Fourier notch filter method and system

Abstract

The invention discloses a technology for determining the temperature of a semiconductor film in the growth process of a multi-quantum well (MQW) film through a metal organic chemical vapor deposition (MOCVD) method. The temperature is determined in real time as the film grows and increases in thickness. On each incremental thickness, a spectrum based on diffusely scattered light from the thin film is generated. Each spectrum is reference divided to correct for device artifacts. And determining the film thickness and the optical absorption edge wavelength value according to the spectrum. The film temperature is determined as a function of the optical absorption edge wavelength and the film thickness by using the spectrum, the thickness calibration table, and the temperature calibration table. The determination of the film temperature is accurate to + / -0.25 DEG C by using a fast Fourier transform (FFT) and a notch filter.

Description

CROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims priority to U.S. Provisional Patent Application Serial No. 63 / 507,823, filed on June 13, 2023, the entire disclosure of which is incorporated by reference into this application and relied upon. Technical Field

[0002] The present invention relates generally to non-contact temperature measurement of thin films as they are deposited onto substrates, and more particularly to improved techniques for determining the location of band edge wavelengths from scattered spectra collected in real time from the thin film. Background Art

[0003] Semiconductor nanostructures and LEDs are typically fabricated using any of a variety of deposition techniques, including chemical vapor deposition, molecular beam deposition, and sputtering. Advanced manufacturing processes involving the deposition of thin films on substrates often depend on the ability to monitor and control substrate temperature with high precision and repeatability.

[0004] In many applications, accurate temperature measurement during the growth of thin films on semiconductor wafers or substrates is critical to the final quality of the finished coated wafer and the performance of the corresponding optoelectronic devices built on the wafer. Variations in substrate temperature, including variations in temperature within the wafer, ultimately affect the quality and composition of the deposited material layer. During the deposition process, the substrate wafer is heated from the backside and rotated around a central axis. Typically, a resistive heater close to the wafer provides a heat source for raising the temperature of the wafer to a predetermined value. Careful control of this heater is required to accurately manage the growth of thin films on semiconductor wafers or substrates.

[0005] The patentee of this application, k-Space Associates, Inc. (kSA) of Dexter, Michigan, USA, is a Bandit TM The BandiT system is a well-established and highly regarded method and apparatus for measuring the temperature of semiconductor substrates. TMThe system is a non-contact, non-invasive, real-time, absolute wafer temperature sensor. Diffuse scattered light from the wafer is detected to measure the optical absorption edge wavelength. The optical absorption edge wavelength (or "band edge wavelength") enables the film temperature to be accurately determined. The BandiT system from kSA is described in detail in U.S. Patents No. 7,837,383 and No. 9,239,265. U.S. Patent No. 8,786,841 improves the BandiT system from kSA for applications where the substrate is optically transparent and therefore does not generate diffuse scattered light. According to US8,786,841 (the entire disclosure of which is also incorporated by reference), diffuse scattered light reflected from the film is collected and its spectrum is analyzed to determine the optical absorption edge as a function of film thickness.

[0006] A chemical vapor deposition technique that is advantageous in certain modern situations is called Metal Organic Chemical Vapor Deposition, or MOCVD for short. The MOCVD process must be performed within a very specific and narrow temperature range. Temperature fluctuations outside of the optimal range can affect the quality and composition of the deposited material layers. In the case of LED devices, this can result in inconsistent color output. In the MOCVD process, the substrate wafer is heated from the backside while being rotated within a controlled chamber.

[0007] Although the BandiT system from kSA has many advantages, including improvements over the technology described in US8,786,841, new technologies such as growing multiple quantum well (MQW) films by MOCVD require higher resolution to control the color output of LEDs and for other purposes. These newer technologies require higher precision to determine the location of the band edge wavelengths used to measure the film temperature.

[0008] Therefore, there is a need in the art for solutions suitable for modern thin film deposition processes that are accurate and fast enough to provide real-time data feedback so that the temperature of the substrate can be controlled with extremely high accuracy and repeatability. Summary of the invention

[0009] According to one aspect of the present invention, there is provided a method for determining the temperature of a semiconductor thin film having a measurable optical absorption edge and deposited on a substrate without a measurable optical absorption edge. The method includes the steps of providing a substrate material without a measurable optical absorption edge and depositing a thin film of semiconductor material having a measurable optical absorption edge and a measurable thickness on the substrate. The method also includes causing light to interact with the thin film deposited on the substrate to produce diffusely scattered light. The method also includes collecting the diffusely scattered light from the thin film and generating a spectrum representing the optical absorption of the thin film based on the diffusely scattered light from the thin film. The method also includes determining the film thickness. The method also includes determining the optical absorption edge wavelength of the thin film and determining the film temperature at the film thickness as a function of the film thickness and the optical absorption edge wavelength. The step of generating the spectrum includes reducing the effect of thin film interference oscillations in the optical absorption edge wavelength by using a fast Fourier transform (FFT) and a notch filter.

[0010] The present invention provides highly accurate real-time measurement of film temperature as a function of film thickness when deposited on substrates without a measurable optical absorption edge. During the film growth process, the film temperature is more accurately resolved so that the resulting products, such as light emitting diodes (LEDs), can be produced to higher standards. For example, by using the principles of the present invention, the color output of a light emitting diode can be controlled to a variation of less than 1 nm. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] These and other features and advantages of the present invention will be more readily understood in conjunction with the following detailed description and accompanying drawings, in which:

[0012] Figure 1 is a graph showing thin film interference oscillations observed in the below-bandgap (i.e., wavelength greater than the optical absorption edge) portion of the processed spectrum;

[0013] Figure 2 is a simplified flow chart summarizing the steps of the present invention;

[0014] Figure 3 Yes Figure 1 Shown but is a plot of the spectrum after the oscillations have been made periodic in wavelength by varying the variables, wherein the wavelength is replaced by the refractive index of the film material divided by the wavelength;

[0015] Figure 4 shows a typical amplitude spectrum obtained by applying the Fast Fourier Transform (FFT);

[0016] Figure 5 Yes Figure 4 The shown diagram illustrates identification of the optimum center frequency of a notch filter;

[0017] Figure 6Yes Figure 3 The plot shown is after applying the notch filter and calculating the inverse FFT, where a reduction in the oscillations can be observed;

[0018] Figure 7 Shows Figure 1 The combination of the original spectrum of and the filtered and smoothed version using the Savitsky-Göre (SG) digital smoothing filter;

[0019] Figure 8 is another example of a spectrum after applying the FFT notch filter;

[0020] Fig. 9 shows side-by-side graphs comparing the effect of finding the band edge wavelength across a semiconductor wafer without and with the additional FFT processing of the present invention;

[0021] Fig.10 The thin film interference oscillations observed in the below-bandgap (i.e., wavelength greater than the optical absorption edge) portion of the processed spectrum are shown, along with a set of integer orders m corresponding to the interference extrema. i .

[0022] Fig.11 shows that a set of integer orders m is found by calculating the sequence that minimizes the discreteness of a set of calculated thickness values i ;as well as

[0023] Fig.12 is a flow chart that concisely summarizes an auxiliary technique called optimized dispersion curve. DETAILED DESCRIPTION

[0024] The present invention relates to a method, apparatus and system for determining the temperature of a sample, the sample comprising a semiconductor thin film having a measurable optical absorption edge and a measurable thickness, and deposited on a substrate without a measurable optical absorption edge. Examples of substrates include, but are not limited to, wafers of sapphire (Al2O3), silicon dioxide (SiO2), glass, amorphous silicon carbide (SiC), and metals such as thin rolled steel, copper (Cu), aluminum (Al), molybdenum (Mo) and tantalum (Ta). Examples of thin films include semiconductor materials such as gallium nitride (GaN) used as a component of blue and white light emitting diodes (LEDs). Such thin films can be grown by any known process performed in a standard deposition chamber using known multiple quantum well (MQW) technology. Examples of processes for depositing thin films on substrates include, but are not limited to, chemical vapor deposition processes such as metal organic vapor phase epitaxy (MOVPE) and metal organic chemical vapor deposition (MOCVD), or molecular deposition processes such as molecular beam epitaxy (MBE), sputtering, or other thin film deposition processes.

[0025] As is well documented in U.S. Pat. No. 8,786,841, a controllable heat source inside the deposition chamber heats the substrate and the film. A light source mounted outside the deposition chamber produces diffusely scattered light from the sample. Optical absorption edges are also called band edges. Spectrometers (such as solid-state spectrometers or array spectrometers) generate spectra derived from or based on the diffusely scattered light from the film. The optical absorption edge wavelength of the film is determined based on the spectrum, which is usually required for metrology of semiconductor materials and film thickness.

[0026] Many new applications are emerging that require accurate real-time monitoring of thin film properties, such as temperature, as the thin film is deposited on a substrate formed of a material that does not absorb light and thus has no measurable optical absorption edge. Manufacturers of LEDs formed of sapphire substrates and GaN thin films typically require that the substrate maintain a nearly constant temperature, including deviations of 1.0°C or less, as the thin film is deposited onto the substrate. In some cases, temperature control must be maintained at + / -0.25°C.

[0027] The general correlation for the transmittance of light in a semiconductor material is given by Equation 1: I(d) / I(0)=exp(-αd)

[0028] Where d is the film thickness, I(d) is the intensity of diffusely scattered light collected from the film at film thickness d, I(0) is the intensity of diffusely scattered light collected on a substrate without the film, and α is the absorption coefficient of the film material at the band gap energy of the material. The absorption coefficient (α) of the material explains the dependence of optical absorption on the band gap energy of the material, which is temperature-dependent. The absorption coefficient (α) is also referred to as α(hν) in the formula: α(hν) = α g exp[(hn-E g ) / E0].

[0029] Equation 1 shows that the optical absorption of a film is thickness dependent and that the nature of the optical absorption is exponential. In applications where the substrate does not have a measurable optical absorption edge wavelength, such as non-semiconductors, the light is not affected by the substrate. The substrate is usually transparent (such as glass or sapphire) or completely reflective (such as steel or other metals). Therefore, the light is only affected by the film. Because the film is so thin, an increase in the film thickness can have a significant effect on the measured optical absorption edge wavelength of the film.

[0030] Incremental changes in film thickness can be accommodated by determining the optical absorption edge wavelength of the film as a function of film thickness. During the manufacturing process, the optical absorption edge wavelength and temperature are determined when the film can be adjusted to correct for non-ideal temperatures that cause undesirable properties.

[0031] The method includes depositing a thin film of semiconductor material on a substrate, the thin film of semiconductor material having a measurable optical absorption edge and a measurable thickness, heating the substrate and the thin film, and allowing an optical signal to interact with the thin film deposited on the substrate to produce diffusely scattered light. The method then includes generating a spectrum showing the optical absorption of the thin film based on the diffusely scattered light from the thin film. The method also includes determining the thickness of the thin film, determining the optical absorption edge wavelength of the thin film, and determining the temperature of the thin film at the thickness of the thin film as a function of the thickness of the thin film and the optical absorption edge wavelength.

[0032] The first step may include performing a spectral acquisition to correct for potential errors due to equipment artifacts, such as the non-uniform response of silicon-based detectors used for spectral analysis from 350 nm to 600 nm, and the non-uniform output light signals of tungsten-halogen or xenon lamps in the same wavelength range. These errors may prevent the raw diffuse light signal from producing a measurable optical absorption edge at the correct wavelength location. When performing a spectral acquisition, the errors can be assumed to be in a steady state.

[0033] Spectral acquisition first involves generating a reference spectrum that represents the overall response of the system, i.e., a combination of the lamp output characteristics and the detector response, both of which are wavelength dependent. The reference spectrum is generated by allowing the light to interact with a substrate without a film (e.g., bare sapphire) and collecting any diffusely scattered light into the detector. Subsequently, a spectrometer is used to generate a reference spectrum based on the diffusely scattered light collected by allowing the light to interact only with the substrate. Spectral acquisition ends with normalizing the reference spectrum.

[0034] Each time a raw spectrum is generated based on diffusely scattered light from the film, the method includes normalizing the raw spectrum and dividing the normalized raw spectrum by a normalized reference spectrum to produce a result spectrum. Dividing the raw spectrum by the reference spectrum for each input raw spectrum is required to determine the accurate film thickness in addition to enhancing the optical absorption edge characteristics. The result spectrum is normalized and used to determine the optical absorption edge wavelength. The result spectrum provides a resolvable optical absorption edge wavelength, which is used to determine the temperature or other properties of the film.

[0035] Spectral acquisition, which includes the creation of a normalized reference spectrum, is performed each time a system component is changed. For example, the detector's viewport can be covered over time, which affects the light collected. Reference spectra can be acquired once per run, daily, weekly, or at other intervals as needed. Acquiring a reference spectrum with each run generally provides more accurate results.

[0036] The spectra of the present method and system, including reference spectra, raw spectra, and resultant spectra, are generally generated by decomposing the optical signal from the substrate into discrete wavelength components of specific light intensities. The spectrum shows the optical absorption of the film based on the diffusely scattered light from the film. The spectrum generally includes a graph of wavelength as a function of light intensity. However, the spectrum can also provide optical absorption information in other forms such as tables.

[0037] The resulting spectrum is used to determine the optical absorption edge wavelength. The optical absorption edge wavelength is the wavelength at which a material suddenly increases its absorption of electromagnetic radiation. The optical absorption edge wavelength is related to the specific material, material temperature, and material thickness. The optical absorption edge wavelength can be identified from the spectrum; it is the wavelength where the intensity transitions sharply from very low (i.e., strong absorption) to very high (i.e., strong transmission). The optical absorption edge wavelength is used to determine the substrate temperature.

[0038] The method includes creating a wavelength versus temperature calibration table for a film at a single thickness (i.e., a temperature calibration table). The temperature calibration table may also be provided to a user of the method, rather than created by the user of the method. The temperature calibration table shows the relationship between the optical absorption edge wavelength and the temperature of the film at a constant thickness. The temperature calibration table provides subsequent temperature measurements of the film based on the optical absorption edge wavelength obtained from the spectrum. The film temperature is determined by measuring the effect of the film thickness on the optical absorption edge wavelength or the dependence of the optical absorption edge wavelength on the film thickness.

[0039] The film thickness can be determined by a variety of methods. In one embodiment of the present invention, the film thickness is conveniently determined by a spectrum generated by diffusely scattered light from the film and used to determine the optical absorption edge wavelength. The spectrum includes oscillations at wavelengths above the optical absorption edge region of the spectrum (i.e., below the band gap). These oscillations are the result of thin film interference, which is similar to the interference rings that can be observed on a thin film of oil. The wavelength-related peaks and valleys of these oscillations are analyzed to determine the film thickness. The following formula 2 can be used to determine the film thickness: Wherein, d is the film thickness, λ1 is the wavelength at the first peak of the oscillation, λ2 is the wavelength at the second peak of the oscillation adjacent to the first peak, or, λ1 is the wavelength at the first valley of the oscillation, λ2 is the wavelength at the second valley of the oscillation adjacent to the first valley, n1 is the predetermined refractive index associated with the semiconductor material at λ1, and n2 is the predetermined refractive index associated with the semiconductor material at λ2. The wavelengths used for λ1 and λ2 can be any two consecutive peaks or any two consecutive valleys of the oscillation. The values ​​of oscillation and film thickness obtained have a nonlinear correlation with all layers of the film. The film thickness can also be determined by using other methods. For example, it can be based on previously measured thickness as a function of deposition time, or by a laser-based reflectivity system (such as Rate Rat from k-Space Associates, Inc., Dexter, Michigan, USA). TM product) to estimate thickness.

[0040] As described above, the step of determining the optical absorption edge of the thin film as a function of the film thickness includes measuring the relationship between the optical absorption of the thin film and the film thickness. Due to the step of depositing a thin film of a semiconductor material having a measurable optical absorption edge and a measurable thickness on a substrate, the step of determining the optical absorption edge of the thin film as a function of the film thickness may also include adjusting the measured optical absorption edge wavelength value of the thin film obtained from the spectrum. The step of determining the optical absorption edge of the thin film as a function of the film thickness may also include identifying the semiconductor material of the thin film and adjusting the measured optical absorption edge wavelength value determined from the spectrum based on the semiconductor material and the film thickness to obtain an adjusted absorption edge wavelength.

[0041] The steps to determine the optical absorption edge of a film as a function of film thickness typically involve the use of a thickness calibration table. Each semiconductor material has a unique thickness calibration table. The thickness calibration table shows the wavelength of the optical absorption edge versus the thickness of the film at a constant temperature.

[0042] The thickness calibration table can be obtained by growing a thin film of semiconductor material at a constant temperature and measuring the optical absorption edge wavelength at each incremental thickness increase to produce a spectrum for each thickness, which spectrum is referred to as an '841 spectrum with reference to U.S. Patent No. 8,786,841. Specifically, if the spectrum is produced according to any teaching found in the entire disclosure of US8,786,841 (including its background section), then the spectrum used for real-time determination of the band edge temperature of the film according to the teachings of US8,786,841 is referred to herein as an '841 spectrum. The thickness calibration table can also be prepared by depositing a thin film on a substrate at a constant temperature and measuring the optical absorption edge wavelength of the film at the constant temperature and multiple thicknesses. Preparing the thickness calibration table at a constant temperature also enables the user to determine the correlation of the optical absorption edge wavelength with the thickness.

[0043] As described above, spectral acquisition is performed for each spectrum. Next, from each spectrum, the raw optical absorption edge wavelength value for each thickness at a constant temperature is determined. A certain n-order polynomial fit is performed on each raw optical absorption edge wavelength value to generate a curve of optical absorption edge wavelength versus thickness, where n is the order of the polynomial that provides the best fit to the data. This n-order polynomial correlation is used to create a thickness calibration table. The thickness calibration table is used as a thickness correction lookup for subsequent temperature measurements. The thickness calibration table illustrates the correlation between the optical absorption edge wavelength and the thickness of the film. The optical absorption edge wavelength increases as the thickness of the film increases. The thickness calibration table is made for each unique semiconductor material because different materials will produce different results. The thickness calibration table can also be provided to the user of the method instead of being made by the user. However, for each unique material, only one thickness calibration table is required to determine the film temperature at different thicknesses and temperatures. The method may include identifying the semiconductor material of the film and providing a thickness calibration table and a temperature calibration table for the identified semiconductor material. The film temperature at a certain thickness is determined based on the spectrum, the thickness calibration table, and the temperature calibration table.

[0044] Although the methods, apparatus, and systems detailed in U.S. Pat. No. 8,786,841 are generally effective, newer techniques have emerged that necessitate more accurate determination of the band edge wavelength locations for measuring film temperature. Accordingly, the present invention includes an improved method, improved apparatus, and improved system to more accurately resolve film temperature, particularly during growth of multiple quantum well (MQW) films via metal organic chemical vapor deposition (MOCVD). Preliminary testing and prototyping have demonstrated that film temperature can be determined to + / -0.25°C by utilizing the improved principles described below. Resolution accuracy in the vicinity of + / -0.25°C is necessary to control the color output of an LED to within a variation of less than 1 nm. This improved technique, referred to herein as the Fourier notch filter step, enables LED product manufacturers to achieve highly advantageous, accurate control of LED color output.

[0045] A Fourier notch filter step can be used to deconvolute the '841 spectra using a fast Fourier transform (FFT) and a notch filter. The Fourier notch filter step can reduce the intensity of thin film interference oscillations observed in the below-bandgap portion of the '841 spectra, as these oscillations can negatively impact the determination of band-edge temperatures and other important parameters obtained from the analysis of these spectra. Any step used for this application must be precise enough, yet fast enough to be implemented for real-time applications. Figure 1 An example of thin film interference oscillations observed in the below-bandgap portion of the '841 spectrum is provided.

[0046] Figure 2A summary of the steps by which the improved process is achieved is provided.

[0047] A fundamental element of this invention is the realization that the oscillations can be made periodic by varying variables. Specifically, the wavelength is replaced by the refractive index of the film material divided by the wavelength, as shown in the following formula:

[0048] Note that the refractive index is also a function of wavelength. Applying the previous formula yields Figure 3 spectrum.

[0049] Once the spectrum has been made periodic by varying the variables, its amplitude spectrum can be determined by using FFT. Figure 4 A typical amplitude spectrum produced by FFT is shown. The oscillation frequency can then be identified and removed by using a notch filter. Examples of suitable notch filters include, but are not limited to, n-Gaussian functions. Key parameters of the notch filter include the center wavelength (x0), the width parameter (Γ), and the shape parameter (n).

[0050] This functional form is useful because it tends to be flat on top and has a Gaussian falloff at the edges. Larger values ​​of the shape parameter result in steeper edges and thus stronger repulsion of neighboring values.

[0051] Note that the peak of the amplitude spectrum can be determined by finding the minimum of the second derivative. This is readily accomplished by using a digital smoothing filter, such as the filter developed by Savitsky and Gore (see "Smoothing and Differentiation of Data by Simplified Least Squares Procedures", Analytical Chemistry, pages 1627-1639, Vol. 36, No. 8, 1964.). The Savitsky-Gore (SG) digital smoothing filter is particularly well suited for this application because it achieves a high degree of smoothing while still preserving the higher order moments of the original distribution (i.e., without flattening the peaks as with other such filters). It has the added benefit that, since a polynomial is fit to the original data, the smoothed derivative is then readily obtained. The ability to conveniently determine the optimum center frequency of the notch filter allows the process to be automated, making it suitable for real-time applications. See Figure 5 .

[0052] After applying the notch filter, the inverse FFT is calculated. Figure 6As can be seen in , the oscillations are greatly attenuated. The changes in the variables are then reversed to reconvert the filtered spectrum back into a function of the original wavelength. Note that the filtered result can be further improved by the optional application of the Savitsky-Göret (SG) digital smoothing filter mentioned above. Figure 7 The original spectrum and the filtered and smoothed version are shown. As can be seen from the figure, the spectrum is now significantly clearer and easier to analyze. Figure 8 Another example of a spectrum after applying a Fourier notch filter is shown.

[0053] Fig. 9 The effect of finding the band edge wavelength across a semiconductor wafer is shown without additional FFT processing (left) and with additional FFT processing as a comparison (right). The jumps in the data without FFT processing are recognized as artifacts of the measurement process. When thin film interference fringes remain at the band edge, the measured band edge position will jump. After FFT processing, the band edge wavelength is smooth and continuous across the wafer.

[0054] The present invention contemplates an auxiliary technique known as optimized dispersion curve. Optimized dispersion curve can be used to improve the performance of the Fourier notch filter step. Optimized dispersion curve includes improving the accuracy of the dispersion curve used for notch filter calculation. At any extreme wavelength λ i In this case, the optical thickness (thickness multiplied by the refractive index) is an integer multiple of λ / 4:

[0055] See also Fig.10 .

[0056] A set of integer order m i It can be obtained by calculating the sequence that minimizes the dispersion of a set of calculated thickness values. Note that this requires the corresponding extreme wavelength λ i The refractive index n i knowledge.

[0057] See also Fig.11 .

[0058] This result can be used to solve the optimal index n at each extreme value i :

[0059] Here d is the average of a set of calculated thickness values.

[0060] The resulting dispersion curve of refractive index as a function of extreme wavelength can be conveniently parameterized by fitting a 3-parameter Sellmeier function of the type used by the '841 spectroscopy: Note that there are other possible parameterizations, but this formalism has the advantage of being simple and works well in the energy range below the bandgap for many commonly used semiconductor materials.

[0061] Fig.12 The flowchart of provides a concise summary of the auxiliary technique for optimizing the dispersion curve. By using this method, the dispersion n(λ) of the film can be more accurately determined, which in turn produces a more accurate FFT and notch filter that removes the dominant frequency in the spectrum.

[0062] In addition, the present invention also contemplates an alternative method by which the film thickness can be determined by the method. This can be referred to as a step for estimating the film thickness based on a pair of extreme wavelengths. According to this improvement, the film thickness is determined by analyzing the wavelength position of the interference extremes. At any extreme value λ i At each point, the optical thickness is an integer multiple of λ / 4:

[0063] The film thickness can be solved:

[0064] The following conditions hold at any extremum:

[0065] Furthermore, adjacent extreme values ​​(such as a given peak and its adjacent valley) have an integer order difference of 1:

[0066] One of ordinary skill in the art can use this to estimate d from any pair of adjacent extreme values: The values ​​of wavelength and rate used here may correspond to any peak and its preceding or succeeding trough in the oscillation.

[0067] By observing the change in d calculated for all adjacent extrema, one of ordinary skill in the art can generate an estimate of the error or "precision" in the thickness measurement. From the above results, the interference order of the extrema with the longest wavelength can also be estimated:

[0068] In summary, the present invention is an improvement over the teaching of U.S. Pat. No. 8,786,841, which allows the film temperature to be more accurately resolved to + / -0.25°C during the growth of multiple quantum well (MQW) films by metal organic chemical vapor deposition (MOCVD). This high level of resolution is considered necessary to control the color output of LEDs within a variation range of <1nm and is considered to be a technology of extremely high commercial value when it comes to controlling the color output of LEDs.

[0069] The above invention is described in accordance with relevant legal standards, so the description is exemplary rather than restrictive in nature. For those skilled in the art, changes and modifications to the disclosed embodiments may be obvious and fall within the scope of the present invention.

Claims

1. A method for determining the temperature of a semiconductor thin film, characterized in that The semiconductor film has a measurable optical absorption edge and is deposited on a substrate material without a measurable optical absorption edge, and the method comprises the following steps: a) providing a substrate of a material having no measurable optical absorption edge; b) depositing a thin film of semiconductor material having a measurable optical absorption edge and a measurable thickness on a substrate; c) causing the light to interact with a thin film deposited on a substrate to produce diffusely scattered light; d) collecting diffusely scattered light from the film; e) generating a spectrum showing the optical absorption of the film based on diffusely scattered light from the film; f) determine the film thickness; g) determining the optical absorption edge wavelength of the film based on the spectrum, and h) determining the film temperature at said film thickness as a function of the film thickness and the optical absorption edge wavelength; i) wherein the step of generating the spectrum includes reducing the intensity of thin film interference oscillations in the optical absorption edge wavelength by using a fast Fourier transform and a notch filter.

2. The method according to claim 1, characterized in that The notch filter of step i) comprises an n-Gaussian function.

3. The method according to claim 1, characterized in that The notch filter of step i) comprises the function: Where x0 is the central wavelength, Γ is the width parameter, and n is the shape parameter.

4. The method according to claim 1, characterized in that The notch filtering of step i) involves determining the peaks in the amplitude spectrum by finding the minimum of the second derivative.

5. The method according to claim 1, characterized in that The notch filter of step i) comprises determining the optimum center frequency by using a digital smoothing filter.

6. The method according to claim 1, characterized in that It also includes the step of calculating the inverse FFT after applying the notch filter of step i).

7. The method according to claim 6, characterized in that It also includes applying a digital smoothing filter after calculating the inverse FFT.

8. The method according to claim 1, characterized in that The notch filter of step i) includes drawing a dispersion curve and further includes the step of optimizing the dispersion curve by determining a set of integer orders.

9. The method according to claim 8, characterized in that The step of determining a set of integer orders includes calculating a sequence that minimizes the dispersion of a set of calculated thickness values.

10. The method according to claim 1, characterized in that Determining the film thickness in step f) includes estimating the film thickness based on a pair of extreme wavelengths.

11. The method according to claim 10, characterized in that Estimating the film thickness involves analyzing the wavelength position of the interference extrema according to the following formula: So that at any extreme value λ i At λ / 4, the optical thickness is an integer multiple of λ / 4.

12. The method according to claim 11, characterized in that It also includes solving the film thickness d according to the following formula:

13. The method according to claim 12, characterized in that It also includes estimating the film thickness d from any pair of adjacent extreme values ​​according to the following formula:

14. The method according to claim 13, characterized in that Also included is generating an estimate of the error in the film thickness measurement as a function of d based on the calculation results of all adjacent extreme values.

15. The method according to claim 14, characterized in that It also includes an estimate of the interference order for the longest wavelength extremum according to the following formula:

16. The method according to claim 1, characterized in that This includes re-determining the film temperature at incremental thicknesses to detect temperature changes resulting from step b).

17. The method according to claim 1, characterized in that The method comprises providing a temperature calibration table, wherein the temperature calibration table shows the relationship between the optical absorption edge wavelength and the temperature of the film at a constant thickness, providing a thickness calibration table, wherein the thickness calibration table shows the relationship between the optical absorption edge wavelength and the thickness of the film at a constant temperature, and The step h) comprises determining the difference between the optical absorption edge wavelength at the thickness determined in the step f) and the optical absorption edge wavelength at the thickness in the temperature calibration table by using the thickness calibration table to obtain a wavelength difference; subtracting the wavelength difference from the optical absorption edge wavelength determined in step g) to provide an adjusted optical absorption wavelength value; as well as The film temperature at the thickness determined in step f) is determined by using the adjusted optical absorption wavelength value.

18. The method according to claim 1, characterized in that The step g) includes metering the semiconductor material and the film thickness.

19. The method according to claim 1, characterized in that The step b) comprises a metal organic chemical vapor deposition method.

20. A method for determining the temperature of a semiconductor film during the growth of a multi-quantum well film, the semiconductor film having a measurable optical absorption edge and deposited on a substrate material having no measurable optical absorption edge, the method comprising the following steps: a) providing a substrate of a material having no measurable optical absorption edge; b) depositing a semiconductor material thin film having a measurable optical absorption edge and a measurable thickness on a substrate by metal organic chemical vapor deposition; c) causing the light to interact with a thin film deposited on a substrate to produce diffusely scattered light; d) collecting diffusely scattered light from the film; e) generating a spectrum showing the optical absorption of the film based on the diffusely scattered light from the film; f) determine the film thickness; g) determining the optical absorption edge wavelength of the film based on the spectrum, and h) determining the film temperature at the film thickness as a function of the film thickness and the optical absorption edge wavelength; i) wherein the step of generating the spectrum includes reducing the intensity of thin film interference oscillations in the optical absorption edge wavelength by using a fast Fourier transform and a notch filter.

Images