Scattered melting detection system and method of using the same

Abstract

A high-bandwidth time and spatial resolution scattered phase transition microscope system configured to detect the onset of melting in a wafer processed by a laser annealing system having a very short dwell time and spot size.

Description

Technical Field

[0001] This application claims the benefit of priority of U.S. Provisional Patent Application No. 63 / 337,714, entitled "Scattering Melting Detection System and Method of Use Thereof", filed on May 3, 2022, which is hereby incorporated by reference in its entirety.

[0002] The present disclosure generally relates to temperature measurement. In particular, the present disclosure relates to phase transition microscopy.

Background Art

[0003] Pulsed laser melting annealing systems are used for rapid heat treatment of chips in the manufacture of advanced integrated circuit (IC) chips. The heat treatment provided by such laser systems can be used for various effects such as dopant activation, bond formation, and other changes in the material and electrical properties of the chips.

Summary of the Invention

Problems to be Solved by the Invention

[0004] However, accurate measurement and control of the wafer temperature during annealing using a pulsed laser system are important, and as the duration of the laser raising the wafer temperature to the melting temperature and the surface area of the wafer on which the laser acts decrease, the measurement and control may become difficult. Accurate measurement and control of the wafer temperature are also important for the calibration of the annealing system.

Means for Solving the Problems

[0005] In one embodiment, the present disclosure relates to a method for determining the occurrence of a solid-to-liquid material phase transition of a semiconductor material during a laser annealing process. The method includes forming an annealing image on the surface of the semiconductor material using an annealing laser beam from an annealing laser, forming a scattered image of a region of the semiconductor surface including the annealing image, and identifying a solid-to-liquid material phase transition of the semiconductor material according to the intensity of light collected in the scattered image.

[0006] In another embodiment, the present disclosure relates to a method for analyzing scattered image data. The method includes receiving a scattered image of a region of a semiconductor surface captured during a laser annealing process of the semiconductor, and identifying a solid-to-liquid material phase transition of the semiconductor material from the image, and this identification includes identifying statistical, time-series, or geometric non-linear variations of the image intensity as a function of one or more annealing process control parameters.

[0007] In yet another embodiment, the present disclosure relates to a method for performing melting calibration for a laser annealing process of a semiconductor material. The method includes annealing a region of the semiconductor material with an annealing laser beam from an annealing laser, which includes gradually changing output levels below and above a melting threshold output level at which a solid-to-liquid material phase transition of the semiconductor material occurs, capturing a scattered image of the region of the semiconductor material during annealing, and determining at least one melting threshold setting value output level of the annealing laser from the captured scattered image.

[0008] In yet another embodiment, the present disclosure relates to a scattered melting detection system. The scattered melting detection system includes a stage configured to support a semiconductor wafer, an annealing laser configured to anneal the semiconductor wafer, a scattered melting detector including an image capture device including a focal plane array (FPA), and a scattered optical arrangement configured to block or avoid specular reflection of a beam formed by the annealing laser and form a scattered image of the surface of the semiconductor wafer on the FPA.

[0009] In yet another embodiment, the present disclosure relates to a laser annealing system. The system includes an annealing laser, a storage device including at least one annealing laser melting threshold calibration setting value for specifying the operating parameters of the annealing laser to raise the temperature of a local region of a semiconductor material to the melting threshold under a set of predetermined annealing process conditions, the annealing laser melting threshold calibration setting value being derived from a scattered image of the semiconductor material obtained during the laser annealing process of the semiconductor material, and a controller configured to control the annealing laser according to the annealing laser melting threshold calibration setting value.

[0010] In yet another embodiment, the present disclosure relates to a non-transitory machine-readable storage medium. The non-transitory machine-readable storage medium includes at least one annealing laser melting threshold calibration setting value for specifying the operating parameters of the annealing laser to raise the temperature of a local region of a semiconductor material to the melting threshold under a set of predetermined annealing process conditions, the annealing laser melting threshold calibration setting value being derived from a scattered image of the semiconductor material obtained during the laser annealing process of the semiconductor material.

[0011] In yet another embodiment, the present disclosure relates to a method for forming a non-melt state reference image and parameters for use in a melt detection algorithm. The method includes annealing a wafer in a non-melt state, capturing an image frame using a scattered melt detector, recording the output level of the annealing laser, creating a non-melt state reference image from the captured image frame, processing the non-melt state reference image to generate a plurality of one-dimensional stage scan direction intensity profiles in the non-melt state, identifying a region of interest (ROIref) in the non-melt state reference image, and creating an average one-dimensional stage scan direction intensity profile.

[0012] In yet another embodiment, the present disclosure relates to a method for detecting whether a wafer material has melted during an annealing process. The method includes annealing the wafer below, at, or above a melting threshold, capturing an image frame using a scattered melt detector during annealing, recording a laser output level, creating an averaged image from the captured image frames, and creating a melt detection image that associates the averaged image with a corresponding non-melt state reference image of the same wafer type.

[0013] In yet another embodiment, the present disclosure relates to a method for detecting whether a wafer material has melted during an annealing process. The method includes annealing the wafer below, at, or above a melting threshold, capturing an image frame using a scattered melt detector during annealing, calculating a statistical variation of pixel values within a region of interest (ROIref) of the image frame that is the region of the image where an anneal laser image is formed on the wafer, and comparing the calculated statistical variation to identify the onset of melting.

Brief Description of the Drawings

[0014] For purposes of illustrating the present disclosure, the drawings show aspects of one or more embodiments of the present disclosure. However, it should be understood that the present disclosure is not limited to the exact arrangements and instrumentalities shown in the drawings.

Figure 1

Figure 2

Figure 3A

Figure 3B

Figure 3C

Figure 4A

Figure 4B

Figure 5A

Figure 5B

Figure 6A

Figure 6B

Figure 6C

Figure 7A

Figure 7B

Figure 7C

Figure 8

Figure 9A

Figure 9B

Figure 10

Figure 11

Figure 12

Figure 13

Figure 14

DETAILED DESCRIPTION OF THE INVENTION

[0015] Aspects of the present disclosure are configured to detect the onset of melting of a semiconductor wafer processed by a scanning CW or QCW (pulse duration << dwell time) laser annealing system having a very short dwell time and a small spot size, and include a phase transition microscope system having high bandwidth time and spatial resolution for calibrating such a system for an annealing process. In some examples, the melting detection system of the present disclosure is configured to accurately and reliably detect the onset of melting for lasers having a very short dwell time in the range of 10 ns to 500 ns or 25 ns to 250 ns, and a spot size in the range of 10 μm to 150 μm × 10 μm to 150 μm.

[0016] Melting annealing involves locally raising the temperature of the wafer surface or subsurface to its melting point using an incident laser beam. For example, if the wafer surface is composed of a thin silicon layer over a layer of material having a lower melting temperature (e.g., germanium or germanium-silicon), the wafer surface may remain solid while the material beneath its surface melts. The melting detection system described herein may be configured to detect melting in a surface melting process where melting occurs on the wafer surface, and / or may be configured to detect melting in a subsurface melting process where the wafer surface remains solid and the volume of material beneath the wafer surface melts.

[0017] FIG. 1 is a schematic diagram of an exemplary quasi - continuous wave (QCW) laser annealing system (“system”) 100 into which the melt detection system of the present disclosure can be incorporated. As will be understood, system 100 is described and illustrated by way of example, and the melt detection system of the present disclosure may be used in any of a variety of other laser annealing systems. Certain aspects of an exemplary embodiment of system 100 are described in U.S. Patent No. 10,083,843, entitled “Laser Annealing Systems and Methods Having a Very Short Residence Time,” issued on September 25, 2018, the content of which is hereby incorporated by reference in its entirety. References incorporated by reference in U.S. Patent No. 10,083,843, including U.S. Patent Nos. 8,309,474, 8,546,805, 8,865,603, and U.S. Patent Application No. 14 / 497,006, now issued as U.S. Patent No. 9,343,307, are hereby incorporated by reference in their entirety.

[0018] System 100 includes a wafer stage 120 that operably supports a chuck 130 having an upper surface 132. The wafer stage 120 is operably connected to a stage controller 124 and is configured to move within the X-Y plane and optionally also in the Z direction via the operation of the stage controller. The chuck upper surface 132 is configured to operably support a wafer 10 having a surface 12. In some examples, the wafer 10 may be processed and may include patterns defined by semiconductor structures typically associated with various stages of manufacturing integrated circuit chips, and often has an optically deep and highly patterned multi-layer structure. In other examples, the wafer 10 may be unprocessed and have no manufactured structures, and the surface 12 may be an extremely smooth mirror-like surface. As will be described in more detail below, the unprocessed wafer 10 may be useful in the measurement systems disclosed herein for calibrating and tracking the performance of a laser annealing tool. The measurement systems disclosed herein, when used with a sufficiently calibrated annealing system, can perform calibration of the wafer-specific annealing laser output and can also determine, for example, the melting threshold setting output levels for specific materials and surfaces. In one example, the chuck 130 includes a heater that heats the wafer 10 to a temperature higher than a uniform ambient temperature. The wafer stage 120 is operably connected to a stage controller 124.

[0019] System 100 includes an annealing laser system 250 configured to generate a scanning laser beam 268 that forms an annealing image 270 (FIG. 2) on the wafer surface 12. The annealing laser system 250 includes an annealing laser 260 that emits an initial laser beam 262, a modulator 264 operably connected to a modulator driver, and a scanning optical system 266 that receives the initial laser beam and generates the scanning laser beam 268. In one example, the annealing laser 260 may be a QCW fiber laser that emits a scanning laser beam 268 having a wavelength of, for example, 532 nm. The foregoing is provided by way of example, and other annealing laser systems now known or later developed may be used with the measurement systems disclosed herein. In one example, the modulator 264 is an acousto-optic modulator (AOM) used with the scanning optical system 266 that selectively and alternately blocks and passes the initial laser beam 262 to control the scanning of the annealing image 270, resulting in a sweeping motion of the annealing image across the entire wafer surface.

[0020] In some examples, system 100 may optionally also include a preheating laser system 150 configured to operate in conjunction with the annealing laser system 250 to generate a preheating laser beam 168. The preheating laser beam 168 is used to preheat the wafer surface 12 by raising the temperature T S of the wafer surface (or beneath the wafer surface) to a pre-annealing temperature T M lower than the melting temperature T PA up to.

[0021] The preheating laser system 150 may include a preheating laser 160 and a line-forming optical system 166. The preheating laser 160 may be a diode laser, a fiber laser, or a continuous wave (CW) p-polarized 10.6 micron CO 2 laser such as a CO 2A laser may be included. In one example, the line-forming optical system 166 is configured such that the preheating laser beam 168 is incident on the wafer surface 12 at an incident angle that is substantially perpendicular to the wafer surface 12 or at a large oblique incident angle. In one example embodiment, the incident angle of the preheating laser beam 168 is approximately equal to the Brewster angle with respect to the wafer surface 12, such that the adverse effect on the pattern density due to non-uniform light absorption is reduced or minimized.

[0022] The line-forming optical system 166 is configured to receive an initial laser beam 162 from the preheating laser 160 and form a preheating line image 170 (FIGS. 2 and 3) on the wafer surface 12 therefrom. The line-forming optical system 166 and the scanning optical system 266 can each include lenses, mirrors, apertures, filters, active optical elements (e.g., variable attenuators, etc.), and combinations thereof. In one example, one or both of the line-forming optical system 166 and the scanning optical system 266 can be configured to perform beam conditioning, for example, to homogenize the respective laser beams 162 and 262 and / or to impart a selected cross-sectional shape to the laser beams. Non-limiting examples of optical systems suitable for performing such beam conditioning are disclosed in U.S. Patent Nos. 7,514,305, 7,494,942, 7,399,945, and 6,366,308, the contents of which are hereby incorporated by reference in their entirety. In one example, the initial laser beam 262 from the annealing laser 260 has high quality (e.g., is substantially Gaussian) and is used with little (in some cases, no) beam conditioning.

[0023] As will be described in more detail below, system 100 also includes a scattered melting detector 280 configured with a scattered optical arrangement for detecting scattered light from a region of interest on wafer surface 12, decomposing the scattered light temporally and spatially, and detecting the onset of melting. The scattered optical arrangements of the present disclosure include systems based on a spatial resolution sensor or sensor array that has no far-field mirror diffraction limit of the probe beam and does not detect specular reflections from a flat wafer surface. One example form of the scattered optical arrangement is dark field imaging. Probe control (such as polarization and wavelength) and post-optical measurement selection (such as polarization filters, birefringent crystals, etc.) can also be used to capture the same or similar scattered information.

[0024] Exemplary laser annealing system 100 further includes a controller 300. In one example, controller 300 is or includes a computer such as a personal computer or a workstation. Controller 300 preferably includes any of a commercially available microprocessor, an appropriate bus architecture for connecting the processor to a storage device such as a hard disk drive, and appropriate input devices (e.g., keyboard) and output devices (e.g., display). Controller 300 is programmable via instructions (software) recorded on a non-transitory computer-readable medium (e.g., memory, processor, or both), and those instructions cause the controller to perform the various functions of system 100 to effect annealing of wafer 10.

[0025] Controller 300 is operably connected to preheating laser system 150 and scanning laser system 250 and controls the operation of these systems. Controller 300 is electrically connected to modulator 264 and controls the operation of the modulator using control signal SMod. In one example, controller 300 includes digital signal processors (DSPs) (not shown) for controlling the scanning function in preheating laser system 150 and scanning laser system 250. In some examples, controller 300 may be operably connected to scatter melt detector 280 and scanning optical system 266 and may be configured to receive and process melt signal SM generated by the scatter melt detector during the laser annealing process for use in controlling annealing laser system 250 or preheating laser system 150, similar to adjusting the output level of scanning laser beam 268 to achieve desired process conditions. In other examples, controller 300 may control the output level of annealing laser system 250 according to previously acquired annealing laser melt calibration setting value 1024 stored in computer-readable storage device 1016 (FIG. 10). Controller 300 or another computing device may also be configured to control components of scatter melt detector 280, such as controlling the image capture device (ICD) of the melt detector.

[0026] In an example of the operation of system 100, system controller 300 transmits a first control signal S1 to preheating laser 160, and preheating laser 160 generates an initial laser beam 162 in response thereto. This initial laser beam 162 is received by optical system 166, and optical system 166 forms a preheating laser beam 168 therefrom, and preheating laser beam 168 forms a preheating line image 170 on wafer surface 12. System controller 300 also transmits a second control signal S2 to annealing laser 260 as a second control signal S2, and annealing laser 260 generates an initial laser beam 262 in response thereto. This initial laser beam 262 is received by scanning optical system 266, and scanning optical system 266 forms a scanning laser beam 268 under the control of control signal SS, and then scanning laser beam 268 forms an annealing image 270 on wafer surface 12. System controller 300 also transmits a third control signal S3 to stage controller 124, and stage 120 performs a controlled movement to move (scan) wafer 10 relative to preheating line image 170 and annealing image 270. In an example where chuck 130 performs wafer preheating, system controller 300 may transmit another control signal (not shown) to the chuck controller to start or control wafer preheating. The normal chuck preheating range is from room temperature (25°C) to 400°C.

[0027] FIG. 2 is a top view of wafer 10 illustrating a method of scanning a wafer surface using preheating line image 170 in an example where annealing image 270 and preheating laser system 150 are used. Moving preheating line image 170 relative to wafer surface 12 can be achieved by moving stage 120 in a first direction, indicated by arrow AR1, which is also referred to herein as the preheating scan direction or the stage scan direction, for example, in the y direction. The portion of wafer surface 12 associated with preheating line image 170 has a wafer surface temperature T S is the preliminary annealing temperature T PAIt represents a locally preheated portion of the wafer surface that is raised up to. The anneal image 270 formed by the scanning laser beam 268 moves relative to the wafer surface 12 in a second direction orthogonal to the first direction, for example, the x-direction, as indicated by the arrow AR2, and is also referred to herein as the anneal scanning direction or the sweeping direction.

[0028] The scanning optical system 266 is configured to scan or sweep the anneal image 270 on the preheat line image 170 in the x-direction over the length of the preheat line image. In one example, the scanning speed of the anneal image 270 is sufficiently fast compared to the movement of the preheat line image 170, and during the scanning of the anneal image, the preheat line image is substantially stationary.

[0029] When the anneal image 270 reaches the end of the preheat line image 170, the scanning laser beam 268 and the corresponding anneal image 270 are turned off by activating the modulator 264 (FIG. 1). As a result, the modulator blocks the transmission of the initial laser beam 262. While the scanning laser beam 268 is "off", the preheat line image 170 moves in the y-direction, thereby enabling the next portion of the wafer surface 12 to be scanned. In one example, the movement of the preheat line image 170 may be continuous, for example, by continuously moving the stage 120 in the y-direction. When the preheat line image 170 is positioned at a predetermined location, the scanning laser beam 268 is turned on again by setting the modulator 264 to the transmission mode when the scanning optical system 266 can direct the scanning laser beam 268 and the corresponding anneal image 270 to the starting position of the first end of the newly positioned preheat line image 170. Then, the scanning of the anneal image 270 on the newly positioned preheat line image 170 is performed. FIG. 2 shows an exemplary method of substantially scanning the entire wafer surface 12 (e.g., at least the patterned portion) by repeating the above-described scanning method. In other examples where the anneal system does not include a preheat laser such as the preheat laser system 150, the preheat line image 170 is omitted, and the anneal image 270 can be similarly scanned at high speed in the x-direction while the wafer 10 is moved in the y-direction by the stage 120.

[0030] Figure 3A is an enlarged top view of one of the preheating line image 170 and the annealing image 270 formed on the wafer surface 12. The preheating line image 170 has a proximal end 172, a distal end 174, and opposing side portions 173. The preheating line image 170 has a longitudinal (dimension) extending from the proximal end 172 to the distal end 174 and has a length L1. The preheating line image 170 also has a narrow dimension (dimension) measured between the opposing side portions 173 having a width W1. In one example, the length L1 ranges from 5 mm to 20 mm, and an exemplary length L1 ranges from 7 mm to 12 mm. Also, in one example, the width W1 ranges from 50 μm to 200 μm, and an exemplary width W1 is 150 μm. In one embodiment, the preheating laser beam 168 (FIG. 1) has a Gaussian intensity profile in the scanning direction (e.g., the y direction) and has a relatively flat top distribution in the longitudinal (scanning transverse direction) direction (e.g., the x direction). The beam width W1 can be defined by the 1 / e 2 intensity value or the full width at half maximum (FWHM) of the Gaussian profile.

[0031] The anneal image 270 overlaps a part of the preheat line image 170, and the overlapping area is referred to as the scanning overlap area (SOR) in this specification. The anneal image 270 has a long dimension with length L2 and a narrow dimension with width W2. In the illustrated example, the anneal image 270 has a substantially Gaussian intensity distribution in the x and y directions. The long dimension L2 of the anneal image 270 is oriented in the direction of the short dimension W1 of the preheat line image 170. In one example, the length L2 ranges from 100 μm to 500 μm, the width ranges from 10 μm to 50 μm, and in some examples, the width ranges from 15 μm to 20 μm or from 16 μm to 18 μm. The scanning direction AR2 of the anneal image 270 is perpendicular (orthogonal) to its longitudinal direction. The scanning direction AR2 is also referred to as the anneal scanning direction or the sweeping direction and is substantially orthogonal to the preheat scanning direction AR1. The width W2 of the anneal image 270 defines the width of the scanning overlap area SOR in the anneal scanning direction AR2. By rapidly scanning the anneal image 270 in the scanning direction AR2, the SOR forms an anneal irradiation line 272 having substantially the same size as the preheat line image 170 that is sequentially moved in the y direction (preheat scanning direction AR1) along the wafer surface 12 by the movement of the stage 120 in the y direction.

[0032] In one example, the length L2 is substantially larger than the width W1 (e.g., two to four times larger), and as a result, as shown in FIG. 3A, the ends of the anneal image 270 extend beyond the sides 173 of the preheat line image 170. This makes it relatively easy to align the preheat line image 170 and the anneal image 270 to define the scanning overlap area SOR. This configuration utilizes the high-intensity portion in the center of the anneal image 270 to raise the wafer surface temperature T S to the melting temperature T M until it rises.

[0033] In one example, the width of the residence time t d of the scanning overlap area SOR is in the range of 10 ns ≤ t d ≤ 500 ns, and in another example, 25 ns ≤ t dis within the range of ≦250 ns. In the case of a width W2 = 15 μm and a residence time of 25 ns, the scanning speed of the annealing image 270 and thus the scanning speed of the scanning overlap region SOR is v s =(W2) / t d = 600 m / s. When the residence time t d is 250 ns, the scanning speed v s is 60 m / s. When the residence time t d is 500 ns, the scanning speed v s is 30 m / s. When the residence time t d is 10 ns, the scanning speed v s is 1500 m / s. In one example, the scanning optical system 266 is configured to provide these speeds or any speed therebetween.

[0034] FIG. 3B is an enlarged top view of another exemplary embodiment of the system 100 in which the preheating laser system 150 is omitted. The annealing laser system 250 is designed and configured to raise the wafer surface temperature T S to a desired annealing temperature. By the high-speed scanning of the annealing image 270 in the scanning direction AR2, an annealing illumination line 272 that is sequentially moved in the y direction (arrow AR1) along the wafer surface 12 is formed by the movement of the stage 120 in the y direction. In one example, the length L1 of the annealing irradiation line 272 ranges from 3 mm to 12 or less, the width W2 of the annealing image 270 ranges from 10 μm to 50 μm, and the length L2 of the annealing image 270 ranges from 100 μm to 500 μm.

[0035] FIG. 3C is a top view of a portion of the wafer surface 12 showing three points in time (labeled a, b, and c) of the anneal image 270 moving in the scan direction AR2 along the anneal irradiation line 272. FIG. 3C conceptually shows a melt pool 310 formed by the scanning laser beam 268, in which a local region of the wafer surface 12 is melted and partially or completely transitions from the solid phase to the liquid phase, resulting in an interface 312 between the solid phase and the liquid phase. In the example shown, the size of the melt pool 310 increases along the scan direction AR2, which is potentially due to an increase in the optical output density of the laser beam 268 through the sweep of the anneal image 270 across the anneal irradiation line 272. The phase transition of the wafer material within the melt pool 310 is a first-order phase transition that results in discontinuities in both the density of the wafer material and the form of the interatomic or intermolecular interactions. Due to the phase transition, the refractive index of the material changes significantly due to the difference in electromagnetic interactions in the two different phases (solid and liquid) of the material. In one example, the refractive index of the liquid phase within the melt pool 310 is substantially higher than the refractive index of the solid-phase wafer material outside the melt pool 310, e.g., 100% to 300% higher than the refractive index of a typical IC semiconductor material.

[0036] The interface 312 between the melt pool 310 and the solid-phase material surrounding the melt pool functions as a major source of scattering of the incident light, and the magnitude of the wide-field structure factor increases with the radius r of the melt pool 310. l The scattering imaging system disclosed herein is designed and configured to detect the scattered light that occurs at the start of melting and when the melt pool 310 first begins to form. The intensity of the scattered light increases rapidly and non-linearly as the size of the melt pool 310 increases, and the scattering imaging system of the present disclosure may be configured to detect the increasing intensity of the scattered light due to the increasing extent of the melted wafer material.

[0037] In the non-molten state of the solid phase, the relationship between the intensity of the reflected light signal and the material temperature can be approximated as approximately linear, γT, where γ is a scaling coefficient and T is the material temperature. For a melt pool 310 with a wavelength of the incident light smaller than, for example, the wavelength of the laser beam 268, the scattering intensity of the melt pool 310 and the interface 312 is approximately αV 4 / 3 and is approximated here, where α is a scaling coefficient and V is the volume of the melt pool, or the fourth power of the radius of the melt pool, where α >> γ. For a melt pool 310 much larger than the wavelength of the incident light, a discontinuity on the melt pool with a radius r l becomes the main scattering source for the wide field of view, and its size is approximately αV 2 / 3 and is approximated as such. In this way, the overall scaling relationship between the increase in the light output and the size of the melt pool is highly non-linear and may ultimately saturate the scattering imaging system. In the example shown in Figure 3C, due to the initial formation of the melt pool 310, a sharp discontinuity occurs in the light detected by the scattered melt detector 280 due to the sharp increase in the scattered light emitted from the region surrounding the melt pool. As the size of the melt pool 310 increases (e.g., the sizes of the melt pools 310b and 310c compared to the melt pool 310a), the intensity of the light detected by the scattered melt detector 280 increases non-linearly on the order of the fourth power of the increase in the radius of the melt pool, which may rapidly saturate the melt signal SM generated by the melt detector. The scattered melt detector 280 and the scattering imaging channel disclosed herein are designed and configured to be highly sensitive to the onset of melting where a discontinuous increase in the intensity of the light captured by the melt detector occurs substantially instantaneously at the onset of melting, resulting in an increase in the intensity of the light signal by several orders of magnitude.

[0038] During the annealing process, a scattered melt detector 280 can be used to capture an annealing image 270 that includes light diffracted, reflected, or refracted by an optical structure within the region of the annealing image from the surrounding area. In one example, the scattered melt detector 280 generates a melt signal SM representative of the detected light and transmits the melt signal to a controller 300. In one example, the controller 300 receives the melt signal SM and uses the signal to create a feedback loop that controls one or more characteristics, such as the output level of the annealing laser system 250 and, in some examples, the preheating laser system 150, resulting in the ability to control the wafer surface temperature T S In some examples, the scattered melt detector 280 may be used to calibrate and track the optical performance of the annealing laser system 250. Also, in some examples, the scattered melt detector 280 may be used to develop a wafer-specific laser output calibration that provides a melt threshold laser output setting corresponding to a set of predetermined annealing process conditions.

[0039] FIG. 4A shows a side view of one embodiment of the scatter melt detector 280. In the illustrated example, the melt detector 280 receives light from the wafer surface 12 along a scatter detector path 402 having an angle α1 with respect to the surface normal N that is different from the angle α2 of the scanning laser beam 268. In the illustrated example, the scanning laser beam 268 is used simultaneously as a probe for scatter temperature measurement and as an anneal laser beam. In other examples, a light source other than the scanning laser beam 268 can be used as the probe. Using the scanning laser beam 268 as the light source for scatter measurement provides the advantage of a simpler arrangement because no additional light source is required and it irradiates only the field of view around the small anneal image 270 that can travel rapidly, e.g., up to about 1 km / s, across the wafer surface 12. Using the scanning laser beam 268 as the light source for scatter measurement results in the advantage of a simpler arrangement because no additional light source is needed and it irradiates only the field of view around the small anneal image 270 that can move at high speed, e.g., up to about 1 km / s, across the wafer surface 12.

[0040] The scanning laser beam 268 has an incident angle α2 that is different from the angle α1 with respect to the surface normal N and forms an anneal image 270 on the wafer surface 12. The angle α1 of the scatter detector path 402 is designed so that the specular reflection 268R of the scanning laser beam 268 from the surface 12 of the wafer 10 does not enter the scatter detector path 402, thereby enabling scatter observation by the melt detector 280.

[0041] The melt detector 280 includes an optical element 404 and an image capture device (ICD) 406. The ICD 406 includes at least one pixelated two-dimensional CCD or CMOS focal plane array (FPA) and associated electronics known in the art of image capture devices, including a processor and a memory unit for storing the captured images, or is electrically connected thereto. The optical element 404 is configured to image the wafer surface 12 onto the FPA. In one example, the ICD 406 is configured to capture a scattered image at an exposure time frame that is relatively long compared to the dwell time of the laser beam 268, e.g., at an image frame capture rate of about 50 us, such that at least one sweep of the anneal image 270 across the anneal irradiation line 272 is captured in a single image frame.

[0042] The optical element 404 may include one or more lenses, mirrors, filters, diffraction gratings, apertures, or other components. Light collected along the scattered detector path 402, e.g., scattered light, is collected by the optical element 404, and in some cases, reflected and / or filtered, and imaged onto the FPA of the image capture device 406. The field of view (FOV) of the optical element 404 may include all or a portion of the wafer surface 12. For example, the FOV of the optical element 404 may have a width in the scan direction AR2 (see FIG. 3B) that is approximately equal to the length L1 of the anneal irradiation line 272, or a width in the direction AR2 that is about 5% to about 50% greater than the length L1 of the anneal irradiation line. The width of the FOV of the optical element 404 in the stage scan direction AR1 (see FIG. 3B) may be the same as the width of the FOV in the anneal image scan direction AR2, or may be larger or smaller than the width of the FOV in the scan direction AR2. In one example, the width of the FOV of the optical element 404 in the stage scan direction AR1 is at least twice the diameter or outer extent of the anneal image 270 in the stage scan direction, such that the wafer surface 12 in the non-molten state and outside the anneal image is captured in the same image as the portion within the anneal image. The FOV of the optical element 404 in the sweep direction AR2 may be at least 10% greater than the length L1 of the illumination line 272, and in some examples, at least 20% greater. The spatial resolution of the ICD 406 may be from 10 μm / pixel to 50 μm / pixel, and in some examples, about 30 μm / pixel.

[0043] As shown, a light source, here an annealing laser system 250, generates a scanning laser beam 268 that forms an annealing image 270 on the wafer surface 12. The first portion of the light of the scanning laser beam 268 is specularly reflected by the wafer surface 12 and does not reach the melt detector 280. When the temperature of the wafer surface 12 along the annealing irradiation line 272 is below the melting temperature and the wafer material is in the solid phase, the scattered melt detector 280 may detect, for example, defects on the wafer surface 12 such as small thickness variations, or light from photoluminescence and scattered light due to large-scale defects such as scratches or dust, or in the case of a processed wafer, diffraction, reflection, and / or refraction of light from optical discontinuities of the wafer such as the wafer surface or a fabricated pattern below the surface, i.e., a certain amount of background light. Thus, in the case of an untreated and highly polished wafer 10, the melt detector 280 may detect little or no light along the scattered detector path 402, and the field of view of the melt detector 280 is substantially dark. In the case of a processed wafer, scattered light associated with the non-molten solid-phase state may be captured to some extent.

[0044] The output density of the scanning laser beam 268 is the melting threshold energy density E mWhen it rises to, a phase transition of the wafer material occurs on or below the wafer surface 12, resulting in a volume of molten liquid-phase material located near the annealing image 270. The reflectivity of the molten liquid-phase material is substantially different from that of the material in the solid phase. The interface between the solid phase and the liquid phase of the newly formed liquid volume (e.g., interface 312 (FIG. 3C)) creates a new optical discontinuity on the wafer surface that strongly scatters light in a hemisphere of angles above the molten region of the wafer. The reflection observed in the molten region (e.g., molten pool 310) is diffuse reflection and, in some examples, approaches Lambertian reflection. The intensity resulting from the imaging thermal reflection signal of the molten region increases substantially compared to the solid phase, for example, by 50%. On the other hand, the change in the intensity of the scattered imaging thermal reflection signal along the scattered detector path 402 due to the phase transition of the material from solid to liquid is on the order of, for example, 100,000% higher intensity and is extreme, providing excellent sensitivity for detecting the onset of melting.

[0045] In one example, the ICD 406 captures a scattered image over a time longer than the dwell time of the annealing image 270 at a corresponding frequency f ≤ 1 / t d The scattered melting detector 280 decomposes time along the annealing laser sweep direction AR2 using the annealing image 270 moving at a velocity v and x = v * t. In this way, the captured scattered image can be used to decompose the scattered signal data along the x-direction (sweep direction AR2) into output fluctuation spectrum measurements in the bandwidth up to f = 1 / t d Using a scattered image frame that includes the entire sweep or multiple sweeps of the annealing image 270, the scattered melting detector 280 can be used to measure the uniformity of the process over the entire annealing image sweep stripe within a single image frame. In one example, the scattered melting detector 280 supplements a spatially resolvable scattered image for all points x = v·(t ± t d ) within a single frame, thereby providing a scattered image at frequencies f ≤ 1 / t dconfigured to measure the melting behavior and uniformity therein, which is very practical and useful in the optimization of the annealing laser system 250.

[0046] FIG. 4B shows another exemplary embodiment of the scattered melting detector 280 in which a probe light source 410 different from the annealing laser system 250 is used as a probe for scattering measurements. The illustrated example has a configuration similar to the example shown in FIG. 4A and includes a detector including an optical element 412 that can include the same or similar components as the optical element 404 and the ICD 406. The probe light source 410 may be a pulsed light source that emits probe light 420 incident on the wafer surface 12 in a region overlapping the annealing image 270. Similar to the example shown in FIG. 4A, the scattered melting detector 280 of FIG. 4B has a scattered detector path 416 such that the optical element 412 does not capture the specular reflection 420R of the probe light 420 or the specular reflection 268R of the scanning laser beam 268. In the illustrated example, the melting detector 280 receives light from the wafer surface 12 along a scattered detector path 416 having an angle α1 with respect to the surface normal N different from the angle α2 of the scanning laser beam 268 and the angle α3 of the probe light 420. The probe light source 410 may be a diode or an optical fiber connected to a laser source, and may emit narrow-band light such as green, blue, NIR, or communication-band light such as between 1260 nm and 1675 nm, for example. The optical element 412 may be configured to remove wavelengths of light other than the probe light 420. The probe light source 410 can be configured to emit pulses of the probe light 420 at a frequency higher than the frame rate of the ICD 406 or at short time intervals so as to be able to image discrete portions of the annealing illumination line 272.

[0047] Figures 5A and 5B conceptually show two alternative optical arrangements 500, 501 for scattered imaging. Apparatus 500 includes an objective lens 502 and an imaging lens 504 for forming a scattered image of a region of the wafer surface 12 including a melt pool 310 on the FPA of an ICD such as ICD406. Arrangement 500 corresponds to the example shown in FIGS. 4A and 4B, where light such as a scanning laser beam 268 (example of FIG. 4A) or another probe light source (e.g., probe light 420 of the example of FIG. 4B) is incident on an object such as melt pool 310. From the perspective of Fourier optics, the zero-order light 530 (non-scattered light) does not reach the objective lens 502 due to the spatial position and numerical aperture of lenses 502 and 504. As a result, the image formed on the FPA is due to the higher-order diffraction intensity 532 scattered by the melt pool 310 and / or other light scatterers or wafer surface or subsurface discontinuities.

[0048] Figure 5B shows an alternative optical arrangement 501, where at least one opaque light blocker 520 is used to prevent the zero-order light 530 (non-scattered light) from reaching the FPA of an ICD such as ICD406. As a result, the image formed on the FPA by the objective lens 510 and imaging lens 512 is due to the higher-order diffraction intensity 532 scattered by the melt pool 310 and / or other light scatterers or discontinuities on or below the wafer surface. The scattered melt detector 280 of FIG. 4A or the scattered melt detector of FIG. 4B may be modified to have an optical arrangement 501 that uses at least one opaque light blocker 520 to form a scattered image.

[0049] Figure 6A conceptually and graphically shows data captured within an image frame by the ICD406 of the scattered melt detector 280 during the sweep of the anneal image 270 when the wafer surface 12 is in a non-melt state. Figure 6B conceptually shows the change in the data captured in the image when the optical output of the anneal laser system 250 increases and the temperature of the wafer surface 12 exceeds the melt threshold and the wafer material begins to melt.

[0050] In the non-molten state, the image captured by the melt detector 280 is the probe intensity I(x → ) scattered by the roughness of the wafer and, in some cases, the resonant scatterers embedded in the wafer surface 12. In the thin film approximation, the roughness due to the height difference δ(x → ) acts as a random holographic phase grating φ(x → ), which acts on the electric field as e iφ(δ(x→)) = e iαδ(x→) . The fluctuating output spectrum, i.e., the square of the Fourier transform of δ(x → ), |F(δ(x → ))| 2 = |δ(k → )| 2 is not zero, usually far exceeding k → = 4π / λ, in the manufacturing process of a uniform wafer. As a result, I(x → ) is scattered uniformly over the upper hemisphere of the wafer plane. FIG. 6A conceptually shows this relatively flat response at low illumination levels that can occur in the non-molten state.

[0051] In the case of a processed and highly patterned wafer 10, some material manufacturing processes include rare light scatterers in the thin film, which can generate a random image that may provide a false melt signal. However, when the image data from the melt detector 280 is averaged over an edited image frame, the false melt signals occurring in the non-molten state appear in the image frame where the high-intensity signals are uniformly edited over the entire sweep across the annealing irradiation line 272 with the increase in the optical output density of the annealing laser 250, so that they can be distinguished from the true melt signal. Aspects of the present disclosure include editing a set of high-resolution images of the non-molten wafer 10 to sample the shape of the irradiation probe beam, e.g., the scanning laser beam 268 or the probe light 420 from the probe light source 410. Then, the non-molten state image data is used with a radius r melt < r beamIt can be distinguished from the molten state image data including the high-strength artifacts of the molten pool 310 and the related interface 312. By reducing the absorption length of the probe beam in the liquid phase portion of the wafer material within the molten pool, the scattering caused by the subsurface optical discontinuities in the processed wafer is reduced, thereby simplifying the scattered light response when the melting threshold is exceeded and making it easier to detect the onset of melting.

[0052] FIG. 6C shows an image frame of test data from an exemplary embodiment of a system 100 that uses the amplitude of the captured light to identify melting. FIG. 6C shows an example of a system that is not properly optimized. The start (top of the image) and end (bottom) of the anneal image sweep are both in the molten state, and the intermediate light-dark region is just before the non-molten state. Since the optical output density of the anneal laser system 250 was not properly calibrated, unwanted variations in the optical output occurred, resulting in the start and end of the sweep being in the molten state and the middle of the sweep being in the non-molten state.

[0053] Figures 7A through 7C show test data from a melt threshold calibration at a specific chuck temperature for a specific processed wafer type having a specific material laminate. Figures 7A through 7C show one or more sweeps from three different stripes of a scanning laser beam 268, where the stripes are a series of continuous sweeps. In Figure 7A, the output of the annealing laser system 250 is set to 132 W, in the next stripe of Figure 7B the output level is increased to 135 W, and in the third stripe of Figure 7C the output is increased to 138 W. The arrows indicate the positions within the images where the signal is saturated, indicating high light intensity in these regions of the image due to a sharp increase in scattered light caused by melting of the wafer material. The images of Figures 7A through 7C also show low-intensity artifacts caused by the inherent roughness and non-uniformity of the processed wafer (some of the lower intensity artifacts are enclosed by dotted lines in each figure). However, the high-intensity positions corresponding to the melt regions of the wafer have a much greater spatial density and magnitudes that are orders of magnitude greater than the non-melted background light in the scattered images, standing out strongly from the background scatter. Figures 7A through 7C show an example of how the melt threshold can be calibrated for a specific processed wafer type having a specific material laminate by increasing the laser output and studying the melt measurements.

[0054] FIG. 8 shows additional data from the same calibration process as shown in FIGS. 7A through 7C. The Y-axis is the cumulative intensity of the light captured by the FPA of the ICD406. More specifically, the Y-axis is the sum of all gray values of all pixels in the image, and each pixel of the FPA records the level of the intensity of the incident light on a gray level scale between 0 (black) and 1 (white), as is known in the art. The X-axis is the number of sweeps of the anneal image. FIG. 8 shows the first plurality of sweeps of the anneal image at 132 W, where the sum of the gray values is a low initial value, which indicates the non-molten state and the background light scattering captured by the FPA of the ICD406. FIG. 8 shows the increase in gray values when the output is increased to 135 W, which was determined to be the melt threshold, and then the laser output is increased to 138 W, which was determined to exceed the melt threshold, showing a sharp increase in the intensity of the light captured by the ICD406 when the liquid volume of the wafer material is consistently formed over most of the anneal image sweep.

[0055] FIGS. 9A and 9B show test data of a fully optimized system having the output of the anneal laser system 250 set to what was determined to be the melt threshold of the semiconductor wafer. FIGS. 9A and 9B show the same data in two different formats. Each is a set of profile analyses from many image frames in the high-speed spot sweep direction. In FIG. 9A, the Y-axis is intensity and the X-axis is the position along the anneal image sweep, with the light intensity profiles of a plurality of consecutive sweeps superimposed. In FIG. 9B, the same sweep intensity profiles are shown on a three-dimensional color map where the Y-axis is the position in the sweep direction of the anneal image, the X-axis is the position in the image frame and stage scan direction, and the color indicates the intensity of the optical signal. FIG. 9B is a mosaic of a plurality of image frames captured by the scattered melt detector 280.

[0056] The position within the image frame can be converted to a specific position along the sweep of the anneal image at a certain point in time using knowledge of the speed of the anneal image and the spatial resolution of the ICD. As an example, if the sweep speed of the anneal image is 500 m / s and the spatial resolution of the ICD is 30 μm / pixel, each pixel represents a time of 60 ns, and the sweep image of the anneal image with an exposure of 50 ns can be decomposed into time units of 60 ns. More generally, the scattered melting detector 280, for example, the example shown in FIG. 4A, spatially resolves all points x = v·(t±t d ) to measure the melting behavior and uniformity at a frequency f ≦ 1 / t d , which is practical and useful for the optimization of the anneal laser 250. In an example where a pulse probe separate from the process laser beam is used as shown in FIG. 4B, data can be obtained at a higher frequency.

[0057] In some examples, there may be slight variations in the optical output density of the scanning laser beam 268 over the sweep of the anneal image 270. Therefore, a method of calibrating the anneal laser system 250 may include analyzing the melting signal from the scattered melting detector 280 at specific points of the sweep over multiple sweeps to eliminate the influence from the variability of the optical output along the sweep from the calibration process. As an example, the position 902 in FIG. 9B indicates a specific sweep position where the melting signals from multiple sweeps at position 902 can be analyzed. For example, by adjusting the output level of the anneal laser system 250 stepwise in units of less than 1% of the full output, the percentage of the sweep of the anneal image where melting is detected can be identified. For example, for a given chuck temperature T ch and dwell time t dwell , there exists a melting threshold energy density Em(T ch ,t d ), and the actually supplied energy density is E m ±δ E . The set output E m -δ EThus, since the output sometimes reaches the melting threshold, melting is rarely detected. Melting is always detected at E m +δ E . By varying the output of the scanning laser beam 268 from less than E m -δ E to above E m +δ E over a plurality of stripes and analyzing the percentage of melting detected at a particular sweep position (e.g., position 902), the range below the energy at which melting is never detected and the range above the energy at which melting is detected in all or almost all sweeps can be used to independently identify both E m and δ E . In one example, during the annealing process, the output level of the scanning laser beam 268 is changed stepwise across the sweep of the annealing image. Thus, by repeating the above calibration process at multiple sweep positions, the melting threshold output level setting value can be identified across the sweep of the annealing image.

[0058] Controller

[0059] FIG. 10 is a block diagram illustrating the physical components of one exemplary implementation of one or more computing devices 1000 that can be utilized with the annealing system of the present disclosure, including providing functionality for a controller 300, a stage controller 124, and a melt detector 280. At least one processor 1002 connected to a chipset 1004 is shown. Also connected to the chipset 1004 are a memory 1006, a network adapter 1008, and a communication module 1010. Peripheral devices 1012 and a display 1014 are connected to the computing device 1000. In another embodiment, the memory 1006 is directly connected to the processor 1002. A storage device 1016 is also connected to the chipset 1004.

[0060] The memory device 1016 may be any non-transitory computer-readable storage medium such as a hard drive, a compact disc read-only memory (CD-ROM), a DVD, or a solid-state memory device. The memory device 1016 may contain any software or data that can be stored in a computer storage device that communicates with an annealing system known in the technical field of laser annealing systems. FIG. 10 shows a non-limiting exemplary list of applications 1020 that can be stored in the memory device 1016, as well as a melting threshold 1022 and an annealing laser melting calibration setting value 1024.

[0061] In the illustrated example, the application 1020 includes an ICD control application 1030 and an image processing application 1032 that include instructions for controlling the ICD of a scattered melting detector, e.g., the ICD 406 of the scattered melting detector 280. The ICD control application 1030 may include instructions for controlling functions such as frame rate and shutter speed and instructions for processing signals generated by the FPA(s) of the ICD. The image processing application 1032 may be configured to execute one or more image processing algorithms to analyze light collected from a scattered detector path having a FOV that includes an annealing image.

[0062] As described herein, when the anneal image, e.g., anneal image 270, begins to melt the material on or under the wafer surface 12, a substantial change in the optical properties of the material occurs, thereby significantly increasing the intensity of the light collected by the scatter melt detector. Image processing application 1032 may be configured to identify the increase in the intensity of the light collected by the ICD using any of a variety of image processing algorithms. For example, a fitting algorithm, an edge detection algorithm, a principal component decomposition, a feedforward classification, or the detection of any intensity within the captured image frame that exceeds a melt threshold, or one or more of the intensity values corresponding to the light scattering effect caused by the interface between the liquid and solid phases of the wafer material. In one example, image processing application 1032 may be configured to identify the increase in the intensity of the light collected by the ICD in a region smaller than the anneal image 270 formed by the process beam, e.g., laser beam 268. In one example, image processing application 1032 may compare the intensity of the light to a melt threshold 1022, which may include a threshold corresponding to the intensity of the light captured by the scatter melt detector 280 for melt pools of different sizes or a threshold that relates the light intensity to the intensity in the non-melt state and the laser output level. Referring to FIG. 3C as an example, image processing application 1032 may be configured to detect the onset of melting and the formation of the melt pool 310 and the corresponding interface 312 by detecting the intensity of the light captured along the scatter detector path 402 or 416 that exceeds the relative scatter intensity that exceeds the intensity of the surrounding probe beam. In one example, the threshold may be normalized with respect to the intensity profile of the pre-melt scatter beam light. Image processing application 1032 may be configured to analyze the spatial density or spatial extent of the image frame that includes the intensity of the light that exceeds the melt threshold in order to distinguish between the true melt signal and the discrete false signals caused by non-melt scatterers.The image processing application 1032 may be configured to compare the spatial position of high-intensity signals within an image frame with the known position of the annealed image 270 and / or the position of the annealing illumination line 272 to verify whether the high-intensity signals overlap or are immediately adjacent to the annealed image or the annealing illumination line.

[0063] FIG. 11 conceptually shows a portion of an exemplary image frame 1100 captured by the ICD 406 during the annealing process. FIG. 11 shows a grid of pixels 1102 and a portion of the illumination line 272. The image frame 1100 is composed of a plurality of columns 1104 (two columns 1104a and 1104b are labeled) aligned in the stage scanning direction labeled y in FIG. 11, and a plurality of rows 1106 (one row 1106a is labeled) aligned in the sweep direction of the annealed image labeled x in FIG. 11. To show an example of varying the gray level of pixels along a given column 1104, one column of pixels includes cross-hatched shading.

[0064] FIGS. 12 through 14 show examples of melt detection algorithms that may be configured to be executed by an image processing application such as the image processing application 1032 (FIG. 10). FIG. 12 shows an exemplary method 1200 for generating a control image and associated parameters in a non-melt state. In block 1203, the method 1200 may include annealing the wafer in a non-melt state, capturing an image frame with a scattered melt detector, and recording the annealing laser output level. In one example, the laser output is a voltage V corresponding to the optical output density of an annealing laser beam such as the annealing laser beam 268. refIt can be measured using an on-board detector (OBD) that generates

[0065] In block 1205, method 1200 may include creating a non-molten state reference image I ref from the captured image frame, which may include averaging the pixel values of multiple image frames or excluding any image including the edge of the wafer within the image. In block 1207, the non-molten state reference image is processed to generate a plurality of one-dimensional stage scan direction intensity profiles in the non-molten state. Block 1207 executes a one-dimensional peak detection algorithm or fitting algorithm for one or more columns 1104 of pixels in the stage scan direction, and for each one-dimensional intensity profile, parameters such as amplitude A ref (x), width w ref (x), and centroid or maximum position y 0,ref (x) may be generated. As an example, the width w ref (x) is the distance from y 0,ref (x) to ±y, and the full width is 2·w ref (x). Here, y refers to the stage scan direction and x refers to the sweep direction (see FIG. 11). Thus, block 1207 returns the intensity profile of each column 1104 or group of columns along the y direction. Block 1207 may also include saving a plurality of profiles as sweep profiles.

[0066] In block 1209, an area of interest (ROI) of a non-molten reference image, referred to herein as ROI ref may be identified, and that area corresponds to the spatial position of the irradiation line 272 within the image. ROI ref may be determined by identifying pixels with an intensity above a minimum value located within the scan profile. In one example, ROI ref is determined from the scan profile parameters determined in block 1207 and has an amplitude greater than a value obtained by adding to the black level the number of gray levels corresponding to the background outside the annealed image, such as 2 gray levels on a scale of 255 gray levels of the FPA, and is a pixel of the reference image having a y position within the range of y = y 0,ref (x) ± 2·w ref (x). In block 1211, method 1200 may include creating an average one-dimensional stage scan direction intensity profile, for example, by summing the gray values of the pixels within the ROI ref of each row 1106 and dividing by the number of columns 1104 (see FIG. 11).

[0067] FIG. 13 shows one exemplary method 1300 of a melt detection algorithm, where melt is detected from the non-linearity of the amplitude within the image frame. In block 1303, method 1300 anneals the wafer below, at, or above a melt threshold, captures the image frame using a scattered melt detector, and V, referred to herein, indicates the output level of the annealing laser beam when the image is captured meltIt may also include recording the corresponding OBD voltage, which is called. Block 1303 may also include receiving wafer information such as a specific ID so that the appropriate non-molten state reference data obtained by method 1200 can be referenced. In block 1305, method 1300 may include creating an averaged image from the captured image frames by averaging all of the image frames of the scan performed at the same laser output level, excluding any frames including the edges of the wafer, and saving the averaged image. In block 1307, the averaged image created in block 1305 is associated with a non-molten state reference image corresponding to the same wafer type to generate a melt detection image I data is generated. In one example, the pixel values of the averaged image from block 1305 and the corresponding laser output values are associated with the pixel values of the non-molten state reference image and the corresponding laser output generated in block 1205 of FIG. 12 as follows. Here, I data are the pixel values of the averaged image generated in block 1307, I ref are the pixel values of the non-molten state reference image, V ref and V melt are the corresponding OBD signal values. In block 1309, the melt image values can be compared with a threshold value such as 1.1 to determine whether melting has occurred within the image. For example, the pixel values within the ROI ref in the melt image can be averaged and compared with the threshold value.

[0068]

Number

[0069] In addition to, or instead of, executing block 1309, at block 1311, the method may include applying one-dimensional profile analysis of the melt detection image generated at block 1307 to identify a narrow peak in the stage scan direction having a width smaller than the width of the laser beam. At block 1313, the width w(x) of the identified peak is determined and compared to the width w(x) of a control image in the non-melt state. If the width of the identified peak is less than a smaller width, e.g., 50% of the width of the control image in the non-melt state w ref (x), it is determined that melting has occurred. The above is based on the energy profile of the laser beam, assuming a Gaussian profile in the stage scan direction y, such that melting occurs only in a narrow region of the larger anneal image 270. Thus, a detected peak of intensity whose width in the stage scan direction (perpendicular to the sweep direction) is smaller than the width of the anneal image in the stage scan direction indicates a local melt region or melt pool within the anneal image. As an example, referring to FIG. 11, pixels 1102a through 1102e along column 1104 include cross-hatching indicating different amplitudes of the melt detection image created at block 1307. For example, method 1300 may determine that melting has occurred if pixel 1102c or pixels 1102b through 1102d have an amplitude greater than that of surrounding pixels, e.g., an amplitude exceeding a threshold, and the width of pixels 1102b through 1102c in the stage scan direction y is sufficiently narrow compared to the width of the anneal illumination line 272 or corresponding ROI ref pixels. Steps 1311 and 1313 are most useful for processed wafers with high scatter and can be achieved using a scatter melt detector with a high-resolution optical element.

[0070] FIG. 14 shows a method 1400, which is another exemplary algorithm for detecting melting according to melt statistics. At block 1403, method 1400 anneals the wafer below, at, or above a melt threshold, captures an image frame using a scatter melt detector, and an OBD voltage signal V meltmay also include recording. In block 1405, the standard deviation of pixel values within the ROI ref (determined in step 1209) of the image frame is calculated. In one example, block 1405 may include calculating a fractional standard deviation that is the standard deviation of the ratio of the pixel values within the ROI ref to the average of the pixel values within the ROI ref . A large standard deviation indicates that there are large fluctuations in the intensity values within the ROI ref , and as shown in the example of FIG. 9B, it indicates that the laser output is set to a melting threshold with a large variation in pixels in the molten and non-molten states. In block 1407, the calculated standard deviation is compared with a threshold value to determine the presence of melting. Method 1400 may be most useful for wafers with minimal scattering, such as an untreated Si wafer.

[0071] Methods 1200, 1300, and 1400 may be selectively used according to a specific wafer type and application as a melting detection subroutine for calibrating a laser system and for determining a laser power calibration table or curve for a specific combination of process conditions including wafer type, chuck temperature, and dwell time for a laser annealing process. ref , ROI ref , melt parameters and reference data such as the one-dimensional stage scan direction intensity profile in the non-molten state may be stored in a memory such as the storage device 1016 and be accessible by the image processing application 1032 to execute any image processing algorithm disclosed herein.

[0072] A computing device 1000 having an anneal laser control application 1034 may be configured to control one or more parameters of the preheat and / or anneal laser system 150, 250 to achieve a desired process condition above the melt threshold. The anneal laser control application 1034 may be configured to control the power level of the anneal laser system 250 according to an anneal laser melt calibration setpoint 1024. The anneal laser melt calibration setpoint 1024 may specify laser power level setpoints to achieve sub-melt, melt threshold, or above melt threshold annealing conditions for a particular set of process conditions, such as a particular chuck heater temperature, anneal image dwell time, and type of wafer 10. As described herein, the scattering melt detector 280 may be used to perform a melt calibration of the anneal laser system 250 to determine the anneal laser melt calibration setpoint 1024.

[0073] In one example, the ICD 406 is configured to scan the annealed image at a sweep frequency f sweep The FOV of the scattering melt detector 280 may include at least one anneal radiation line 272 (i.e., at least one sweep of the anneal image), and in some cases may include multiple anneal radiation lines (multiple successive sweeps of the anneal image). Thus, a single image captured by one exposure of the ICD 406 may include at least one sweep, and in some instances multiple sweeps, of the anneal image 270. The image processing application 1032 may use predetermined information regarding the known intensity profile of the material phase transformation and extract one or more statistical correlation functions from the captured intensities of multiple sweeps from one or more image frames to characterize the consistency of the melt process, or act as a multi-sampler on the device wafer, or act as a spatial lock-in frequency and simplify melt detection using a fast Fourier transform.

[0074] In an example where the scatter melt detector of the present disclosure includes a probe separate from the process laser, such as probe light source 410 (FIG. 4B), application 1020 may also include a probe control application 1036 for controlling the intensity, phase, and / or frequency of the pulsed radiation emitted by probe light source 410, and / or the pulse timing of the radiation emitted by the probe, and / or the sensor exposure timing such as the exposure timing of ICD 406 for various lengths or periods. Probe control application 1036 synchronizes the probe and the ICD via a common clock to a sweep frequency f sweep and determines pulse frequencies f Probe and pulse phases φ Probe for the exposure of probe 410 and the ICD as a function of the sweep frequency, sweep phase, and aspects of the phase transition process of the measured wafer material. In one example, since f exposure is sufficiently smaller than f Probe , n exposure per ICD exposure becomes 1. In one example, the frequency ratio R = f sweep / f probes may be any integer, and as a result, statistics regarding the stability of the phase transition process may be extracted. In one example, f sweep may not be present, rather, a random distribution of pulses at known times may be collected for the extraction of broadband statistics. In some examples, f Probe may be a rational beat frequency having f Probe for determining and optimizing the uniformity of the sweep. In one example, the probe pulse frequency f Probe is f sweep . Probe is f Probeis selected to be synchronized with the sweep of the annealing image 270 in the x direction, a subset of the synchronized instants in the set of annealing image sweeps is probed with individual pulses, and multiple pulses are collected by the ICD 406 into a single image frame, whereby the high-bandwidth coherence of the phase transition may be extracted from the contrast of the molten region. In some examples, the probe 410 is configured to emit pulses of light having a duration of from 3 ns to 20 ns, in some examples from 5 ns to 15 ns, in some examples from 5 ns to 10 ns, and in some examples from 5 ns to 7 ns.

[0075] The memory 1006 may hold instructions and data used by the processor 1002. The network adapter 1008 connects the computing device 1000 to a local area network or a wide area network, and the communication module 1010 provides additional channels for wired or wireless communication. As is known in the art, the computing device 1000 may have different and / or other components than those shown in FIG. 10. Additionally, the computing device 1000 may lack certain of the illustrated components. In some examples, the storage device 1016 may be local to and / or remote from the computing device 1000, such as another storage device, a cold storage device, a storage area network (SAN), or a cloud-based storage architecture.

[0076] As is known in the art, computing device 1000 is adapted to execute computer program modules for providing the functions described herein. As used herein, the term module refers to computer program logic used to provide a specified function. Thus, a module may be implemented in hardware, firmware, and / or software. In one embodiment, the program modules are stored in storage device 1016, loaded into memory 1006, and executed by processor 1002.

[0077] Some portions of the present disclosure are described by way of example from the perspective of algorithmic processes or operations. These algorithmic descriptions and representations are commonly used by those skilled in the data processing arts to effectively convey the substance of their work to others skilled in the art. These operations, while described functionally, computationally, or logically, are understood to be implemented by a computer program comprising instructions for execution by a processor or equivalent electrical circuit, microcode, etc. Further, it has proven convenient at times to refer to these configurations of functional operations as modules.

[0078] FIG. 10 shows a single computing device 1000 and storage device 1016, but it will be understood that the functions and storage provided by computing device 1000 and storage device 1016 may be implemented in any number of computing devices and storage devices. By way of example, a first computing device 1000 may be used to implement controller 300, and one or more other computing devices 1000 may be used to execute other functions disclosed herein, such as stage controller 124, chuck heater, melt detector 280, etc.

[0079] Computing device 1000 may be configured to communicate with other computing devices of system 100 via one or more networks that may include any combination of local area networks and / or wide area networks, using both wired and / or wireless communication systems. In one embodiment, the network uses standard communication technologies and / or protocols. For example, the network includes communication links using technologies such as Ethernet, 802.11, worldwide interoperability for microwave access (WiMAX), 3G, 4G, code division multiple access (CDMA), digital subscriber line (DSL), and examples of networking protocols used for communication over the network include MPLS (multiprotocol label switching), TCP / IP (transmission control protocol / internet protocol), HTTP (hypertext transport protocol), SMTP (simple mail transfer protocol), and FTP (file transfer protocol). Data exchanged over the network may be represented using any suitable format such as hypertext markup language (HTML) or extensible markup language (XML). Those skilled in the art will recognize that encryption using other suitable technologies is appropriate for various applications based on the nature of the network.

[0080] Examples of the present disclosure include a method for analyzing scattered image data. The method includes receiving a scattered image of an area of a semiconductor surface captured during a laser annealing process of a semiconductor, and identifying a solid-to-liquid material phase transition of the semiconductor material from the image, where identifying includes identifying statistical, time-series, or geometric non-linear variations of the image intensity as a function of one or more annealing process control parameters. In some examples, the one or more annealing process control parameters include laser output or optical output density.

[0081] Examples of the present disclosure also include a method for performing melting calibration for a laser annealing process of a semiconductor material. The method includes annealing an area of the semiconductor material with an annealing laser beam from an annealing laser, where the output level of the annealing laser is varied stepwise above and below a melting threshold output level at which a transition from the solid phase to the liquid phase of the semiconductor material occurs, capturing a scattered image of the area of the semiconductor material during annealing, and determining at least one melting threshold setpoint output level of the annealing laser from the captured scattered image.

[0082] Examples of the present disclosure include a method for controlling an operating parameter of an annealing laser during a laser annealing process. The method includes controlling an operating parameter of the annealing laser during an annealing process of a semiconductor material according to a laser annealing temperature calibration created by performing the method described in the previous paragraph.

[0083] Examples of the present disclosure also include machine-readable instructions configured to cause a processor of a controller of a laser annealing system to control an annealing laser according to an annealing process parameter determined according to the method disclosed herein and at least one annealing laser melting threshold calibration setpoint.

[0084] Examples of the present disclosure include a method of forming a non-molten state reference image and parameters for use with a melt detection algorithm. The method includes annealing a non-molten state wafer, capturing an image frame using a scattered melt detector, recording an annealing laser output level, creating a non-molten state reference image from the captured image frame, processing the non-molten state reference image to generate a plurality of one-dimensional stage scan direction intensity profiles in the non-molten state, identifying a region of interest (ROIref) in the non-molten state reference image, and creating an averaged one-dimensional stage scan direction intensity profile.

[0085] Examples of the present disclosure include a method of detecting whether a wafer material has melted during an annealing process. The method includes annealing the wafer below, at, or above a melt threshold, capturing an image frame using a scattered melt detector during annealing, recording a laser output level, creating an averaged image from the captured image frame, and creating a melt detection image that associates the averaged image with a corresponding non-molten state reference image of the same wafer type.

[0086] The method according to paragraph

[0082] , further comprising comparing the values of the melt detection image with a threshold to determine whether melting has occurred in one or more regions of the image.

[0087] The method according to paragraph

[0082] , further comprising applying a one-dimensional profile analysis of the melt detection image to identify a narrow peak in the stage scan direction having a width smaller than the width of the process laser beam.

[0088] The non-molten state reference image is obtained by performing any method using any system disclosed herein. The method according to paragraph

[0082] .

[0089] Examples of the present disclosure also include a method for detecting whether a wafer material has melted during an annealing process. The method includes annealing the wafer below, at, or above a melting threshold, capturing an image frame during annealing using a scattered melt detector, calculating a statistical variation of pixel values within a region of interest ROIref of the image frame that is an area of an annealing laser image formed on the wafer, and comparing the calculated statistical variation to identify the onset of melting.

[0090] The foregoing is a detailed description of illustrative embodiments of the present disclosure. In this specification and the appended claims, conjunctive language such as "at least one of X, Y, and Z" and "one or more of X, Y, and Z" is to be construed, unless specifically stated or indicated otherwise, as meaning that each item in the conjunctive list can be present in any number, excluding any other item in the list, or in any number in combination with any or all of the other items in the conjunctive list, and each of those can also be present in any number. It should be noted that applying this general rule, the conjunctive phrase in the foregoing example where the conjunctive list consists of X, Y, and Z would include, respectively, one or more of X, one or more of Y, one or more of Z, one or more of X and one or more of Y, one or more of Y and one or more of Z, one or more of X and one or more of Z, and one or more of X, one or more of Y, and one or more of Z.

[0091] Various modifications and additions can be made without departing from the spirit and scope of the present disclosure. Each feature of the various embodiments described above may, if necessary, be combined with the features of other described embodiments in order to provide combinations of a number of features in related new embodiments. Further, while the above describes several other embodiments, what is described herein is merely an example of the application of the principles of the present disclosure. Additionally, while a particular method herein may be illustrated and / or described as being performed in a particular order, the ordering is highly variable within the scope of those skilled in the art in order to achieve aspects of the present disclosure. Accordingly, this description is intended to be construed only as an example and is not intended to limit the scope of the present disclosure.

Claims

1. A method for determining that a material phase transition from solid to liquid of a semiconductor material has occurred during a laser annealing process, comprising: forming an annealing image on the surface of the semiconductor material by an annealing laser beam from an annealing laser; forming a scattered image of a region of the semiconductor surface including the annealing image; and identifying a material phase transition of the semiconductor material from solid to liquid according to the intensity of light collected in the scattered image. A method comprising the above steps.

2. The method according to claim 1, wherein the identifying includes detecting an intensity in the scattered image that reaches or exceeds a threshold value, and the threshold value corresponds to the start of the phase transition of the semiconductor material from solid to liquid.

3. The method according to claim 2, wherein the identifying includes detecting an intensity of the scattered image in a region of the scattered image that is within the annealing image and smaller than a region of the annealing image.

4. The method according to claim 1, wherein the identifying includes identifying a discontinuity in the intensity of light collected in the scattered image as compared with a control image in a non-melted state.

5. The method according to claim 1, wherein the identifying includes identifying a non-linear scaling relationship between the intensity of the light and a process parameter.

6. The method according to claim 5, wherein the process parameter is an output level of the annealing laser or a local output density of the annealing laser.

7. The method according to claim 1, wherein the annealing laser process includes sequentially sweeping the annealing image across the semiconductor surface, each sweep forming an annealing irradiation line, and forming a scattered image includes forming an image of at least one annealing irradiation line.

8. The method according to claim 7, wherein the identifying step includes analyzing the intensity of light along the annealing irradiation line to determine the degree of material phase transition of the semiconductor material along the annealing irradiation line.

9. The method according to claim 7, wherein the identifying step includes analyzing the intensity of light along the annealing irradiation line to determine the proportion of the annealing irradiation line at which the start of the phase transition from solid to liquid occurs.

10. The identifying step of claim 7 includes analyzing the intensity of light at a first position along the annealing irradiation line in the plurality of sequential sweeps of the annealing image to determine at least one of the melting threshold output level or the statistical output variation of the annealing laser.

11. The laser annealing process includes performing sequential scans across the semiconductor surface at a scan frequency f ｓｗｅｅｐ and the method further includes irradiating the semiconductor surface with a probe light source having radiation pulses at a pulse frequency f Ｐｒｏｂｅ where f Ｐｒｏｂｅ < f ｓｗｅｅｐ The method according to claim 1.

12. The identifying of claim 1 includes applying one or more of a fitting algorithm, an edge detection algorithm, eigenbasis decomposition, feed-forward classification, or detection of any intensity in an image frame exceeding a captured threshold.

13. The identifying of claim 1 includes identifying statistical, time-series, or geometric non-linear variations of image intensity as a function of one or more annealing process control parameters.

14. The one or more annealing process control parameters of claim 13 include laser output or optical output density.

15. The identifying includes creating a melt detection image that associates the scattered image with a corresponding non-melt state reference image of claim 1.

16. The method of claim 15 further includes comparing the values of the melt detection image with a threshold to determine whether melting has occurred in one or more regions of the melt detection image.

17. The method of claim 15 further includes applying a one-dimensional profile analysis of the melt detection image to identify a narrow peak in the stage scan direction having a width smaller than the width of the process laser beam.

18. The non-melt state reference image is obtained by performing the steps of annealing a non-melt state wafer, capturing an image frame using a scattered melt detector, recording the annealing laser output level, creating a non-melt state reference image from the captured image frame, processing the non-melt state reference image to generate a plurality of one-dimensional stage scan direction intensity profiles of the non-melt state, identifying a region of interest (ROIref) of the non-melt state reference image, and creating an average one-dimensional stage scan direction intensity profile of claim 15.

19. The identifying Calculating statistical variations in pixel values within a region of interest (ROIref) of the image frame, where the anneal laser image is the region of the image formed on the wafer, and Identifying the start of the melting by comparing the calculated statistical variations The method according to claim 1, comprising:

20. A stage configured to support a semiconductor wafer, An anneal laser configured to anneal the semiconductor wafer, An image capture device including a focal plane array (FPA), and a scattering optical arrangement configured to block or avoid specular reflection of the beam formed by the anneal laser and form a scattered image of the surface of the semiconductor wafer on the FPA, and a scattering melt detector A scattering melt detection system, comprising:

21. A processor and a machine-readable recording medium including program instructions executed by the processor to perform the operations according to any one of claims 1 to 19, the system according to claim 20, further comprising:

22. An anneal laser, A storage device including at least one anneal laser melt threshold calibration setting value for specifying an operating parameter of the anneal laser to raise the temperature of a local region of a semiconductor material to a melting threshold under a predetermined set of anneal process conditions, the setting value being derived from a scattered image of the semiconductor material obtained during a laser anneal process of the semiconductor material, a recording device, A controller configured to control the anneal laser according to the anneal laser melt threshold calibration setting value A laser anneal system, comprising:

23. The system according to claim 22, wherein the calibration setting value is determined by a process including the steps according to any one of claims 1 to 19.

24. At least one anneal laser melt threshold calibration setting value for specifying an operating parameter of an anneal laser to raise the temperature of a local region of a semiconductor material to a melting threshold under a predetermined set of anneal process conditions, the calibration setting value being derived from a scattered image of the semiconductor material obtained during a laser anneal process of the semiconductor material A non-transitory machine-readable recording medium, comprising:

25. The non-transitory machine-readable recording medium according to claim 24, wherein the calibration setting value is determined by a process including the steps according to any one of claims 1 to 19.

Images