TW200528223A - Laser scanning apparatus and methods for thermal processing - Google Patents

Abstract

Apparatus and methods for thermally processing a substrate with scanned laser radiation are disclosed. The apparatus includes a continuous radiation source and an optical system that forms an image on a substrate. The image is scanned relative to the substrate surface so that each point in the process region receives a pulse of radiation sufficient to thermally process the region.

Description

200528223 IX. Description of the invention: “L is a member of the family” Xuanyuan Beij This application is a part of the application for joint examination (No. 10 / 287,864) on November 6, 2002. 5 Field of the Invention The present invention relates to a laser scanning device and a method for heat treatment of a substrate, and particularly a semiconductor substrate on which an integrated element or a circuit is formed. L Previous Winter; j Background of the Invention 10 Integrated Circuits (ICs) include subjecting-semiconductor substrates to, for example, light & coating, lithographic exposure, light and development, afterglow, polishing and heating or "heat treatment," Several processes. In some applications, heat treatment is used to activate dopants in the substrate doped regions (such as source and drain regions). Heat treatment includes different loading (2 cooling) techniques such as rapid thermal annealing (RTA) and Laser Heat Treatment (LTP). When 15 is used for heat treatment, this technique is sometimes referred to as "laser treatment," or "laser annealing." " Different technologies and systems for laser processing of semiconductor substrates are known and used in circuit (IC) manufacturing. H-ray processing is more completed in a single cycle, so that the temperature of the material is annealed to the annealing temperature and then returned. Initial (eg ambient 20) temperature. If the heat treatment cycle required for the activation or annealing can be fixed at one millisecond or less, the performance of ic can be substantially improved. Thermal cycle times of less than one microsecond can already be obtained by irradiating one or more circuits with a pulse laser. The example system for laser heat treatment using a -pulse laser source has been published.

Thermal is in U.S. Patent No. 6,366,308 B1, titled ". However, in shorter spokes

Processing Apparatus and Method, explains that the pulses can be reduced, and the circuit element itself will easily cause its own temperature difference. For example, '-the polycrystalline material existing on the thick-field oxide insulator is heated more quickly than the shallow junction on the surface of the broken wafer. A longer radiation pulse can obtain a more uniform temperature distribution, which is due to the heating depth It is deeper and has a longer pulse interval for lateral heat transfer to average the temperature of the entire circuit. However, it is impractical to extend the length of the laser pulse so that the period exceeds one micro 10 seconds and exceeds 5 cm2 or more circuit area, because the energy of each pulse becomes too high, and the laser and its power supply The high energy to be provided becomes too large and expensive. Another method using pulsed radiation is to use continuous radiation. An embodiment of a heat treatment device using a continuous radiation source in the form of a laser diode is disclosed in U.S. Patent Application No. 09 / 536,869, entitled " Apparatus Having Line Source of Radiant Energy f〇r

"Exposing a Substrate", whose application was filed on March 27, 2000 and transferred to the same assignee as this application. The laser diode rod array can obtain an output power in the range of 100 W / cm and can form a micron width Line image. 20 It is also very efficient to convert electricity to radiation. In addition, because there are many diodes operating at slightly different wavelengths in the rod, it can form a uniform line image. However, a polar body is used as A continuous radiation source is only suitable for some applications. For example, when annealing source and gate regions below a micron depth, it is better that the radiation is not absorbed by the stone that exceeds this depth. Unfortunately, A typical laser diode with an operating wavelength of 0.8 micrometers has an absorption depth of 20 micrometers at room temperature. Therefore, it is mostly used for the heat treatment of the top layer of the substrate (such as narrower than-micrometers). The bipolar substrate_ penetrates into the broken wafer far beyond the required or desired depth. This will increase the total power required. Although "thin absorber layers can be used to reduce this problem, it is more Added original in process Very complicated complexity. [Summary of the Invention] Summary of the Invention 10 An aspect of the present invention is a device for heat treating a region of a substrate. The device includes a device capable of providing a first beam intensity curve and having a property suitable for heating the substrate. A continuous light source of continuous light emission at a region wavelength. An optical system is arranged at the back of the continuous radiation source to make it receive the light under control and form a second radiation, which forms an image on the substrate. One specific example is 15 cases. In the image, the image is a line image. The device also includes a stage suitable for supporting the substrate. At least one of the optical system and the stage is suitable for scanning the image along the scanning direction for the substrate, using radiation Pulse heating the region to a temperature sufficient to process the region. Another aspect of the present invention is a method of heat treating a region of a substrate. The method includes generating a continuous radiant light having a wavelength that can heat the region of the substrate, and then Scan the area covered by the radiation along the scanning direction so that each point in the area receives a thermal energy that can process the area of the far substrate. The diagram briefly illustrates the first 1A The figure is a simplified diagram of a general specific example of the device of the present invention; 7 200528223 FIG. 1B illustrates a specific example of an ideal line image with a long size L1 and a short size L2 formed on the substrate by the device of FIG. 1A; FIG. 1C is A two-dimensional figure represents the intensity distribution accompanying the actual line image. 5 Figure 1D is a simplified diagram of a specific embodiment of the optical system of the device of Figure 1A, which includes a cone lens to form a line image on the substrate; Figure 2A is a The schematic diagram illustrates a specific implementation example of the laser scanning device in FIG. 1A, which further includes an optical converter disposed between the radiation source and the optical system. FIG. 2B is a schematic diagram illustrating the optical converter in FIG. 2A How to change 10 beam intensity curve of variable radiation; Figure 2C is a cross-sectional view of a specific example of a converter / optical system including a flat Gaussian beam intensity curve converter; Figure 2D is a conversion from Figure 2C Of the modeled beam intensity curve of non-peripheral vignettes radiated by the reflector / optical system; Figure 2E is similar to Figure 2D. The peripheral vignette aperture removes the edge beam to reduce the intensity peaks at the endpoints of the image. ; Figure 3 is a category The schematic diagram of the device similar to FIG. 1A, with additional elements representing different embodiments of the invention; FIG. 4 illustrates a specific embodiment of the reflected radiation light monitor of FIG. 3 whose 20 incident angle Φ is equal to or close to 0 ° FIG. 5 is a detailed enlarged view of a specific example of the implementation of the analysis system 300 of FIG. 3 for measuring the temperature of the scan image 100 on or near the substrate. Figure 6 is the intensity-temperature blackbody temperature curve (picture) at 1410 ° C. The temperature is slightly higher than the temperature of the dopant in the source and drain 8 200528223 region used to activate a semiconductor transistor; Figure 7 is a detailed enlarged isometric view of a substrate with incident and reflected laser light with a calibration feature in a 45-degree plane showing the characteristics of the light thumb pattern relative to the light thumb pattern. 5 Figure 8 shows a 10.6-micron wavelength laser. Radiation light reflected from the following surface ρ and s-polarization direction reflectance versus incident angle (a) pure Shixi, made) 0.5 micron oxide insulation layer on the top of Shixi, (c) 0.5 micron oxidation on silicon Polysilicon flow channel with 0.1 micron at the top of the insulating layer, and (d) silicon oxide layer of infinite depth; FIG. 9 is a specific example of the device of the present invention for processing a gate pattern semiconductor wafer tape 10 type substrate 60 formed thereon The top view illustrates the operation of the substrate in an optimal radiant light structure. Figure 10 is a plan view of a substrate illustrating a pattern of a scan image folded on the surface of the substrate. Figure 11 is a cross-section of a specific example of an optical system. Figure, basin includes 15 movable Sighting mirror; Figure 12 is a plan view of the four substrates with the ability to rotate and linearly move on the stage. 'Plans that generate a spiral cat pattern on the substrate. Figures 13A and 13B are plan views of the substrate. — "Pattern" where the cat scan path is separated by a space that allows the substrate to cool before scanning adjacent to the scan path 20; Figure 14 shows the device's spiral scan method, optical cat scan method, and fold scan Method in the substrate factory], the simulation output of the time versus the microsecond dwell time; Figure 15 is-similar to the detailed implementation example of the MLTP system 9 200528223 Enlarged diagram, which also includes configuration to get reflection A circular optical system that radiates light and guides it back to the substrate as circulating light; Figure 16 is a cross-sectional view of a specific embodiment of the circular optical system of Figure 15, which includes a right-angle mirror and a collecting / focusing lens; Fig. 17 is a cross-sectional view of a modified embodiment of the cyclic optical system in Fig. 16, in which the right-angle mirror is displaced (deviated) from the axis AR by the amount of AD, resulting in direct incident and cyclic radiated light. The angle of incidence has a deviation; Figure 18 is a simplified cross-sectional view of a specific example of the circulating optical system in Figure 15 'which includes an enlarged relay group and a roof mirror; 10 Figure 19 is another implementation of the circulating optical system in Figure 15 A schematic cross-sectional view of a specific example, which includes a collimating / focusing lens and a grating; and FIG. 20 is a schematic cross-sectional view of a specific example of an LTP system using two laser diode arrays and two opposite LTP optical systems The substrate is irradiated at a similar incident angle across the substrate normal. • The different components described in the drawings are only used for representation and not to scale. 2 ° Certain parts may be enlarged and others may be reduced. These drawings are used to illustrate different implementation methods of the present invention, which can be understood and used by those skilled in the art. [Thirst application method] 〇 Detailed description of the preferred embodiment In the following detailed description of the specific examples of the present invention, refer to the accompanying drawings-and not the specific examples that illustrate the feasibility of the present invention. These specific examples are fully explained so that those skilled in the art can practice the present invention, and it should be noted that, without departing from the scope of the present invention, it can be used for 200528223 specific examples and make other changes. Therefore, the following detailed description is not limited to the concept, and the scope of the present invention is only defined by the scope of the attached patent application. General Apparatus and Method FIG. 1A is a simplified diagram of a laser scanning apparatus of a general specific example of the present invention. The device 10 of Fig. 1A includes a continuous radiation source 12 that radiates along an optical axis A1 and has continuous output light 14A with output power and beam intensity curve P1 perpendicular to the optical axis. In one embodiment, the continuous radiation 14A is collimated. Also in a specific embodiment, the radiation source 12 is a laser and the radiation light 14A is a laser light. In addition, in this embodiment, the radiation source 12 is a carbon dioxide (c02) laser with an operating wavelength of 10 "between about 9.4 microns and about 10.8 microns. (^ 〇2 lasers can convert electrical energy very efficiently Is radiation, and its output light is typically very homogeneous, so curve P1 is a Gaussian curve. In addition, as described below, the infrared wavelengths generated by this CO2 laser are suitable for processing (such as heating) silicon (such as Silicon substrate of a silicon wafer). Also in a specific embodiment, the radiated light 14A 15 is linearly polarized and can be manipulated so that the radiation incident on the substrate includes a single P-polarized state P, or a single S-polarized state s Or both. Since the radiation source 12 emits a continuous radiation source 14A, it is referred to herein as a "continuous radiation source". Generally, the radiation light 14A includes radiation whose wavelength is absorbed by the substrate, so it is suitable for heating The substrate. 20 The device also includes an optical system 20 that converts (e.g., focuses or shapes) light light μα into radiant light 14B at the rear section of the wheel source 12. The optical system 20 may be composed of a single element (such as a single lens element or mirror) By The optical system 20 may also include a movable element, such as a scanning lens, which will be described in more detail below. 11 200528223 The clothing set 10 further includes an upper surface 42 on the rear section of the optical system 2 (). Chuck 40. The chuck 40 is supported by the carrier post supported by the platen 50. In another embodiment, the chuck is tilted into the carrier 46. In another embodiment, The stage 46 is movable. In addition, in one embodiment, the base plate stage 46 can be rotated along the-or M, γAZ axis. The upper surface 42 of the chuck can support a surface 62 having a surface normal N and Substrate 60 in cross section 63. In a conventional embodiment, the substrate 60 includes a reference feature 64 to facilitate calibration of the substrate in the device 10, as described below. In an embodiment, the reference feature 64 is also used to The crystal orientation of a single crystal substrate 60 is defined. In one embodiment, the substrate 60 is a single crystal silicon wafer, such as in Semiconductor Equipment and Materials International (SEMI), 3081 Zanker Road, San Jose 95134. The obtained literature #semi M1 -600, "Spec certifications for Polished Monocrystalline

Silicon Wafers ", which is also incorporated herein by reference. 15 In addition, in one embodiment, the substrate 60 includes source and drain regions 66A and 66B formed on or near the surface 62 as Part of a circuit (such as a transistor) 67 formed in a substrate. In one embodiment, the source and drain regions 66A and 66B are shallower and have a depth of one micron or less into the substrate. 20 Axis A1 and the substrate The normal line N forms an angle Φ, which is the incidence angle of the radiated light 14B (and the axis A1) and the substrate surface normal line N as φ. In an embodiment, the radiated light 14B has an incident angle of φ > 0 to ensure that The radiated light reflected from the substrate surface 62 does not return to the radiation source 12. Generally, the incident angle can be changed in the range of 0. φ < 90 °. However, in some applications, 12 200528223 radiation is selected within this range. It is beneficial to make this arrangement, as described in more detail below. "In the embodiment, the device 10 further includes a connection via a connection line (line P to the light source 12 and connection via line 78). Control to the stage controller 76 °° 70. The stage controller 76 is The line 80 is connected to the carrier 46 to control the movement of the carrier 4. In the specific embodiment, the controller controls the radiation source 12, the carrier controller% and the optical system 20 through signals, 92 and 94. (Such as the movement of each element therein). In a consistent embodiment, one or more of the wires 72, 78, 80, and 82 are electric wires and the corresponding one or more signals 90, 92, and 94 are electronic signals, however In another embodiment, one or more of the wires are optical fibers and the corresponding one or more signals are optical signals. In a specific embodiment, the controller 70 is a computer such as a personal computer or a workstation. The "Electrical Device" can be obtained from any of a number of well-known computer companies such as Dell Computer, Inc, Austin Texas. The controller 70 preferably includes 15 such as 1ntel PENTIUM series, or aMD Κ6 * Κ7 processors, etc. Any commercially available microprocessor, a bus suitable for connecting the processor to a memory element such as a hard disk component, and suitable input and output components (such as a keyboard and a display, respectively). Continue to refer to Figure 1A, Radiation Light 14B is guided by optical system 20 along axis A1 20 is led to the substrate surface 62. In an embodiment, the optical system 20 focuses the radiant light 14B to form an image 100 on the substrate surface 62. The term "image" is used herein, which generally refers to the radiation light The distribution of the light formed by 14B on the substrate surface 62. Therefore, the image 100 need not have the accompanying objects in the general concept. In addition, the image 100 need not be shaped by the beam line to a point of focus 13 200528223. For example, Like a circular light spot formed by a circularly symmetrical optical system focusing normal incident light, the image 100 can form an elliptical light spot by the deformed optical system 20. Similarly, the term "image" includes the light distribution formed by the substrate 60 intercepting the light 14B on the substrate surface 62. 5 The image 100 can have any number of shapes, such as square, rectangular, oval, etc. Similarly, the image 100 can have a variety of different intensity distributions, including the distribution of related uniform line images. FIG. 1B illustrates that the image 100 is a line image in a specific embodiment. An idealized line image 100 has a long dimension (length) L1, a short dimension (width) L2, and a uniform (ie, flat) intensity of 10. In fact, the line image 100 is not completely uniform due to the diffraction effect. Figure 1C is a two-dimensional diagram showing the intensity distribution of the actual line image. For most applications, the integral section at the short dimension L2 need only be completely uniform at the long dimension L1 $, and the uniformity of the integral intensity distribution is about ± 2% of the operable part of the image. 15 With continued reference to Figures 1B and 1C, in one embodiment, the length li ranges from about 1.25 cm to 4.4 cm and the width is about 50 microns. In another embodiment, the length L1 is 1 cm or less. In addition, in an embodiment, the image 100 has an intensity range from 50 kW / cm2 to 150 kW / cm2. The intensity of the image 100 is determined according to the energy required to be provided to the substrate in practical applications, the image width L2, and the speed at which the image scans the substrate surface 62. Figure 1D is a simplified diagram of an optical system 20 including a conical lens M1, M2, and 馗 3 to form a line image on the surface of the substrate. The optical system 20 of FIG. 1D illustrates how a reflection cone segment can be used to focus a collimated light into a line image 100. In one embodiment, the optical system 20 includes a parabolic cylindrical lens 14 200528223 segments M1 and M2 and a conical lens segment M3. The cone segment M3 has an axis A3 associated with the entire cone (shown in dotted lines). The axis A3 is parallel to the collimated light 14A and is in a direction along the substrate surface 62. The line image 100 is formed on the substrate surface 62 along the axis A3. The advantage of this arrangement of the optical system 20 5 is that it produces a narrow diffraction-limiting image 100 with minimal variation at the angle of incidence φ. The length L1 of the line image is mainly determined by the incident angle φ and the size of the collimated light measured in the y direction. Different incident angles φ can be achieved by converting different cone lens segments (such as mirror M3) into the radiated light 14A. The length L1 of the image 100 can be changed by using a collimating lens 10 that can be adjusted (such as zoom) to change the size of the collimated light. Continuing to refer to FIG. 1D, in an embodiment, the size of the collimated light 14A can be changed by using parabolic cylindrical mirrors M1 and M2. The collimated light 14A is first brought by the forward cylindrical parabolic mirror M1 to focus on the line at point F. Before the focal point of the point F is reached, the 'focused light 14A' is intercepted by the negative parabolic cylinder M2 to collimate the focused light. The two cylindrical parabolic mirrors M1 and M2 only change the width of the collimated light in the y-direction. Therefore, the parabolic mirrors M1 and M2 also change the length L1 of the line image 100 on the substrate surface 62, but do not change the width L2 of the line image along a plane perpendicular to the figure. Also shown in Figure 1D are the alternate parabolic mirrors mi, and m2, and a 20-cone conical lens M3 ', all of which can be taken to the predetermined path in the optical path using, for example, the marked wheels 106, 108, and 110. Fixed position. Referring to FIG. 1A again, in an embodiment, the substrate surface 62 is scanned under the image 100 using one of several scanning patterns described in detail below. Scanning can be achieved by several methods, including moving the substrate stage 46 or radiating the light 14B. Therefore, the term "scanning seedlings" as used herein includes the movement of the image relative to the surface of the substrate, no matter how it is achieved. The use of concealment-continuous radiation on the substrate surface 62, as selected on = Areas such as the areas 66A and _, or-or more such as the 5th fast path of the transistor ⑺, get a radiation pulse at each irradiated point on the substrate. In one embodiment, a dwell time of -200 microseconds is used. (That is, the image stays at one point), the energy obtained by each scanning seedling on the substrate in a single scanning seedling = the range is 5. overlapping overlapping scanning seedlings increases the total absorption energy. Therefore, the device 10 can make a continuous radiation source more than a pulsed radiation source. 10 is suitable for providing a pulse or burst that can be controlled with sufficient energy to process-or more areas, such as a circuit or a circuit element formed in or on it. The term treatment used at each point on the radiation plate includes other terms such as selective refining, explosive recrystallization, and doping activation. The term "treatment used here" does not include laser vapor-plated 15 substrates. Laser cleaning or The resistive lithography exposure and subsequent chemical activity are reversed. Through the implementation of the example, the image 100 scans the substrate 60 to provide enough ... month b, plus one or more surface temperatures to process the In one or more regions or B source activation and dopant regions and B activation doping or change the crystal structure of the Xuan or more regions. In a specific embodiment 20 of a heat treatment, the device 10 is used for rapid heating and Cool and thereby activate shallow source and; and electrode regions, such as source and drain regions 66A and 66B, having transistors 67 one micrometer or more into the substrate from the surface 62. The placement 10 has several A different specific example will be explained by the embodiment discussed below. 16 200528223 A specific example with an optical converter In an embodiment shown in FIG. 1A, the beam intensity curve P1 of the radiated light 14A is uneven. This situation may occur when the radiation source 12 is a completely synchronized laser, and the energy distribution generated by the collimated light is a 5 Gaussian distribution, which results in similar energy to when the collimated light is developed on a substrate Lining. In some applications, it is more desirable to Make the radiated light 14A and 14B more uniformly distributed and change the size, so that the image 100 has the intensity distribution and size of the substrate suitable for heat treatment δHai application. Figure 2A is a schematic diagram illustrating the lightning in Figure 1A The embodiment 10 of the radio-scanning device 10 includes a light converter 150 arranged between the optical system 20 and the continuous radiation source 12 along the axis A1. The light converter 15 will have a beam intensity curve The light beam 14A of P1 is transformed into light beam 14A ′ having a beam intensity curve P2. In an embodiment, the light converter 15 and the optical system 20 are combined to form a single converter / optical system. Although the light converter The 15 display shows that 15 is arranged before the optical system 20, and it can also be arranged after the optical system. Figure 2B is a diagram illustrating how the light converter 15 can radiate the light 14A having the beam intensity curve P1. It is converted into radiated light 14A ′ having a beam intensity curve P2. The radiant light 14A and 14A 'show that it is composed of a light beam no whose distance is related to the relative intensity distribution in the radiant light. The light converter 150 adjusts the relative distance (ie, the density) of the 20 light beam 70 to change the beam intensity curve P1 of the radiated light 14A to form a radiated light 14A with a beam intensity curve P2. In one embodiment, the light converter 150 is a refractive, reflective, or refracting lens system. Figure 2C is a converter / optical system 16 with a converter 150 that converts the radiated light 14A with a Gaussian beam intensity curve P1 into a flat (ie uniform) beam intensity curve 17 200528223 P2, and a formation A cross-sectional view of the optical system 20 focusing the radiated light 14b and the line image 100. The converter / focusing system 160 of FIG. 2C includes cylindrical lenses L1 to L5. Here, "lens" can refer to individual lens elements or a group of lens elements, that is, a lens group. The first two cylindrical lenses L1 and L2 use 5 to shrink the diameter of the radiated light 14A, and the cylindrical lenses L3 and L4 are used to increase The radiated light is about the original size, but because of the spherical phase difference in the lens, the radiated light beam intensity curve A14 is changed. The fifth cylindrical lens L5 is used as the optical system 20 and is rotated 90 ° relative to other lenses., Therefore, its effectiveness is outside the plane of the figure. The lens L5 forms radiant light 14B and sequentially forms a line image 100 on the substrate 60 10. In an embodiment, the converter / optical system 160 in FIG. 2C also includes a Configured at the peripheral edge of the lens L1, the vignetting aperture 18 °. This removes the outermost beam of the incident light 14A, which will be over-corrected by the spherical difference of this system, which will cause other flat beam intensity curves. The intensity of edge 15 suddenly increases. Figure 2D is a graph of an exemplary beam intensity curve p2 of non-peripheral dark-angle uniform radiant light 14A 'formed by a typical light converter 150. Typically, a flat beam The degree curve P2 has a flat portion 200 at most of its length, and includes an intensity peak 21 near the end of the beam 204. By removing the outer portion of the beam by the peripheral dark angle 20 aperture 180, this can also be compared The beam intensity curve P2 of the uniform sentence is shown in Fig. 2E. Although the outermost beam of the peripheral dark angle radiation UA can be used to avoid the increase in the intensity of the beam end 204, sometimes the intensity near the beam end It is desirable to increase to generate uniform heating. The heat at the beam end depends on the direction of the parallel and perpendicular 18 200528223 linear image 100 (Figure 1B). Therefore, a higher intensity at the beam end 204 can help compensate Higher heat loss. This results in a more uniform temperature distribution curve on the substrate when the substrate 60 is scanned at image 100. More specific example 5 Figure 3 is a device 10 similar to Figure 1A Schematic diagram, which includes several additional elements located on the top of the figure and above the substrate 60. These additional elements, individually or in combination, have been included to illustrate additional embodiments of the present invention. In the following In each specific example, How many additional elements are introduced in Figure 3 to operate are clear to those skilled in the art10, and whether or not they have been described in the previous implementation examples, they are also required in the implementation examples that will be explained For simplicity, Figure 3 has been shown to include all the elements required in these additional implementation examples, and some of these implementation examples are based on the previously described specific examples. These additional implementation examples are as follows Be explained.

15 MB With reference to FIG. 3, in an embodiment, the device ι〇 includes an attenuation arranged at the rear section of the X-ray source 12 | § 226 'Selectively subtract 1A and 14A of light according to the position of the attenuator, and radiate Light 14A 'or light 14B. In an embodiment, the radiated light 14A is polarized in a specific direction (such as p, s or a combination of both), and the attenuator 226 includes a polarizer 227 that can be rotated relative to the polarization direction of the radiated light to attenuate it. The light. In another specific embodiment, the attenuator 226 includes at least one movable garment minus element or a programmable controllable attenuation disk with a plurality of attenuator elements. In an embodiment, the attenuator 226 is connected to the control 19 200528223 via the line 228 and controlled by the signal 229 of the controller. Quarter wave plate 5

10 In another embodiment, the light 14 A is linearly polarized light, and the device 10 includes a quarter-wave plate 23G after the radiation source 12 to convert the linearly polarized light into circularly polarized light. The quarter-wave plate 230 functions in combination with the subtractor 226 in the embodiment including the polarizer 227 to prevent the radiation light reflected or scattered by the substrate surface 62 from returning to the radiation source. Specifically, in the return path, the reflected circularly polarized radiation light is converted into linearly polarized light, which is then blocked by the polarizing plate 227. This configuration is particularly useful when the angle of incidence is at or near zero (that is, at or near normal incidence). Optical energy monitoring system In another embodiment, the device 10 includes a light energy monitoring system 250 arranged on the axis A1 of the rear section of the radiation source 12 to monitor the energy of individual light beams. The system 250 is connected to the controller 70 via a line 252 and is provided to the controller 15 to represent a signal of the detected light energy. Folding Mirror In another embodiment, the device 10 includes a folding mirror 260 to make the devices closer to each other or to form a specific device structure. In one embodiment, the folding mirror 260 is movable to adjust the direction of the light 14A ′. In addition, in one embodiment, the folding mirror 260 is connected to the controller 70 via a line 262 and is controlled by a signal 264 from the controller. Reflected radiant light monitor Continuing to refer to FIG. 3, in another embodiment, the device 10 includes a reflective wheel light monitor 280 to obtain the radiation 20 200528223 emitted light 281 reflected from the surface 62 of the substrate. The monitor 280 is connected to the controller 70 via a line 282, and provides a signal 284 representing the amount of reflected light light 281 it detects to the controller. Fig. 4 illustrates a specific embodiment of a reflected radiation light monitor 280, which is used in a device 10 whose incident angle? (Fig. 1 or 2A) is equal to or close to zero. The implementation of 5 specific examples. The reflected light beam monitor 280 uses a split beam 285 along the axis a 1 to guide a small portion of the reflected radiation light 281 (FIG. 3) to a detector 287. The monitor 280 is connected to the controller 70 via a line 282, and provides a signal 284 on behalf of the light beam it detects to the controller. In an embodiment, a focusing lens 290 for focusing the reflected radiation 281 is included in the detector 287. The reflected radiation light monitor 280 has several applications. In one mode of operation, the image 100 is made as small as possible, and a signal difference in the reflected radiation monitor signal 284 is detected. This signal is then used to determine the reflectance difference across the substrate. This mode of operation requires the inverse of a detector (such as detector 287). 15 The duration of day-to-day interval is equal to or less than the dwell time of the hidden light. By adjusting the incident angle ^, the polarization direction of the incident light 14B or both is adjusted to minimize the reflectance. In the second mode of inserting, the light energy monitoring signal 254 (No. 3®) of the light energy monitoring system 250 is combined with the stray light monitoring signal to accurately measure the amount of light absorbed and received. The energy in the radiant light 14B is then adjusted to maintain a fixed absorbed radiation. In the standard second mode, the reflected radiant light monitoring signal is compared with a critical value, and a signal exceeding the critical value is used as a warning of undesired abnormalities, requiring further investigation. In an embodiment, the difference data related to the reflected radiation is archived (eg, stored in the controller 21 200528223). After the substrate processing is completed, the corresponding substrate identification code can help determine the cause of any abnormality. . Analysis system In many heat treatments, it is helpful to know the maximum temperature or temperature_5 time curve of the surface being treated. For example, in the case of joint annealing, the maximum temperature needs to be controlled very closely at LTp. Proximity control is achieved by using the so-called measured temperature to control the output power of the continuous light source. Ideally, the response performance of such a control system is faster than, or approximately equal to, the retention time of the obscure image. 10 Therefore, referring to FIG. 3 again, in another embodiment, the device 10 includes an analysis system in communication with the substrate 60. The analysis system 300 is connected to the controller 70 via line 302 and is adapted to perform specific analysis operations, such as measuring the degree of the substrate 62. The analysis line 3GG provides a signal 304, such as an analytical measurement of the substrate temperature, to the controller. 15 Refer to FIG. 4 again when the incident angle Φ is or close to 0. The analysis system 300 is rotated out of the path of the focusing optical system 20.

FIG. 5 is a detailed enlarged view of the analysis system 300 for measuring the temperature at or near the scan image 100 in a specific example. The system 300 in FIG. 5 includes a collecting lens 340 along the axis A2 to collect the radiated light 31 and 20 to separate the collected radiated light 310 and guide the radiated light to the passing lines 302A and 302B. Connected to the two detectors 3608 and 35B of the controller 70 respectively. The detectors 350A and 350B detect 31 different bands of radiated light. A very simple analysis system 300 configuration includes a single debt test 22 200528223 =, such as-㈣ hall, 35A, align it to the hottest point on the trailing edge of stray light (Figure 3). Usually, the signal 304 obtained by this detector will change, because the image 100 on the substrate encounters different films (not shown) with different reflectances. For example, Shi Xi, Xi Shi Xi, and thin polycrystalline Shi Xi 5 films on the oxide layer all have different reflectances in the direction of normal incidence and cause different thermal emissivities. One way to deal with this problem is to estimate the temperature using only the highest signal obtained at a given time interval. This approximation improves accuracy because it reduces the response time of the detector. 10 Continuing to refer to FIG. 5, in one embodiment, the collecting lens 340 is focused on the trailing edge of the image 100 (moving in the direction of arrow 354) to collect the radiant light 31o emitted from the hottest point on the substrate 60. Therefore, the hottest (ie, most asked) temperature on the substrate can be directly monitored or controlled. Controlling the temperature of the substrate can be accomplished by several methods, including changing the energy of the continuous radiation source 12, adjusting the attenuator 226 (Fig. 3) using b, changing the substrate scanning speed or image scanning cat speed 'or any of its Combined. The temperature of the substrate 60 can be measured by monitoring the radiated light 310 emitted at a single wavelength, providing the entire surface 62 with the same thermal emissivity. If the substrate is patterned, the temperature can be measured by monitoring the wavelengths between two phases near the space during the scanning operation, assuming that the thermal emissivity does not change rapidly with the wavelength. Figure 6 is the intensity-temperature blackbody temperature curve at 1410 ° C (Figure). Its temperature is the upper limit used in certain heat treatment applications to activate the doping in the source and drain regions of a semiconductor transistor. Objects, such as regions 66A and 66B of transistor 67 (Figure 3). It can be seen from Figure 6 that the temperature 23 200528223 degrees close to i41 ° C can be monitored by the detectors 350A and 350B in the silicon detector array at 0.8 microns and 1.0 microns. Compared with a single detector, the advantage of using a detector array is that the former can obtain many temperatures along and across the image 100, so that any uneven or irregular temperature can be found quickly. In an embodiment of the active source and the dopants in the 5 drain regions 66A and 66B, the temperature needs to be raised to 1400 ° C and the point-to-point difference between the highest temperatures is less than 1 (rc. In the range of 1400 ° C, the two spectral regions may range from 500 nm to 800 nm and from 800 nm to 1 100 nm. The signal ratio of the two detectors is precisely related to temperature. It is assumed here that two different materials on the substrate surface There is no significant difference in the thermal emissivity of the 10 spectral region. The ratios of the signals 3048 and 3046 obtained by the silicon detectors 350 and 3508 are used to control the temperature, making it relatively easy to achieve a response with a time equivalent to the dwell time. Time control loop bandwidth. An alternative approach is to use detectors 15 350A and 350B in the form of a detector array, where two arrays are developed on the same substrate area but using different spectral areas. This configuration can be The temperature curve of the processing area is obtained, and the maximum temperature and temperature unevenness can be accurately determined. This configuration can also uniformly adjust the beam intensity curve. This configuration allows the use of silicon detection The control loop bandwidth has a response time approximately equal to the dwell time. 20 Another method to compensate for encountering different thermal emissivity films on the substrate is to configure the analysis system 300 to use polarized radiation at a Brewster's close to silicon Observe under the angle of angle. In this case, the wavelength related to the wavelength sensed by the Bing Gamma, s angle calculation and analysis system 300. Since the absorption coefficient in Brewster's angie is very close to a single, it is also the heat 24 200528223 Radiation 4 -In the specific embodiment, this method is combined with the method of using two arrays to obtain signal ratios at two adjacent wavelengths. In this case, the plane of the observation axis of the package A eight analysis system 300 will be perpendicular to the area containing the radiated light and The plane reflecting the radiated light 281 is shown in Fig. 7. The cat scan image can generate a large area on the substrate. However, diffraction and many possible defects in the optical chain can interfere with the formation of the image and cause problems such as Unpredictable results of uneven heating. Therefore, a built-in image monitoring system is strongly needed to directly measure the energy uniformity in the image. 10 15 20 FIG. 5 illustrates a specific implementation example of an image monitoring system 36. In an embodiment, the image monitoring system 360 is arranged on a scanning path and in a plane PS defined by a substrate surface 62. The image monitoring system includes A hole 362 facing the path of the cat, and a detector behind the hole. During operation, the substrate stage 46 is placed so that Lai Zhe 4 can be sampled when the typical image is hidden, and the points on the substrate can be sampled. The image seen is ⑽. The image surveillance system 360 is connected to the controller 70 via a line 366 and provides a signal 368 representing the detected light to the controller. The sampling of the ten image sections provides a determination of the image intensity curve (as required The data can determine the uniformity of substrate heating. Substrate Precalibrator Referring again to FIG. 3, in some examples, the substrate 60 needs to be placed on the chuck 40 in a predetermined direction. For example, the substrate 60 may be a crystalline body (e.g., a crystalline stone wafer). The inventors have found that substrates using crystals for heat treatment applications generally tend to align the crystal axis in a selected direction with respect to image 100 to optimize processing. According to this, in the embodiment, the device 10 includes a pre-calibrator 376, which is connected to the controller 70 by 378. The pre-calibrator accepts a substrate, and uses a reference feature 64 such as a plane or a notch to calibrate it to a 5 2 test position PR, and moves (such as rotates) the substrate to set the reference feature calibration direction. Processing optimization. The signal is transmitted to the controller 70 when the substrate is calibrated. The substrate is then transferred from the pre-calibrator to the carrier stage via a substrate controller 386 as a connection stage and pre-calibrator 376. The substrate controller 386 is connected to the controller 70 via a line 388, and is controlled by a signal 390. The substrate 60 is then placed on the stage 40 in a direction aligned with that on the pre-calibrator 376. Measure the absorbed radiant light Use the light energy monitoring system 250 to measure the energy of one of the radiant light i4A, 14A, or 14B, and use the monitoring system 28 to measure the energy of the reflected radiant light 15 281, which can be absorbed by the substrate 60 Radiant light. This sequence can keep the radiant light absorbed by the substrate 60 fixed, even when the reflectance of the substrate surface 62 changes during scanning. In a specific embodiment, maintaining a constant absorption energy per unit area is achieved by adjusting one or more of the following descriptions: the output power of the continuous radiation source 12; the image 100 is scanned at a speed of 20 on the substrate surface 62 ; And the degree of attenuation of the attenuator 226. In one embodiment, the fixed absorption energy per unit area is achieved by changing the polarization of the radiated light 14B, such as by rotating the quarter-wave plate 230. In another embodiment, the energy absorbed per unit area is changed or fixed by a combination of any of the techniques described above. Selected infrared 26 200528223 The absorption of the wavelength in silicon is completely increased with the improvement of silicon conductivity b impurity. Even if the incident radiation reaches its minimum absorption yield at room temperature, any increase in temperature will increase its absorbance, so an unsatisfactory cycle will occur and all the incident energy will be absorbed by a surface layer that is only a few microns deep. 5 Therefore, the heating depth in a silicon wafer is mainly determined by the diffusion of heat from the silicon surface, not the absorption depth of infrared wavelengths at room temperature. Similarly, doping silicon with η-type or p-type impurities increases the room-temperature absorption and further promotes this runaway cycle, resulting in a strong absorption of the material in the first few microns. The incident angle 10 at or near Brewster's angle In an embodiment, the incident angle Φ is set to be consistent with Brewster * and angle. At the Brewster's angle, all p-polarized radiation light p (Fig. 3) is absorbed in the substrate 60. Brewster's angle is based on the refractive index of the material on the incident radiation. For example, Brewster's angle of Shi Xi at room temperature is 73.69. And the wavelength is λ = 10.6 micrometers. Since about 30% of the incident radiated light 14B is reflected in the direction of normal incidence 15 (φ = 0), the use of ρ-polarized light at or near Brewster ’s angle can significantly reduce the power required for heat treatment per unit area. The use of a relatively large incident angle φ such as Brewster's angle also makes the image 100 wider in one direction by cos- ^, or about 3.5 times the width of the vertically incident image. The effective depth of focus of the image 100 is also reduced by similar elements. 20 When the substrate 60 has a plurality of surfaces 62 in different regions, as in the case of ICs formed by typical semiconductor processing, the angle most suitable for processing can be measured by plotting the reflectivity versus the incident angle Φ in different regions. It is generally found that the P-polarized radiation light has the smallest reflectance in the region of each substrate close to the Brewster's angle. It is usually found that an angle, or a small range of angles, minimizes and equalizes the reflectivity in each area. In the embodiment, the 'incident angle Φ is limited to the angle range around the Brewster's angle. The Brewster's angle is 73.69. In the embodiment, the incident angle φ can be limited to 65. And 80. between. 5 Optimal radiant light structure In one embodiment, the scanning image 100 is used to thermally process the substrate 60 on the surface 62, causing a small volume of material on the substrate surface to be heated to a temperature close to the melting point of the substrate. Accordingly, a large amount of stress and strain are generated in a region where the substrate is heated. In some cases, this stress results in an undesired slip plane extending to the surface 62. Similarly, in one embodiment, the radiated light 14α is polarized. In this case, it is practical to choose the polarization direction of the incident radiation light 14B relative to the substrate surface 62, as the direction of the incident radiation light 14B incident surface 62 results in the most efficient processing. In addition, the heat treatment of the substrate 60 is usually performed after the substrate has undergone more than 15 other processes that may change the properties of the substrate such as the structure and shape. Fig. 7 is a detailed enlarged isometric view of an embodiment of a substrate 60 in the form of a semiconductor wafer having a pattern 400 formed thereon. In one embodiment, the pattern 400 includes lines or edges 404 and 406 to form a grid (ie, a Manhattan geometry) with lines / edges running in the 1 and y_ directions. Lines / edges 404 and 406 correspond to the edges of more than 20 runners, gate and field oxide isolation areas, or 1C wafer boundaries. Generally, in 1C wafer manufacturing, substrates are most often continuously patterned and the patterns are perpendicular to each other. Therefore, when the substrate (wafer) 60 is annealed during the process of forming the IC or the source and drain regions 66A and 66B need to be activated, the surface 62 is quite complex. For example, in a typical 1C process, one area of the surface 62 may be pure silicon, while another area of the surface may have a relatively thick oxide oxide isolation trench, and other areas of the surface may have thin polycrystalline silicon. The Shi Xi conductor traverses this thick oxidation trench. 5 According to this, if you do not pay attention, the image 100 will be reflected or diffracted by a part of the substrate surface 62, and will be selectively absorbed in other areas according to the surface structure, including the main directions of the lines / edges 404 and 406. This is particularly true in the embodiment where the radiated light 14b is polarized light. The result is non-uniform substrate heating, which is usually undesirable during heat treatment. 10 Therefore, with continued reference to FIG. 7, in one embodiment of the present invention, it is desirable to find a ^ best round light structure, i.e., polarization direction, incident angle φ, scanning speed, and image angle Θ for light light. The difference in the absorbance of 14B on the substrate 60 is the smallest. In addition, it is more desirable to find the optimal radiant light structure to minimize the slip plane generated in the substrate. 15 The point-to-point difference in the radiated light 281 reflected by the substrate 60 is caused by several factors, including differences in film composition, the number and ratio of lines / edges 404 and 406, the orientation of the polarization direction, and the incident angle φ. With continued reference to Figure 7, a plane 440 is defined so that it contains radiated light 14B and reflected radiated light 281. The radiation light 14B can be used to irradiate the substrate to minimize the difference in reflectivity due to the presence of the lines / 20 edges 404 and 406, so the plane 440 intersects on the substrate surface 62 and the lines / edges 404 and 406 become 45. . The line image is formed so that its long direction is also calibrated to the same plane 440 or perpendicular to this plane. Therefore, regardless of the incident angle φ, the image angle Θ between the line image 100 and the opposite lines / edges 404 and 406 is 45 °. 29 200528223 The difference in the amount of reflected radiant light 281 caused by different structures (such as lines / edges 404 and 406) on the substrate surface 62 can be further reduced by clearly selecting the incident angle ^. For example, in the case of forming a transistor as part of 1C, when the substrate 60 is ready to anneal or activate the source and drain regions 66A and 5 66B, it typically contains all the following types: a) pure silicon, b ) Oxidation insulators buried in silicon (such as about 0.5 microns thick), and thin (such as 0.1 micron) polycrystalline silicon runners buried on top of buried oxide insulators. Figure 8 shows ρ-polarized radiation? And ^ polarized radiant light s, 10.6 micron long laser light light 'at the above are the group reflectivity of the doped silicon substrate at various room temperature reflectances along the reflectivity of the silicon oxide layer of infinite depth. Figure 8 clearly shows that the reflectance varies greatly with polarization and incident angle φ. The angle of incidence φ is between about 65. To about 80 ip_polarized radiated light ρ (i.e., polarization at plane 440), the reflectance of all four cases is the smallest, and the difference from each case is also the smallest. So by about 65. To about 80. The range of the incident angle ψ is particularly suitable for the device 10 for heat treating a semiconductor substrate (such as activation in a doped region of a silicon substrate), because they all reduce the total power required and the point-to-point difference caused by absorbed radiation. The presence of a dopant or higher temperature makes silicon more metal-like, and moves the minimum value relative to Brewster's angle to a higher angle and higher reflectivity. Therefore, for doped substrates and / or higher temperatures, the optimum angle is higher than Brewster, which is doped material at room temperature, as §16. FIG. 9 is a top isometric view of a device 10 for processing a substrate 60 in the form of a semiconductor wafer, showing operation of the device in an optimal radiated light structure. The wafer 60 includes a grid pattern 400 formed thereon, and each square 30 in the grid 30 200528223 468 represents a 1C wafer (such as the circuit 67 in FIG. 1A). The line image ιΟ〇 is concealed from the substrate (wafer) surface 62 to 470 to make the image angle a shirt. . Calculating the crystal orientation As described above, a crystal substrate such as a single crystal silicon wafer has a crystal surface 5 whose orientation is usually formed by the McCaw feature 64 (as shown in a (Notched (plane) in Figure 9). The scanning of the line image 100 has a large thermal gradient and stress concentration in the direction 474 of the vertical scanning direction 47 ° (Figure 9), which has an adverse effect on the structural integrity of the crystal substrate ® plate. 10 Following the performance of the 9th figure, Shi Xi substrate 60 usually has a (100) crystal orientation, and the lines / edges 404 and 406 are aligned. The two main crystal axes (100) and ( 010) into 45. . A better scanning direction is along the main crystal axis to reduce the slip plane formed in the crystal. Therefore, the better-known direction used to reduce the formation of the moon shift in the crystal is also consistent with the better direction of the broken substrate in the normal case for the line / edge 404 and 406. If the line image 100, I line / edge 404 and 406, and the crystal axes (100) and (010) are kept in a fixed direction, the line image scan relative to the substrate (wafer) 60 must be linear Way (ie backward or forward) rather than round or 圆形. Similarly, since the preferred specific sweeping direction needs to be related to the crystal direction, in one embodiment, the substrate is pre-calibrated on the 20 chuck 40 using a substrate pre-calibrator 376 (Fig. 3). By carefully selecting the orientation between the substrate crystal axes (100) and (010) and the scanning direction 470, the possibility of a slip plane due to thermally induced stress in the crystal substrate can be reduced. The most suitable scanning direction is due to the thermal gradient, which causes the crystal lattice to have the largest sliding resistance. It is generally believed that it depends on the nature of the substrate crystal 31 200528223. However, the direction can be determined by scanning the image 100 in a spiral pattern of a single crystal substrate and inspecting the wafer, and the most suitable temperature gradient of the fork can be experimentally found before the slip occurs. In the substrate 60 in the form of a (100) crystalline silicon wafer, the scanning direction 5 is most suitable to be aligned to the (100) substrate crystal direction or 45 with the pattern grid direction indicated by the lines / edges 404 and 406. . This has been demonstrated by the inventor using a radial pattern scan of a field pattern 100 in a spiral pattern. The increasing maximum temperature and the distance between the substrate's ten centers are a function relationship. Using the comparison and the direction of the crystal axis with the direction of large anti-slip can get the most suitable scanning direction. 10 Image Scanning Folding Scan_ Figure 10 is a plan view of a substrate illustrating a pattern 52 of scanning the image 100 by folding (ie changing backward and forward or “X-γ”) on the substrate surface 62, To generate short thermal pulses at each point on the substrate of the image overrun. The scanning pattern 52 includes 15 linear scanning scanning segments 522. The folding scan pattern 520 can be completed by a conventional bidirectional X-Y stage. However, this stage is typically very heavy and limits its acceleration capability. If a very short dwell time is required (that is, the time when the scan image stops at a specific point on the substrate), the conventional stage will waste a lot of acceleration and deceleration time. This stage also takes up a lot of space. For example, a 100- 20 micron light width retention time of 10 microseconds requires a stage speed of 10 meters per second (m / s). Acceleration at a lg or 9.8m / s2 requires L02 seconds and a movement of 5.1 meters to accelerate / decelerate. It is not necessary to provide 10.2 meters of space for the platform to accelerate and decelerate the system. Optical scanning 32 200528223 On the surface of the substrate 62, the image of 100 cats can be used to move the image, and the substrate can be moved and held / soiled, and the image can be used. ~ The cross-sectional view of the specific example of the implementation of the second optical system 20 with 'or both moving the substrate is as follows: i # / J # 动 —26 G can pass through optical materials to achieve a very efficient acceleration rate (ie load The speed at which the station needs to move to achieve the same sweeping results).

10 In the optical system 20 in Fig. U, the light 14A (or 14A,) is reflected by the thin mirror of the aperture of the flat field continuum optical system formed by the cylinder lens L pulse. In the implementation of various assets, once the mirror is connected to and driven by the line 542 to the horse's horsehead to connect and drive the 'servo motor assembly 540 is controlled by the signal 544 transmitted by the control line 542. The optical system 20 glances at the substrate surface 62 and shoots the "MB to form a moving image 15 of the moving line 100". The stage 46 increases the position of the substrate in the bidirectional concealment direction after each scanning seedling to cover the scanning area required for the substrate. In a specific embodiment, the lens elements L10 to U3 are made of chemistry and can penetrate the radiant light emitted by the C02 laser and the near-IR and visible radiant light emitted by the heated portion of the substrate. Infrared wavelength. This allows a two-color beam splitter 55 to be placed on the front of the path of the radiated light 14A of the scanning lens 20 260 to separate the visible and near IR wavelengths of the radiated light from the substrate caused by the long-wavelength radiated light 14A used to heat the substrate . The emitted radiant light 310 is used to monitor and control the heat treatment of the substrate, and is provided with a collection lens 562 and a detector 564, and is connected to the light analysis system 560 of the controller 7033200528223 via a line 568. In one embodiment, the radiated light 3i0 is filtered and focused to a separate detection array 5 64 (only one is shown). The signal 570 of the amount of the radiated light is measured by the sensor 56 and supplied to the controller 70 via the line sensor. 5 Although Fig. 11 shows that the radiated light 14B has an incident angle 9 = 0, in another embodiment, the incident angle is Φ > 0. In an embodiment, the incident angle φ is changed by appropriately rotating the substrate stage 46 along the axis AR. One of the advantages of optical scanning is that it can be performed at very high speeds, so there is minimal time wasted in accelerating and decelerating the beam or stage. In the commercially available sweeping cat optical 10 system, it can achieve the same effect as the acceleration of 8000g stage. Spiral scan cat In another embodiment, the image 100 is scanned under a spiral pattern relative to the substrate. FIG. 12 is a plan view of four substrates 60 on a stage 46, wherein the stage has the ability to rotate and linearly move with respect to the image 100 to generate a spiral scanning pattern 604. This rotation action is performed by rotating the center 61m to rotate the IIm substrates' to show that four substrates are used for explanation. In another embodiment, the stage 46 includes a linear stage 612 and a rotary stage 614. The duty scan pattern 604 is formed by combining the linear and 20 rotation motions of the substrate, so each substrate is covered by the spiral scan pattern. In order to make each point stay on the substrate for a fixed time, the rotation rate is inversely proportional to the distance between the image 100 and the rotation center 610. In addition to the initial and termination treatments, the spiral sweep cat has no rapid acceleration / cutting secret. Based on this, you can use this configuration to actually get a short dwell time. Another advantage is that several substrates 34 200528223 can be processed in a single scan operation. The alternating raster scan of the folding pattern scan image 100 on the substrate 60 at a small adjacent path pitch will cause the end of the substrate scanning segment to overheat, and the end of the scanning segment is a 5 slice and just 70% When another segment is starting. In this case, the beginning part of the new scan path segment has a thermal gradient of selection due to the scan path segment just completed. This gradient increases the temperature due to the new scan, unless the beam intensity is properly adjusted. This makes it difficult to reach a uniform maximum temperature across the entire substrate during scanning. 10 Figures 13A and 13B are plan views of a substrate 60 illustrating an alternate raster scan path 700 with linear scan path segments 702 and 704. Referring first to FIG. 13A, in the alternate raster scanning path 700, the scanning path segment 702 is performed first, so there is an interval 706 between adjacent scanning paths. In one embodiment, the interval 706 has a size equal to 15 times the positive multiple of the effective length of the line scan. In one embodiment, the width of the interval 706 is the same as or close to the length L1 of the image 100. Next, referring to FIG. 13B, the path segment 704 is scanned to fill the gaps. This scanning method greatly reduces the thermal gradient of the scanning path caused by continuous scanning path segments at close range, making it easier to reach the highest uniform temperature across the entire substrate. 20 Comparison of the output of scanning patterns Figure 14 shows the spiral scanning method (curve 72), the optical scanning method (curve 724) and the folding (χ-Υ) scanning method (curve 726) in a simulated output (substrate / Small day possible) vs. retention time (seconds). This comparison is based on the assumption that a 5 kW laser is used as a continuous radiation source in an embodiment, which is used to generate a Gaussian image with a beam width L2 of 35 200528223 and a width of 100 microns, and is scanned by an overlapping scanning path to achieve Uniformity of radiant light of about ± 2%. From the figure, it can be seen that the spiral scanning method has better yield under all conditions. However, this spiral scanning method processes several substrates at a time, and therefore requires a large area to support the 4-chuck. For example, for four 300mm crystal circles' the surface diameter needs to be greater than about 800mm. Although the disadvantage of this method is that it cannot maintain the crystal orientation between the line scan image and the substrate, it cannot maintain the optimal processing structure for a crystal substrate. The yield of the optical scanning method is almost independent of the dwell time, and it has an advantage over the χ_γ 10 stage scanning seedling system when a short dwell time requires high scanning speed. Cyclic Optical System In the present invention, it is important to transfer energy from the continuous radiation source 12 to the substrate 60 as much as possible. Accordingly, referring briefly to FIG. 19, which will be described in detail below, in one embodiment, the radiated light 14B has a large incident angle range on the substrate. That is, the optical system 20 has several apertures NA = sine14b, where e〗 4b is a half angle formed by the outer beam 15A or 15B of the axis A1 and the radiated light 14B. Note that the incident angle Φμβ formed by the axial beam (axis A1) and the substrate surface normal N is referred to herein as the "center angle," which is the range of angles provided by the radiated light 14B. 20 In an embodiment, The central angle e] 4b can be selected to reduce the difference in reflectivity between different film stacks (not shown) on the substrate. In the past, the portion of the radiant light 14 reflected by the substrate surface 62 is not easy to avoid. Therefore, One embodiment of the present invention includes capturing the reflected radiant light 23R and guiding it back to the substrate as a "circulated radiant light" 23RD, which can be absorbed by the substrate located at the place where the incident radiant light 14B is reflected. 23RD also provides additional thermal energy to one or more substrate areas (such as areas 66A, 66B in FIG. 1) to assist the annealing process. Accordingly, referring to FIG. 15, a laser scanning device according to the present invention is shown. The detailed enlarged schematic diagram of the specific example of the installation of 5 sets of support. The device 10 in Fig. 15 is similar to that in Fig. 1A, but it also includes a configuration to obtain the reflected radiation 23R and guide it back to the substrate. As cyclic spokes The circular optical system 900 of the light 23rd. The circular optical system 900 is arranged along the axis AR and forms an angle with the surface normal N (p23RD. In order for the circular optical system 900 to obtain the optimal reflection 10 of the radiant light 23R, one implementation In the specific example, the angle CP23RD is the same as the incident angle φΐ4B of the radiant light. It should be noted that in the present invention, the substrate is illuminated by a radiation pulse. As described above, the radiation "pulse" is scanned by the radiant light 14B to make the substrate The selected part is exposed to the light 15 light 14B at a specific time, that is, the light retention time. Strictly speaking, in the specific example, there is a setting 10 of a circulating optical system, and the reflected radiation light 23R actually constitutes Incident radiation light MB is accompanied by a weaker second pulse. This second pulse is delayed by △ kOPL / c in time compared to the first pulse, where 0PL is the reflected radiation track before it returns to the substrate. In the cyclic optical system 900 The length of the optical path traveled in the middle is 20 velocities of light (~ 3xl08 m / s). Since the OPL is -m or less, the delay time between pulses AT is 10 · 9 seconds. When the scanning speed is for丨 meters / second (m / s), the space between the first and second pulses on the substrate surface 62 is separated into ~ (ι m / S) (l (T9S) ~ lG.9m, which is in the case of laser annealing The space is not obvious in the middle system 37 200528223. Therefore, the incident and reflected Korean beams effectively overlap, that is, they can reach the same part of the substrate at the same time for all practical purposes. Therefore, the combination of incident and reflected pulses results in a single light beam with enhanced energy In other words, for all purposes and applications, the incident (first) radiant light 14B and the cyclic (second) 5 radiant light 23RD irradiate the substrate (eg, on one or more areas above it) simultaneously. FIG. 16 is a cross-sectional view of a specific embodiment of a circular optical system 900, which includes a hollow right-angle mirror 910 and a collection / focusing lens having a focal length F and the same distance from the lens along the axis AR to the substrate surface 62 Lens 916. Hollow right-angle reflection 910 has three reflective surfaces that intersect at right angles. Although it is a simplified illustration, only two surfaces 912 and 914 are shown in FIG. 16. In the 900 operation of the cyclic optical system of FIG. 16, the lens 916 collects the reflection and collimates the parallel light 920. The parallel light is reflected by the three reflecting surfaces and is guided back to the lens 916 in completely opposite directions, and on the other side of the axis ar, as the parallel light 920, which constitutes the cyclic radiation light 23R. The parallel light 920 is collected by the lens 916 15 and refocused on the original point 321 on the substrate surface 62. FIG. 17 is a cross-sectional view of a modification of the specific example described in FIG. 16 where the right-angle mirror 910 is displaced relative to the axis AR (the amount of deviation. This causes the reflected radiation light 23R and the cyclic radiation light 23RD to enter the substrate There is a deviation of the angle. Note that the position of the beam on the substrate is still the same-only the angle of incidence 20 changes. The relative deviation between the angles of incidence of the two beams can be used to prevent the reflected radiation from traveling back to the continuous radiation source 12 (Figure 15) In this specific embodiment, all internal reflections are used at right angles because it cannot maintain the polarization of the light beam instead of what is desired. Figure 18 is a cross-section of a circular optical system 900 in another embodiment 38 200528223 Figure. The substrate 60 includes a cylindrical lens 950, a first cylindrical lens 352, an aperture 954, a second cylindrical lens 956, and a polarized light-maintaining ridge mirror 960 in order along the axis AR. In the example, the first and second cylindrical lenses 352 and 956 have the same focal length (F ') and are separated to twice the focal length of 5 focal lengths (2F') and form an aperture 954 halfway through them. Roof mirror 960 It is located at a distance from the cylindrical lens 956F 'and the roof mirror 960 faces the direction of the p-polarized radiation reflection. In the specific embodiment of the circular optical system 900 shown in FIG. 18, it is assumed that the radiated light 14B is focused by the optical system 200 and is on the substrate An image 100 is formed on the image (Fig. 15 10). The cylindrical mirror 950 obtains and collimates the reflected radiant light 23R, which then penetrates the cylindrical lenses 952 and 956. The roof mirror 960 is configured to change the direction of the radiated light and pass back. The cylindrical lens reaches the cylindrical mirror and returns to the surface of the substrate. The tilting of the ridge mirror 960 with respect to the incident radiation 23 determines the angle of the preheated radiation 23RD incident on the substrate 60 in a different direction to the substrate 60. In one embodiment, polarization 15 The light-retaining roof mirror 960 includes a micro-tilt design to prevent circulating radiation 23RD from returning to the continuous radiation source 12. The radiation returning to the laser or laser diode cavity can cause operational problems such as laser output power FIG. 19 is a cross-sectional view of another embodiment of the cyclic optical system 900, which includes a collimating / focusing lens 1050 and a 20 grating 1060 having a grating surface 1062. In a specific example, the lens 1050 is a high-resolution, telecentric replacement lens having first and second lenses 1070 and 1072, and an aperture light block 1074 located between the first and second lenses. In addition, in this embodiment, The lens 1050 has a focal length F1 on the substrate side and a focal length F2 'on the grating side, and the lens is configured so that the substrate surface 62 is located away from the lens 1 〇〇〇〇〇〇〇〇３０２５２５２ measuring the distance F1 along the axis AR, and the grating 1060 is located at a distance F2 from the lens 1072 along the axis ar. The two lenses 1070 and 1072 are also separated to the same distance as the sum of their two focal lengths. The thumb surface 1062 is preferably adapted to optimize the wavelength of the radiated light in the diffracted reflected radiant light 23r, and to limit the radiated light incident on the grating surface to be diffracted to return along the incident path. The best grating period? Is p = nAy2sin (pG where λ is the wavelength of the radiant light, (pG is the angle of the grating relative to the normal NG of the grating surface 'and η is the refractive index of the medium around the grating (n = 1 for air). Grating The purpose of is to compensate the tilted focus plane on the substrate. The other 10 faces will cause the returned image to be out of focus according to the amount of the axial plane distance between point 321 and the replacement lens 1050 in Figure 19. Note that in this structure The replacement lens 1050 is operated at -IX, (pG = 9) 4B = 923R = 923RD. Generally tan (pG = Mtan (p23R, where M is the magnification from the substrate to the grating replacement lens 1050). 15 Operation The reflected radiant light 23R is collimated by using a telecentric replacement lens 1050, which includes a lens 1070 and a lens 1072, which bring the radiated light to a focus on the grating surface 1062. The grating surface 1062 changes direction (or more precisely, The radiant light returns to the replacement lens 5050, which guides the current cyclic radiant light 23RD back to the substrate surface 62 at or near the point 321, which is where the reflected radiant light 20 is generated. The specific example of Fig. 19 is disadvantageous. 23R is formed on the grating for reflecting the radiated light. After a certain time, the image may cause the grating to eventually melt or be damaged. A similar problem encountered is the use of a vertical incidence mirror (not shown) instead of the grating. Therefore, the implementation of the circular optical system 900 in FIG. 200528223 example, care must be taken when operating the device. Figure 20 is a schematic cross-sectional view of a specific example of a laser scanning device used to anneal the substrate 60. These devices use two-dimensional laser diode arrays Radiation sources 12 and 12 ', and two optical systems 20 and 20' arranged along the axes A1 and AΓ, respectively. The continuous radiation light sources 12 and 12 'are effectively connected to the controller 70 and emit radiation light 14 A and 14 respectively. A '. Each radiant light is received by the opposite optical systems 20 and 20'. The optical systems 20 and 20 are formed on the substrate surface 62 in sequence by the radiant light 14A and 14A 'generated by the radiant light 14B and 14B'. Images 100 and 100 '. 10 In one embodiment, the optical systems 20 and 20 at least partially overlap each other to form an image 100 on the substrate. In another embodiment, the images 100 and 100' are line images. In another embodiment, at least one of the annealing light 14B and 14B ′ is incident on the substrate surface 62 at an incidence angle of φΜΒ and φΜΒ, which is at or near Brewster's angle cpB of Shi Xi. 15 This configuration reduces the continuous radiation source. 12 and 12, requirements for outputting high-power radiant light 14B and 14B '. The specific embodiment of the device in Figure 20 is not limited to two radiant lights 14 and 14B. Generally, any reasonable continuous radiation source 12, 12', 12 "'and the relative number of optical systems 20, 20, 2, 20, etc., can be used to form relative images 100, 100, 100 on the substrate surface 62, 20 (such as line image) to achieve the desired annealing effect. Many features and advantages of the present invention have been specifically described by the detailed description, and therefore, 'is meant to cover all features and advantages of the device following the description of the true spirit and scope of the present invention by the additional application scope. In addition, since those skilled in the art will quickly think of several improvements and changes, there is no need to limit the actual structure and operation of the present invention. Accordingly, other specific examples are within the scope of the attached patent application. [Schematic description] Figure 1A is a simplified diagram of a general specific example of the device of the present invention; 5 Figure 1B illustrates the implementation of an ideal line image with a long size L1 and a short size L2 on the substrate by the device of Figure 1A Specific example; Figure 1C is a two-dimensional figure representing the intensity distribution accompanying the actual line image. Figure 1D is a simplified diagram of a specific example of an optical system implementation of the device of Figure 1A, which includes a cone lens to form a line image on a substrate; Figure 2A is a simplified diagram illustrating the implementation of the laser scanning device of Figure 1A The specific example further includes a light converter disposed between the radiation source and the optical system; FIG. 2B is a schematic diagram illustrating how the light converter in FIG. 2A changes the beam intensity curve of the radiated light; FIG. 2C A cross-sectional view of a specific embodiment of a converter / optical system including a flat Gaussian beam intensity curve converter; FIG. 2D is a non-peripheral dark-angled uniform radiant light formed by the converter / optical system of FIG. 2C A diagram of an exemplary beam intensity curve; FIG. 2E is similar to FIG. 2D, and the edge beam 20 is removed by a peripheral vignette to reduce the intensity peak of the image; FIG. 3 is a simplified diagram of a device similar to FIG. , Its additional components represent different embodiments of the present invention; FIG. 4 illustrates the embodiment of the reflective radiation monitor of FIG. 3 whose incident angle Φ is equal to or close to 0 °; 42 200528223 FIG. 5 is a diagram of FIG. 3 Used to measure the scan image on the substrate A detailed enlarged view of a specific example of the analysis system 300 of the temperature at or near the 100 position. Figure 6 is the intensity-temperature blackbody temperature curve at 14HTC (Figure). Its temperature is slightly higher than the temperature of the dopants in the source and drain regions used to activate a semiconductor transistor. Figure 7 This is a detailed enlarged isometric view of a substrate with grating features that shows incident and reflected laser light in a 45-degree plane with respect to the features of the grating pattern. Figure 8 plots laser radiation with a wavelength of 106 micrometers from the following surface. 10 P & s polarization direction reflectance versus incidence angle (a) Pure Shixi, (b) 0.5 micron oxide insulation layer on top of Shixi, and 0.5 micron oxide top on silicon. 1 micron polycrystalline silicon flow channel, and (d) an infinitely deep silicon oxide layer; FIG. 9 is a plan view of a specific example of the device of the present invention for processing a gate pattern semiconductor wafer-form substrate 60 formed thereon, illustrating the substrate Operate in the best 15 radiant light structures; Figure 10 is a plan view of a substrate illustrating a pattern of a scanning image folded on the surface of the substrate; Figure 11 is a cross-sectional view of an embodiment of an optical system, including a movable Scanning Mirror 20 Figure 12 is a plan view of the four substrates with the ability to rotate and linearly move the image on the stage, creating a spiral scanning pattern on the substrate; Figures 13A and 13B are plan views of the substrate illustrating an alternating light scan Miao pattern, where the scanning path is separated by a space that allows the substrate to cool before scanning adjacent Miao Miao road; 43 200528223 FIG. 14 is a spiral scanning method, optical scanning method and The fold scan method is based on the simulation output of the substrate / hour versus the dwell time in microseconds. Figure 15 is a detailed diagram of a specific embodiment of the LTP system similar to Figure 1A. 5 A simplified diagram, which also includes configuration To obtain the reflected radiant light and guide it back to the substrate as a cyclic optical system of cyclic radiant light; FIG. 16 is a cross-sectional view of a specific embodiment of the cyclic optical system of FIG. 15, which includes a right-angle mirror and a collecting / focusing lens Figure 17 is a cross-sectional view of Modification 10 of the embodiment of the circular optical system shown in Figure 16, in which the right-angle mirror is displaced (deviated) from the axis AR by the amount of AD, resulting in direct incident and cyclic The incident light has a deviation at the angle of incidence; FIG. 18 is a simplified cross-sectional view of a specific embodiment of the cycle optical system in FIG. 15, which includes an enlarged relay group and a roof mirror; FIG. 19 is another view of the cycle optical system in FIG. 15. A 15-section schematic diagram of the specific embodiment, which includes a collimating / focusing lens and a grating; and FIG. 20 is a schematic diagram of a specific embodiment of an LTP system, which uses two laser diode arrays and two opposite LTPs. The optical system irradiates the substrate at a similar angle of incidence across the substrate normal. [Description of main component symbols] 10 ... device 40 ... chuck 12 ... continuous radiation source 42 ... upper surface 14A ... radiated light 46 ... stage 14B ... radiated light 50 ... press plate 20 ... optical system 60 ... substrate 44 200528223 62 ... substrate surface 200 ... flat portion 63 ... cross section 204 ... beam end point 64 ... reference characteristic 210 ... intensity peak 66A ... source region 226 ... attenuator 66B ... drain region 227 ... polarizer 67 ... circuit 228 ... line 70 ... controller 229 ... signal 72 ... line 230 ... quarter wave plate 76 ... stage controller 250 ... light energy monitoring system 78 ... line 252 ... line 80 ... line 254 ... light energy monitoring signal 82 ... line 260 ... folder 90 ... signal 262 ... line 92 ... signal 264 ... signal 94 ... signal 280 ... reflected radiation monitor 100 ... image 281 ... reflected radiation 104 ... collimator lens 282 ... line 106 ... · Wheel 284 ... Reflected radiation monitoring signal 108 ... Wheel 285 ... Splitter 110 ... Wheel 287 ... Detector 150 ... Light converter 290 ... Focus lens 160 ... · Converter / optical system 300 ... Analysis system 170 ... beam 302 ... line 18 0 ... Peripheral Vignette Iris 302A ... Line 45 200528223 302B ... Line 406 ... Line / Side 304 ... Signal 468 ... Square 310 ... Radiant Light 470 ... Scanning Direction 304A ... Signal 474 ... Direction 304B ... ·· Signal 520 ... Linear scanning pattern 340 ... Collecting lens 522 ... Linear scanning scanning segment 346 ... · Splitter 540 ... Servo motor assembly 350A ... Silicon detector 542 ... Line 350B ... Silicon detector 544 · ·· Signal 352 ··· The first cylindrical lens 550 ... Dual color beam splitter 354 ... Arrow 560 ... Light analysis system 360 ... Image monitoring system 562 ... Collecting lens 362 ... Hole 564 ... Sensor 364 ... Ballast detector 568 ... line 366 ... line 570 ... signal 368 ... signal 604 ... spiral scanning pattern 376 ... pre-calibrator 610 ... rotation center 378 ... line 612 ... linear stage 380 ... signal 614 ... rotary stage 386 ... substrate controller 700 ... alternate light path 388 ... line 702 ... linear scan path segment 390 ... signal 704 ... linear scan path segment 400 ... pattern 706 ... interval 404 ... line / edge 720 ... curve 46 200528223

724 ... curve 726 ... curve 900 ... circulating optical system 910 ... hollow right-angle mirror 912 ... surface 914 ... surface 916 ... lens 920 ... reflected radiation parallel light 920 '... reflected radiation parallel light 950 ... cylindrical mirror 954 ··························································································· 10 23RD ... Cyclic radiation

47

Claims (1)

200528223 X. Scope of patent application: 1. A device for heat treating a region of a substrate, comprising: a continuous radiation source capable of providing a continuous first radiant light and a wavelength capable of heating the substrate region; 5 a suitable for obtaining the first An optical system that radiates light and thereby forms a second radiant light to form an image on the substrate; a cyclic optical system configured to obtain radiant light reflected from the substrate, and directing the reflected radiant light back to the substrate as a cyclic reflection Light; and a stage adapted to support the substrate, and scan the 10 substrates relative to the image to utilize the first pulse of radiation from the optical system and the second pulse of radiation from the circulating optical system to heat the area sufficiently The temperature of the area is heat treated. 2. For the device in the scope of patent application, the image is a line image. 3. The device according to item 1 of the patent application scope, wherein the circulating optical system 15 includes a collimating / focusing lens and a corner reflector. 4. The device according to item 3 of the patent application, wherein the circulating radiation light and the second radiation light each have an incident angle, the circulating optical system has an optical axis, and wherein the right-angle mirror is offset relative to the optical axis Therefore, the circulation and the incident angle of the second radiated light are separated at least partially. 20 5. The device according to item 1 of the patent application scope, wherein the circulating optical system includes a telecentric continuity device and a diffraction grating. 6. The device according to item 1 of the scope of patent application, wherein the circular optical system in sequence along the optical axis from the substrate comprises: a cylindrical mirror, 48 200528223 a repeater magnification group (ιχ); and a device adapted to reflect the The reflected radiant light is returned to the substrate through the repeater amplification group to maintain the ridge mirror. 7. The device according to item 6 of the scope of patent application, wherein the repeater magnification group 5 includes: first and second cylindrical lenses having the same focal length and separated by twice the focal length; and Midway apertures of first and second cylindrical lenses. 8. The device according to item 1 of the patent application scope, wherein the circulating optical system 10 is adapted to guide the circulating radiant light to the substrate at an incidence angle at or near the Brewster's angle. 9. An apparatus for heat-treating a region of a substrate, comprising: two or more continuous light sources each capable of providing a continuous first radiant light and capable of heating the wavelength of the region of the substrate; 15 two or more each adapted to obtain The first radiated light and thereby the second radiated light form an optical system forming an image on the substrate, thereby respectively forming two or more images on the substrate; and an optical system adapted to support the substrate and opposite to the substrate. Two or more images scan the substrate to cause each two or more radiation pulses to heat the area to a stage that is hot enough to process the temperature of the area. 10. The device of claim 9 in which the scope of the patent application is applied, wherein the two or more optical systems are suitable for forming the two or more images as line images. 11. A method for heat treating one or more regions of a substrate, the steps comprising: 49 200528223 a. Generating a continuous beam of light having a wavelength capable of heating the one or more regions; b. Radiating light from the continuous beam Illuminating the substrate as the first radiant light; c. Capturing reflected radiation from one or more regions of the substrate and directing the reflected radiation back to the one or more regions as circulating radiation; and d. In The first radiant light and the cyclic radiant light are scanned on the one or more areas, so that the one or more areas obtain a thermal energy capable of processing the one or more areas. 10 12. The method according to item 11 of the patent application range, wherein the cyclic radiation light is formed so that it has an incident angle corresponding to the minimum reflectance of the substrate at the selected wavelength. 13. The method of claim 11, wherein directing the reflected radiant light back to the one or more regions includes reflecting 15 radiated light obtained by the right-angle mirror. 14. The method of claim 11, wherein directing the reflected radiation back to the one or more areas includes reflected radiation reflected by a roof mirror and a cylindrical mirror. 15. The method according to item 11 of the patent application, wherein directing the reflected radiant light 20 back to the one or more regions includes reflected radiant light diffracted by a diffraction grating inclined with respect to the reflected radiant light, so that The reflected radiant light directed back to the substrate remains focused in the one or more regions. 16. The method of claim 11 in which the reflected radiation is guided back to the one or more regions includes: 50 200528223 guiding the reflected radiation through a cylindrical mirror and a replacement amplifier (IX ) To a polarized light-maintaining ridge mirror, wherein the ridge mirror is adapted to reflect the reflected light rays back to the repeater magnification group to form a focused image on the substrate portion. 5 17. A method for heat-treating a region of a substrate, comprising: generating two or more continuous light beams having a wavelength capable of heating the region of the substrate, and two or more each adapted to obtain the first radiant light An optical system that forms an image on the substrate from the second radiant light thereby obtains the 102 or more consecutive first radiant lights, wherein each of the second radiant light forms an image on the substrate, thereby Two or more at least partially overlapping images are formed on the substrate, respectively; and the substrate is scanned relative to the two or more images to cause each two or more co-existing carousel pulses to heat the region to a temperature sufficient to heat treat the region 15 temperature. 51