. One of ordinary skill in the art of beam scanning will appreciate the advantages of a mirror-based beam deflection system, but as mentioned above, alternatives to other substrate position adjustment mechanisms, such as X, Y translation stages for movement of the substrate and / Lt; / RTI > For example, the substrate position adjustment mechanism 34 may include a very precise X, Y, Z position adjustment mechanism operating at a limited range of movement (much less than one micron). The moving system may be used to translate laser processing optical system components, including lasers, fiber amplifiers, and focusing subsystems, in a more coarse manner. Details of a preferred mobile system are further disclosed in U. S. Patent Application Serial No. 09 / 156,895, filed on September 18, 1998, entitled " High Speed Precision Positioning Apparatus "

A system optical switch 13 in the form of an additional acousto-optic attenuator or pockels cell is arranged in the laser output beam through the laser cavity. Under the control of the computer 33, the system optical switch 13 prevents the beam from reaching the focusing system unless required, and when a processing beam is required, the power of the laser beam can be controlled to a desired power level . During the vaporization procedure, the power level may be reduced to 10 percent of the total laser power, depending on the operating parameters of the system and process. During the alignment procedure in which the laser output beam is aligned with the target structure prior to the vaporization procedure, the power level may be about 0.1 percent of the total laser power. Although the delay of the Fouques elements is fairly small, acousto-optic devices are generally preferred because of their ease of use.

During operation, the position of the wafer 22 (or target or substrate) is controlled by the computer 33. Typically, relative movement occurs at a substantially constant speed across the memory die 24 on the silicon wafer 22, but step and repeat movement of the wafer is also possible. The seed laser 10 is controlled by a timing signal based on a timing signal controlling the position adjustment mechanism. The seed laser 10 typically operates at a constant repetition rate and is synchronized with the position adjustment mechanism by the system optical switch 13.

In the system block diagram of Figure 7, the laser beam is shown as being focused on the wafer 22. In the enlarged view of FIG. 9, the laser beam is shown as being focused on the link element 25 of the memory die 24.

Spot size requirements are becoming more and more challenging in handling fine link structures. The spot size requirement typically has a diameter of 1.5-4 microns, peak power occurs at the center of the spot, is well adapted to the Gaussian distribution, and lower power is generated at the edge. As the diffraction limit is reached, an excellent beam quality of about 1.1 times or more typical beam quality or "m-squared factor" is required. Such drainage diffraction limited quality criteria are well known to those skilled in the art of laser beam analysis. Low side lobes are also desirable to prevent optical crosstalk and undesirable illumination on the components outside the target material area.

The link element 25 is somewhat smaller than the spot size, thus requiring precise positioning and good spot quality. The link may be, for example, 1 micron in width and about 1/3 micron in thickness. In the case illustrated here, the link is made of metal, and the side size (width) and thickness are less than the laser wavelength.

Laser systems - preferred cases

In a preferred embodiment, the laser system of FIG. 5 utilizes a master oscillator, power amplifier (MOPA) configuration. The system generates a laser pulse that causes the amplifier to generate a high power short rise time pulse. Seed lasers are important in generating fast rise times, short pulses at very low energy levels. The system requires a laser amplifier that generates enough energy to process the material. High-speed infrared laser diodes and fiber laser amplifiers with output wavelengths suitable for laser processing applications are desirable. Such a system can be considered as a laser which produces laser pulses of the desired shape and speed, i.e. a fast rise time pulse, a square at the top, and a fast fall time, as shown in the lower part of FIG. Such a pulse shape then provides the desired laser-material interaction result, which reduces metal reflectivity, less energy diffusion into the device, and cracks the top oxide without damaging the bottom oxide.

The MOPA configuration is very new, and the pulse version is considered to be the latest technology. A laser diode with sub-nanosecond rise time in response to a modulated drive waveform is the beginning of a fiber laser MOPA configuration with a laser diode as a gain element. Laser diodes generally have a plurality of longitudinal modes and the subsystem may be set for single mode operation or otherwise tuned to bulk components at the output or alternatively may be integrated And can be tuned to integrated fiber gratings.

For example, it is a feasible configuration that the 'Littman-Metcalf' grid configuration described in the product literature of New Focus Inc. is in an external co-configuration. Figure 6b shows a schematic of a single frequency laser with external cavity tuning and also includes an optical fiber that is cladding pumped by a diode laser pump.

Other diode laser alternatives include a distributed feedback laser (DFB) and a dispersed Bragg laser (DBL), which has an integrated grating and a waveguide structure and, in some cases, ), And the grating filter can be controlled independently. In FIG. 6A, a DBL configuration including a coupler 50 can be seen. It provides flexible mode selection and tuning. The laser frequency can be dynamically selected in a number of configurations by grating and / or mirroring of an external cavity, or, alternatively, by adjustment of bulk components such as a fixed wavelength or a selected mode. The wavelength at which the center wavelength of the diode can be selected is generally about 1.3-1.5 탆 or more at less than 1 탆, and the wavelength of about 1.3-1.5 탆 or more corresponds to that used for fiber optic communication.

Whatever the case, for a laser wavelength chosen for material handling, a critical factor for the purposes of the present invention is the rise time and pulse shape of the "seed" laser diode. Furthermore, one of the considerations of the present invention is that the seed laser wavelength can be used in a spectral band with a low sensitivity to high gain and small wavelength variations in fiber optic amplifiers-that is, to " flat "reaction area that maintains a-to-pulse power output. For ytterbium-doped fibers, the gain is high at moderately broad wavelength bands around 1.1 m, the absorption edge of silicon. Future developments in integrated fiber components or materials can extend the useful wavelength range to provide more flexibility in addressing target material properties, seed laser wavelengths, and fiber emission spectra. For example, in Photonics Spectra , August 1997, page 92, results are reported for the development of state-of-the-art fiber lasers in the wavelength range of 1.1 袖 m to 1.7 袖 m.

The above-mentioned '759 patent describes a specific case in which the operation of a Raman shifter uses a short pulse Q-switching system. If desired, the device may be placed on the output side of the fiber system to shift the output wavelength, for example, to a desired area for improving absorption contrast. Recognizing the importance of pulse width and small spot size requirements in the process as described in the aforementioned '759 patent, the typical operation of a preferred system for metal link processing is in the range of about 1.06 μm or more, for example, Lt; / RTI > wavelength.

The output of the seed laser must be amplified for laser material processing. A preferred fiber optic laser amplifier will provide a gain of about 30 db. The seed laser output is connected directly to the core of the fiber laser or is connected with a bulk optic that splits the beam for fiber delivery. While both techniques are routinely practiced by those skilled in the art of ultrafast laser using chirped pulse amplification, the system of this preferred embodiment is much less complex than such a high speed system. In the system of the present invention, the seed pulse is amplified, and no optical mechanism for pulse stretching and compression is required. The fibers used in the amplification system are cladding pumped with diode lasers of substantially different wavelengths, e. G. 980 nm, from the seed lasers, and the dichroic mirror in the bulk of the bulk optical system arrangement, Lt; RTI ID = 0.0 > isolation. ≪ / RTI > From the viewpoint of cost, size, and ease of alignment, it is preferable to use a coupling arrangement in which the seed laser is directly connected to the fiber amplifier. Pump lasers emit high power diode energy into the cladding structure of rare earth ytterbium (Yb) -doped fibers at a wavelength of, for example, 980 nm, using coupling techniques known to those skilled in the art of fiber laser system design .

Low distortion is an important characteristic of fiber amplifiers. The low distortion causes the output pulse shape to substantially correspond to the seed laser pulse shape or, if possible, further enhance the pulse edge or uniform power profile. The fiber optic gain medium produces the amplifier pulses of FIG. 5 that are transmitted to the optical system and focused on the target material.

Multiple fiber amplifiers may be cascaded if desired for additional gain, provided that the distortion is small. It may be desirable to provide an active optical switch or an optical isolator at the output of the intermediate stages to suppress instantaneous emission. Such techniques are known to those skilled in the art and are disclosed, for example, in U.S. Patent Nos. 5,400,350 and WO 98/92050.

In some cases it may be desirable to shorten the "tail" with a pulse slicer added to the laser subsystem to further improve the pulse shape. It may be in the form of an electro-optical modulator such as a forklith element or preferably a low-delay acoustic optical modulator. This technique can suppress the energy in the pulse tail to a negligible level whenever there is a risk of damage in a small multiple of the "pulse duration" of the processing pulse. For example, if energy is reduced by 20dB (100: 1) within 1.5 times the predetermined pulse duration, there will be no risk of substrate damage in metal link blowing applications in practice. More specifically, if a pulse duration of 8 ns is selected for a rectangular pulse shape in a metal link blowing application and the energy is reduced from 12 ns to 20 ns, the remaining energy is far below what can cause damage to the Si substrate, It is practically more than about 18ns after application of the laser pulse. In the preferred mode of operation, a low delay, high bandwidth pulse slicer will be started near the end of the amplifier pulse duration. The effects of amplifier distortion and the "turn on delay" of the modulator can be compensated to some extent by changing the shape of the seed diode laser waveform during the pulse duration. The temporal pulse shape in the focused beam is thus compensated and has the desired rectangular shape.

In addition, current fiber systems operate optimally at a pulse repetition rate of about 20 KHz, somewhat faster than the throughput rate. An output optical switch, for example a low delay acoustic optical modulator, is operatively connected to the computer by a driver to select a pulse for processing. In this way, the reliability of the fiber amplifier, and therefore of the processing system, is high. It will be appreciated by those skilled in the art that combining a pulse slicer and an output optical switch in a single module is advantageous from an economic point of view.

Laser Systems - Alternatives

Preferred seed laser and fiber amplifier systems have a number of advantages mentioned above. The current modulation of the laser diode through a suitable driver can directly produce the desired gain switching pulse shape, which is amplified with low distortion by a fiber laser amplifier. This method is considered to be the best and most efficient approach to the practice of the present invention. However, those skilled in the art of laser pulse generation and formation will recognize that other less efficient approaches may be used. For example, it is possible to change the Q-switching system beyond 4,483,005 to obtain a relatively flat pulse by using various control functions to drive a Forceless device or an optical switch, provided that the modulator reaction time is fast enough. Do. Modern techniques for implementing pulse widths include, for example, the use of modified output couplers to replace existing glass of Nd: YAG Q-switching lasers with bulk or crystal GaAs. In Q-switching of Nd: YAG lasers with GaAs output couplers, Q-switching pulses with durations of from a few picoseconds to a few nanoseconds are described in "OPTICAL ENGINEERING, 38 (11), 1785-88, November 1999 ".

Laser processing steps and results

The metal link element 25 is supported on the silicon substrate 30 by a silicon dioxide insulation layer 32, which may be, for example, 0.3-0.5 microns thick. Silicon dioxide extends over the link, often an additional silicon nitride insulating layer is present on the SiO ₂ layer. In a link-blowing technique, a laser beam is irradiated on each link to heat the link to its melting point. During heating, the metal is prevented from vaporizing due to the accepting effect of the overlying passivation layers. During the duration of the short pulse, the laser beam gradually heats the metal until the metal expands to such an extent that the insulation ruptures. At this point, the molten material is instantaneously vaporized under too high pressure and blown cleanly through the burst hole.

As described in the aforementioned '759 patent, with very small spot sizes used in small metal links, it can be assumed that the beam spreads essentially in an exponential gradient by heat conduction from the area hitting the target material have. Even though the peak beam power is so high that sufficient energy is delivered to evaporate the link in 8 nanoseconds, preferably substantially less pulses, the heat transfer of the conductive component is very thin, So that the temperature rise in silicon due to conduction and the temperature rise in silicon due to beam absorption can be kept below the temperature threshold at which cumulative unacceptable silicon damage occurs.

Further, the aforementioned '759 patent describes several important aspects regarding the heat transfer characteristics of the link and adjacent structures. The thermal model is expected to have a narrow pulse width, e.g., 3-10 ns, to prevent thermal conduction to, and damage to, the Si substrate for a representative size, which is expected to depend on the thickness of the target material . However, it is extremely important to recognize that other structures neighboring the link can also affect the quality of the laser processing results, as the following experimental results indicate.

The advantages of gain switching square pulse shapes were verified both in experimental results and in computer simulations (finite element analysis). The specifications for the laser used in link-blowing are:

▶ Laser wavelength 1.083 micron

▶ Maximum laser energy 10 micro lines

▶ Pulse width 7ns (FWHM, square pulse)

▶ Repeat rate 10KHz (70KHz laser speed)

Spatial profiles Gaussian, TEM-OO, M ² = 1.02 (time diffraction limited)

▶ Polarization is not polarized

▶ Pulse Rise Time ~ 0.5ns

The selected lasers were a tribe, cladding pumped fiber laser in a MOPA configuration using a 980 nm pump diode and 7 micron diameter single mode fiber.

Experimental results performed on recent memory devices with the specified laser showed excellent performance when compared with a standard Q-switched laser system. This result led us to conclude that short and fast rising pulses of MOPA lasers were the cause of excellent performance. As described above, there are three reasons for this:

1. The 1.083 wavelength is long enough to prevent substrate damage - about 10 times less absorption at 1.083μm compared to the 1.047μm wavelength.

2. Rapidly rising pulses provide thermal shock to the top oxide layer, which facilitates link removal.

3. The high power density of fast rising pulses reduces link reflectivity, which allows efficient energy coupling.

These characteristics provide significant differences from the interactions observed in the Q-switching system. Furthermore, computer finite element models were used to simulate the effects of fast rising pulses on various material thicknesses and link sizes. The results independently confirmed that the link-blowing results were improved with the use of a steep rise time pulse with a roughly rectangular distribution. The results of the computer model generated by the author Bernstein of reference number 3 are shown in Figures 11A and 11B. The following Tables A and B are associated with the graphs of Figures 11A and 11B, respectively:

[Table A]
Model 1 @ 0.7 uJ Square pulse Slow rising pulse First crack 929K @ 1.88ns 978K @ 2.40ns The second crack 1180K @ 2.93ns 1380K @ 3.45ns The third crack 1400K @ 2.05ns no Fourth crack 1520K @ 4.73ns no

delete

Al thickness: 0.8 m Al width: 0.8 m

SiO ₂ : 0.1 μm Si ₃ N ₄ : 0.4 μm

Laser energy: 0.7uJ

[Table B]

Model 2 @ 0.7 uJ Square pulse Slow rising pulse First crack 974K @ 2.03ns 1050K @ 2.55ns The second crack no no The third crack no no Fourth crack no no

Al thickness: 0.8 m Al width: 0.8 m

SiO ₂ : 0.6 μm Si ₃ N ₄ : 0.6 μm

Laser energy: 0.7uJ

Stress and temperature history clearly indicate the importance of fast rising pulses of sub-nanosecond rise time. It is also known that, in the presence of significant pulse energy, Si may be damaged after ablation is complete and several nanoseconds after, for example, 15 ns. A fast fall time with high extinction is also important.

1A schematically shows stress cracking only in the uppermost surface layer of the semiconductor due to the expansion of the vaporized metal;

Fig. 1b schematically shows that a stress crack occurs in the uppermost and lowermost surface layers of the semiconductor due to the expansion of the vaporized metal;

Figure lc shows a conventional prior art laser pulse, which resembles the Gaussian shape, i.e. the coupled form of Gaussian and exponential tails, referred to as the "Q-switched pulse envelope ";

Figure 2 shows a preferred pulse shape of the present invention for processing a metal link when compared to a Q-switched version of the same total energy;

FIGS. 3A and 3B illustrate a method of making modified pulses by combining two short pulses closely spaced to produce a modified pulse; FIG.

Figures 4A and 4B show the results of "pulse slicing" for improving the pulse energy energy enclosure of a typical pulse shape;

5 is a general block diagram of a preferred laser system for laser material processing;

6A is a schematic diagram of one type of MOPA laser system with a Bragg laser dispersed as a semiconductor seed laser, which is a schematic diagram of a fiber optic amplifier that produces a single mode laser and a desired pulse shape;

Figure 6b is a schematic diagram of a single frequency laser with external cavity tuning and a fiber optic amplifier;

Figure 7 is a block schematic diagram of another laser system according to the present invention including a preferred system optical switch and an optional shifter;

Figure 8 is a graphical representation of the temperature at the interface between the silicon substrate and the silicon dioxide layer of Figure 9 as a function of the thickness of the silicon dioxide layer;

9 is a perspective view of a state where a link of a memory is present in the substrate;

FIG. 10 is a drawing of a Gaussian laser beam focused on a small spot on a focal plane that accommodates a metal link represented by emphasizing the fine size of the link as compared to a diffraction-limited beam waist;

11A and 11B show the results of a computer finite element analysis simulation shown in the graph for Q-switching pulses and square pulses, where the time history of temperature and stress is used for metal link processing Graph.

Claims (62)

A system for processing a target material comprising a target structure positioned adjacent to and having a microscopic dimension of at least one material having different thermal or optical properties, A non-q-switched pulsed seed laser for generating a shaped pulse or a short pulse sequence for application to the target structure; A laser amplifier for optically amplifying the shape pulse or short pulse sequence to generate an amplified shape pulse or an amplified short pulse sequence; An optical switch for receiving the amplified shape pulse or the amplified short pulse sequence and controllably selecting the amplified shape pulse or the amplified short pulse sequence; A beam delivery and focusing system for transmitting and focusing the selected amplified shape pulse or the selected amplified short pulse sequence to a spot of the target structure; A position adjustment system for positioning the target material relative to the beam delivery and focusing system; And A controller coupled to the laser, the optical switch, and the position adjustment system, the controller being configured to adjust the timing and delivery of the selected amplified shape pulse or the selected amplified short pulse sequence to the target material ) The target structure is processed. The method according to claim 1, Wherein the amplified shape pulse has a non-q-switched waveform with a rise time of less than 1 nanosecond. The method according to claim 1, Wherein the optical switch comprises an acousto-optic modulator or an electro-optic modulator. delete delete The method according to claim 1, And wherein the seed laser is gain switched. The method according to claim 1, Wherein the at least one shot pulse of the sequence comprises a pulse having a pulse width less than 18 ns. The method according to claim 1, Wherein the laser amplifier amplifies at a low distortion and without significantly changing the shape of the amplified shape pulse or the shape of the pulse of the amplified sequence. The method according to claim 1, Wherein the amplified shape pulse is a substantially square shape pulse and has a pulse power uniformity within 10% for a relatively flat portion of the substantially rectangular shape pulse. Based systems. The method according to claim 1, Lt; RTI ID = 0.0 > 1, < / RTI > wherein the seed laser comprises a semiconductor diode laser. 11. The method of claim 10, Wherein the diode laser comprises a DBR (distributed Bragg reflector) laser or a DFB (distributed feedback) laser for externally controlling at least one gain, wavelength, and phase. Based system. 11. The method of claim 10, Wherein the optical output of the semiconductor diode is directly modulated. The method according to claim 1, Wherein the position control system is controlled by a controller for providing continuous movement of the target material to the beam delivery and focusing system. delete The method according to claim 1, Wherein the pulse width of the amplified shape pulse is less than 18 ns. 16. The method of claim 15, Wherein the pulse width is less than 10 ns. 16. The method of claim 15, Wherein the pulse width is in the range from several picoseconds to a few nanoseconds. 18. The method of claim 17, Wherein the seed laser comprises a GaAs saturated (saturable) absorbing material for reducing the pulse width of the seed laser to within a few nanoseconds from several picoseconds. system. The method according to claim 1, Wherein the power density in a spot on the target structure is greater than 10 ⁹ W / cm ² . The method according to claim 1, Wherein the energy of the amplified shape pulse is at least 0.1 microjoules at the output of the laser amplifier. The method according to claim 1, Characterized in that the amplified shape pulse comprises a sawtooth shaped pulse. The method according to claim 1, Wherein the target structure comprises a conductive link of a memory device. The method according to claim 1, Wherein the amplified short pulse sequence comprises a pulse of variable width. The method according to claim 1, A high power laser diode or diode array pumps and laser-gain-switches the laser amplifier. The method according to claim 1, Further comprising a pulse changing device for changing a portion of the output from the seed laser. The method according to claim 1, Further comprising a pulse changing device for changing a portion of an output from the laser amplifier. The method according to claim 1, ≪ / RTI > wherein the laser amplifier comprises a fiber optic amplifier. 28. The method of claim 27, Wherein the optical fiber amplifier comprises multiple amplifier stages. The method according to claim 1, Wherein the gain of the laser amplifier is greater than or equal to 20 dB. The method according to claim 1, Wherein the pulse of the short pulse sequence has a predetermined non-q-switched shape. The method according to claim 1, And a wavelength shifter for receiving an output of the laser amplifier and shifting the amplified shape pulse or the amplified pulse sequence to another wavelength. A laser-based method for processing a target material comprising a target structure positioned adjacent to and having a microscopic size, the at least one material having different thermal or optical properties, Generating a short pulse sequence or shaped pulse from a non-q-switched seed laser for application to the target structure; Optically amplifying the shape pulse or the pulse sequence to generate an amplified shape pulse or an amplified short pulse sequence; Controllably selecting the amplified shape pulse or the amplified short pulse sequence; Transmitting and focusing the selected amplified shape pulse or the selected amplified short pulse sequence to a spot of the target structure; Positioning the target material for the selected amplified shape pulse or the selected amplified short pulse sequence; And And coordinating delivery and timing of the selected amplified shape pulse to the target material or the selected amplified short pulse sequence to cause the target structure to be processed. 33. The method of claim 32, Wherein the step of optically amplifying generates an amplified shape pulse having a non-q-switched waveshape with a rise time of one nanosecond or less. 33. The method of claim 32, Wherein the controllably selecting comprises selecting the amplified shape pulse or the amplified pulse sequence with an optical switch comprising an acousto-optic modulator or an electro-optic modulator Based laser. delete delete 33. The method of claim 32, Further comprising the step of gain switching the seed laser. 33. The method of claim 32, And generating a short pulse of the short pulse sequence having a pulse width less than 18 ns. 33. The method of claim 32, Further comprising amplifying the short pulse sequence or the shape pulse with low distortion and without significantly changing the shape of the amplified shape pulse or the shape of the pulse of the amplified sequence. 33. The method of claim 32, Amplifying the short pulse sequence or the shape pulse to produce a substantially rectangular shape pulse and generating pulse power uniformity within 10% for a relatively flat portion of the substantially rectangular shape pulse, ≪ / RTI > 33. The method of claim 32, Wherein generating the short pulse sequence or shaped pulse comprises modulating a current input to the semiconductor diode laser. 33. The method of claim 32, The step of generating the short pulse sequence or shaped pulse may comprise generating a shape pulse from a distributed Bragg reflector (DBR) laser or a DFB (distributed feedback) laser for externally controlling at least one gain, And generating the short pulse sequence. 33. The method of claim 32, ≪ / RTI > further comprising gain switching the laser diode receiving the modulated drive waveform. 33. The method of claim 32, Wherein the positioning is controlled to provide continuous movement of the target material to the beam delivery and focusing system. delete 33. The method of claim 32, Generating a pulse width of said amplified shape pulse less than 18 ns. 47. The method of claim 46, Generating a pulse width of said amplified shape pulse less than 10 ns. 47. The method of claim 46, Generating a pulse width of the amplified shape pulse within a range of several picoseconds to a few nanoseconds. 33. The method of claim 32, The step of generating the short pulse sequence or the shaped pulse may comprise generating a GaAs saturated (saturated) pulse to reduce the pulse width to within a few nanoseconds from several picoseconds, Wherein the pulse width of the seed laser is reduced with a saturable absorbing material. 33. The method of claim 32, Further comprising generating a focused shape pulse of the amplified shape pulse having a power density of 10 ⁹ W / cm ² or more. 33. The method of claim 32, Further comprising generating an amplified feature pulse having an energy greater than or equal to 0.1 microjoules after said optically amplifying step. 33. The method of claim 32, Further comprising generating an amplified shape pulse having a sawtooth shape. 33. The method of claim 32, Wherein the propagating and focusing step comprises directing the amplified shape pulse or the amplified short pulse sequence to a conductive link of a memory device. 33. The method of claim 32, And generating an amplified short pulse sequence having a variable width pulse. 33. The method of claim 32, Further comprising gain switching the high power laser diode or diode array for pumping the laser amplifier. 33. The method of claim 32, Further comprising changing a portion of the short pulse sequence or the shape pulse. 33. The method of claim 32, Further comprising modifying a portion of the amplified shape pulse or the amplified short pulse sequence. 33. The method of claim 32, Wherein the step of optically amplifying comprises amplifying with an optical fiber amplifier. 59. The method of claim 58, Wherein amplifying with the optical fiber amplifier comprises amplifying to multiple amplifier stages. 33. The method of claim 32, Wherein the step of optically amplifying comprises amplifying with a gain of at least 20 dB. 33. The method of claim 32, Wherein generating the short pulse sequence comprises generating a short pulse sequence in which the pulse has a predetermined non-q-switching shape. 33. The method of claim 32, And shifting the wavelength of the amplified shape pulse or the wavelength of the amplified pulse sequence to another wavelength.

Images