WO2022180840A1

WO2022180840A1 - Optical integrated circuit manufacturing system and manufacturing method

Info

Publication number: WO2022180840A1
Application number: PCT/JP2021/007529
Authority: WO
Inventors: 慶太山口; 藍柳原; 雅太田; 賢哉鈴木
Original assignee: 日本電信電話株式会社
Priority date: 2021-02-26
Filing date: 2021-02-26
Publication date: 2022-09-01
Also published as: JPWO2022180840A1

Abstract

Disclosed are a manufacturing system, manufacturing method, configuration of an optical circuit for improving a manufacturing method, etc., whereby an inspection result/data that can be acquired in an optical circuit manufacturing process can be more efficiently and precisely reflected in the process. This optical integrated circuit manufacturing method uses wafer in-plane process data of a core film obtained in a core film deposition process to correct a structure value of an optical waveguide in a subsequent optical waveguide fabrication process. The wafer in-plane process data of the core film includes film thickness distribution data and refractive index distribution data of the core film. In an exposure/etching process for fabricating an optical waveguide core, a distribution of correction amounts with respect to a structure value of the optical waveguide in the direction parallel to a wafer surface is determined. A corresponding mask is selected for each exposure unit region of an optical integrated circuit in the wafer in accordance with the determined distribution of correction amounts for the structure value of the optical waveguide.

Description

Optical integrated circuit manufacturing system and manufacturing method

The present invention relates to a method and manufacturing system for improving manufacturing errors and efficiency in the manufacturing process of photoperiodic circuits.

Semiconductor lasers, photodiodes, optical wavelength multiplexers/demultiplexers, optical switches, etc. are optical devices that are key to optical fiber communication. A semiconductor laser, as an optical oscillator, generates a light wave that serves as a carrier for superimposing an information signal. A photodiode operates as a device that converts the intensity of an optical signal into an electrical signal. An optical wavelength multiplexer/demultiplexer represented by an arrayed waveguide grating is used for wavelength division multiplex communication as an element for multiplexing/demultiplexing different wavelengths of light (see, for example, Non-Patent Document 2). Furthermore, optical switches have important functions in ROADM (Reconfigurable Optical Add/Drop Multiplexing) systems as elements that route optical paths. These optical devices may be configured as optical integrated circuits. In optical fiber communication, not only the optical fiber as a transmission medium but also the optical integrated circuits of these optical devices that perform optical signal processing play an important role (see, for example, Non-Patent Document 1).

An optical integrated circuit is generally composed of an optical waveguide formed on a substrate. An optical waveguide consists of a core through which an optical signal propagates and a clad surrounding it. Semiconductor lasers and photodiodes are made of semiconductor materials such as InP, and arrayed waveguide gratings and optical switches are made of optical waveguide materials mainly made of silica glass.

FIG. 1 is a diagram showing a conventional method for manufacturing an optical waveguide. Here, a typical manufacturing process will be described using a quartz-based planar lightwave circuit made of quartz-based glass as an example. First, in 1: a lower clad deposition step, a glass film that will become a lower clad 12 is deposited on a silicon substrate (wafer) 11 . The lower clad 12 is made of SiO ₂ added with P ₂ O ₅ or B ₂ O ₃ deposited by a flame hydrolysis deposition (FHD) method. The soot-like glass particles deposited by the FHD method are heated at a high temperature of 1000° C. or higher to obtain a transparent lower clad 12 . Next, in the 2: core deposition step, the FHD method is also used to deposit a thin film glass that will become the core 13 having a higher refractive index than the lower clad 12 . In depositing the core 13, the desired refractive index value can be obtained by adding GeO ₂ to SiO ₂ . 1: A transparent core 13 is formed by heating at a high temperature of 1000° C. or higher as in the lower clad deposition step.

3: In the photoresist film-forming process, a photoresist film 14 is formed on the substrate by spin coating. Next, in step 4: circuit pattern exposure, the photoresist film is irradiated with UV light 16 through a photomask 15 to expose a circuit pattern corresponding to the mask pattern. Then, in step 5: photoresist development, the circuit pattern of the photoresist film is developed to obtain a photoresist pattern 17 .

Next, 6: in the etching process, the photoresist pattern 17 is transferred to the core by reactive ion etching (RIE) to obtain a core pattern 18 . 7: In the resist removing step, the photoresist remaining on the core is removed by ashing. Finally, in step 8: upper clad deposition, an upper clad 19 is deposited by the same method as the lower clad deposition in step 1: lower clad deposition.

Various inspections such as the pattern size and optical characteristics of the optical circuit are performed on the optical waveguide obtained by the above manufacturing process. Conventionally, in order to reflect these inspection results in the manufacturing process, after the series of processes shown in FIG. 1 have been completed, manufacturing conditions reflecting the inspection results are set for each process. In this method of reflecting inspection results, manufacturing errors accumulate in each process. Due to this accumulated error, the precision of the inspection results obtained in each process becomes lower as the process shown in FIG. Accumulation of manufacturing errors can be suppressed by resetting the manufacturing conditions of the current process or adjusting the manufacturing conditions of the subsequent process based on the inspection results obtained at the end of one process. However, with such a reflection method, the manufacturing process is interrupted after each process, and there is also the problem that it is difficult to improve the throughput of the manufacturing process.

In particular, the core film formed in the core deposition step 2 in the manufacturing process of FIG. 1 has a great influence on the optical characteristics of the core of the optical waveguide. Therefore, the core film deposition process and the core processing process play an important role in determining the overall quality and cost of the manufacturing process.

There is a need for a manufacturing system and manufacturing method for optical integrated circuits that suppresses the effects of errors that occur in the core film deposition process and accurately and efficiently reflects them in the manufacturing process.

One embodiment of the present invention includes the steps of obtaining a film thickness distribution and a refractive index distribution of a core film in the plane of a wafer, and obtaining a structural value of an optical waveguide based on the film thickness distribution and the refractive index distribution determining a correction amount distribution; selecting a mask corresponding to the correction amount distribution for one or more exposure unit areas in the wafer; and exposing a photoresist film on the wafer.

Another embodiment of the present invention is an optical integrated circuit manufacturing system, comprising: selecting a mask corresponding to the correction amount distribution for one or more exposure unit regions in the wafer; and an exposure unit for exposing the photoresist film on the wafer using the selected mask.

It is possible to improve the quality and throughput of the manufacturing process of optical integrated circuits.

It is a figure explaining a series of processes of the manufacturing method of the typical optical waveguide of a prior art. 1 is a diagram showing an outline of a method for manufacturing an optical integrated circuit according to Embodiment 1; FIG. 1 is a diagram of a generalized feedforward system of Embodiment 1; FIG. FIG. 4 is a diagram showing variations in core film characteristics within a wafer surface on which an optical integrated circuit is formed; FIG. 4 is a diagram for explaining calculation of an equivalent refractive index in an optical waveguide; FIG. 5 is a diagram for explaining correction of structural values of an optical waveguide based on core film deposition process data; FIG. 10 is a schematic diagram of a manufacturing system for carrying out the method for manufacturing an optical integrated circuit according to Embodiment 2; FIG. 10 is a flowchart showing rough steps of a method for manufacturing an optical integrated circuit according to Embodiment 2; FIG. 10 is a flowchart of more specific steps of the method for manufacturing an optical integrated circuit according to Embodiment 2; FIG. 10 is a diagram of an example in which the method for manufacturing an optical integrated circuit according to Embodiment 2 is applied to an AWG; FIG. 10 is a diagram of an example in which the manufacturing method of the optical integrated circuit of Embodiment 2 is applied to MZI; FIG. 10 is a diagram showing a schematic configuration of a measuring device according to Embodiment 3; FIG. 10 is a diagram of a lower clad reflection spectrum obtained by the measurement device of Embodiment 3; FIG. 11 is a diagram for explaining a method of calculating a lower clad film thickness and a refractive index according to Embodiment 3; FIG. 10 is a diagram showing a method for measuring a multilayer film according to Embodiment 3; FIG. 10 is a diagram showing reflection spectra of cores obtained by the measurement apparatus of Embodiment 3; FIG. 11 is a diagram for explaining a method of calculating a core film thickness and a refractive index according to Embodiment 3; FIG. 11 is a diagram showing the relationship between the number of measurement points and the refractive index measurement accuracy in Embodiment 4; FIG. 10 is a diagram showing actual measurement points after fitting processing in Embodiment 4;

The inventors reviewed from a new perspective how to use the information data obtained from the inspection results regarding the reflection of the inspection results in the manufacturing process of conventional optical circuits. In the prior art series of manufacturing methods shown in FIG. 1, after the final stage inspection results are obtained, information based on the inspection results is fed back to the previous process. As a result, the values of inspection results such as optical characteristics obtained at the final stage contain cumulative errors in each process. In terms of process improvement through feedback, process conditions were changed and adjusted based on information that included cumulative errors, including processes that were not directly related. It was not sufficiently effective in terms of the amount of feedback, its accuracy, and the destination and timing of feedback.

However, if information on the processing result obtained in a certain process, for example, information on the width of the resist pattern obtained in the photolithography development process, can be known immediately after the development process, the information on the pattern width can be obtained in the subsequent etching process. can be reflected. Also, if the film thickness and refractive index of the core obtained in the core deposition process can be known immediately after the deposition process, it is possible to predict the optical characteristics of the optical waveguide formed in the subsequent photolithography process and etching process. can. In this way, by acquiring the characteristic values of the optical waveguide component obtained in the preceding pre-process during or immediately after the pre-process, the processing conditions of the succeeding post-process are reflected, and the optical properties obtained in the post-process are reflected. Can be used for prediction and improvement.

In the series of inventions of the present disclosure, process information and process data obtained in multiple manufacturing processes are "feed forwarded" to the next subsequent process during the process in which the information and data are obtained. Improvement of the manufacturing process by such feedforward is also advantageous in terms of the throughput of the manufacturing process.

The following embodiments disclose a manufacturing system, a manufacturing method, and a configuration of an optical circuit that improves the manufacturing method, in which inspection results and data that can be obtained in the manufacturing process of the optical circuit can be reflected in the process more efficiently and accurately. In particular, the present specification will focus on a manufacturing system and a manufacturing method that effectively reflect process data acquired in the core film deposition process in subsequent processes and reduce errors in the core film deposition process. First, the basic concept of the manufacturing method of the optical circuit of the present disclosure will be explained with reference to the drawings.

[Embodiment 1]
In the optical circuit manufacturing method of the present embodiment, attention is paid to data about the constituent elements or characteristics of the optical device that are formed and obtained during the execution of one step in the manufacturing process. This data can also be referred to as real data or current data. The constituent elements or characteristics of the optical device are measured, and the manufacturing conditions in the post-process are adjusted or corrected based on the data obtained by the measurement (hereinafter, this system is also referred to as "feedforward system"). The feedforward system makes it possible to suppress variations in the optical characteristics of the optical device and obtain desired optical characteristics of the finally obtained optical device.

FIG. 2 is a diagram showing a method for manufacturing an optical circuit including an optical waveguide according to Embodiment 1 of the present disclosure. The feedforward system "measures" the components of the optical device formed in one process, and performs "optical property estimation" by the optical property estimation processing 21 based on the measurement results. Then, based on the estimation result, the process control processing 22 performs “control” of the post-process.

For example, the refractive index and thickness of the lower clad film formed in the lower clad deposition step 1 in FIG. 2 and the refractive index and thickness of the core layer deposited in the core deposition step 2 are "measured". Based on the results of this measurement, the final optical properties of devices fabricated with standard (nominal) design values are estimated. Then, based on this estimation, in the subsequent etching step 6, the etching intensity or etching time is "controlled".

Specifically, based on the "measured" thickness and refractive index of the core layer and the refractive index of the clad film, the ideal core width of the pattern to meet the performance required for the optical device is estimated (predicted). do. Then, in the etching step 6, etching is performed based on this predicted value. For example, in the case of prediction information that the standard (nominal) design value is "the waveguide width after core processing is thick" and the desired performance cannot be satisfied, the etching process can be used to correct the width of the core to be formed. conduct. As an adjustment (correction) method at this time, there is a method of thickening/thinning the core width by shortening/longening the etching time or weakening/strengthening the etching intensity. Furthermore, the core width and steps in the waveguide pattern formed in the etching step 6 can be "measured". Based on this measurement result, it is also possible to "control" the refractive index of the upper clad film formed in the upper clad deposition step 8 and adjust the optical characteristics of the finally obtained optical waveguide.

As described above, the feedforward system of the embodiment of the present invention measures the shape, characteristics, etc. of the constituent elements of the optical device formed during or after the preceding process among the multiple processes of manufacturing the optical device. do. Based on this measurement result, the manufacturing conditions in the post-process are adjusted or corrected.

FIG. 3 is a generalized diagram of the feedforward system of the first embodiment. The feedforward system includes an optical device manufacturing procedure consisting of M steps, and an optical device, which is an object to be manufactured, is divided into

steps

1, 2, . . . , step i, . M order. Here, when i<j, the process j is a process later than the process i. The feedforward system includes a measurement data processing section 31 and a control data processing section 32 . The measurement data processing unit 31 executes the optical property estimation processing 21 described with reference to FIG. 2, and the control data processing unit 32 executes the process control processing 22 described with reference to FIG. The measurement data processing unit 31 and the control data processing unit 32 can be in the form of a computer configured with a CPU, RAM, ROM, and the like.

In the feedforward system in Figure 3, the solid line indicates the flow of the product according to the process. Also, dashed lines indicate measurement data obtained by "measurement" of each process, and one-dot chain lines indicate control data for "control" of each process. As shown in FIG. 3 , the feedforward system of the present embodiment acquires measurement data from the manufacturing apparatus of that process or its measurement apparatus in process i, and transfers it to the measurement data processing unit 31 . Based on the transferred measurement data, the measurement data processing unit 31 predicts the shape or characteristics of the constituent elements of the optical device formed in step i. Also, based on the measurement data, the optical properties of the finally obtained optical device can be predicted in step i.

The predicted value derived by the measurement data processing unit 31 is passed to the control data processing unit 32. Based on the predicted value, the control data processing unit 32 obtains the manufacturing conditions for the subsequent step j. The control data processing unit 32 supplies control data for the process j to be set in the manufacturing apparatus according to the obtained manufacturing conditions when the process j is performed. The control data based on the previous process, which is supplied when the post-process j is performed, may be only control data based on the previous process i, or a plurality of types of control data based on some of the previous processes. Also good. The form of the control data is, of course, determined according to conditions such as the actually constructed manufacturing apparatus and the manufacturing object.

[Embodiment 2]
In the present embodiment, an optical integrated circuit manufacturing method and manufacturing system that minimizes the influence of an error that occurs in the 2: core deposition process of the feedforward system of the first embodiment on subsequent processes will be described. This corresponds to the improvement from 2: core deposition step to 6: etching step in the prior art manufacturing process of FIG. In the feedforward system of FIG. 3, a mechanism is disclosed for feeding forward to the circuit pattern exposure process based on the process data (core film thickness distribution data and refractive index distribution data) from the process of depositing the core film. First, the effect of the core film deposition process on the post-process and the optical characteristics of the optical integrated circuit will be described, and then the method and system for manufacturing an optical integrated circuit of the present disclosure will be described.

FIG. 4 is a diagram showing variations in core film characteristics within a wafer surface on which an optical integrated circuit is formed. FIG. 4(a) is a diagram three-dimensionally showing a plot of the optical film thickness in the wafer plane and a distribution curved surface of the optical film thickness to which the plotted points are fitted. The optical film thickness (nm) normalized to the vertical axis is plotted over the in-plane (xy plane) of a wafer approximately 120 mm in diameter. The "optical film thickness" is represented by the relationship of optical film thickness=film thickness×refractive index with respect to the core film, and the distribution of this optical film thickness within the wafer surface can be obtained. Also, as core film characteristics, distributions of the film thickness and refractive index of the core film, and the refractive indices of the lower and upper clads can be obtained. These distributions are important for obtaining information on the equivalent refractive index _ne , which is an index representing the propagation characteristics of optical signals, after the optical waveguide is formed in the core film. The equivalent refractive index _ne will be further described later. FIG. 4(b) is a diagram showing the one-dimensional refractive index profile for the core film of another wafer with a diameter of 300 mm. Curves fitted according to Embodiment 4, which will be described later, are also shown.

In FIG. 4(a), it can be clearly seen that the optical film thickness is distributed within the wafer surface. Also in the example of FIG. 4B, the refractive index of the core film is distributed so as to decrease from the center of the wafer toward the periphery. Similarly, the in-plane distribution shape of the film thickness of the core film and the refractive indices of the lower and upper clads can be obtained (details of Embodiments 3 and 4). Variations in the thickness and refractive index of the core film and the refractive indices of the lower and upper clads within the wafer surface lead to variations in optical characteristics among a plurality of chips of optical integrated circuits fabricated on the entire surface of the wafer. When these film characteristics change, the equivalent refractive index _ne , which affects the phase sensed by light, changes, and even if the same circuit pattern is used, the amount of phase advance of light, which will be described later, varies within the wafer. . Further, in an optical integrated circuit having a large number of optical interference circuits in one chip, variations in optical characteristics occur for each optical interference circuit even in one chip. In addition to the optical film thickness that indicates the variation in the optical properties of the core film, another index is the equivalent refractive index n _e of the optical waveguide.

As mentioned at the beginning, an optical integrated circuit in which key optical devices are integrated contains an optical waveguide as an important component. An optical device can also be called an optical waveguide device. Therefore, it is important to suppress manufacturing errors in the structure and optical characteristics of the optical waveguide. Also, many of the important large-scale optical integrated circuits include Mach-Zehnder interferometers (MZIs) and arrayed waveguide gratings (AWGs). The functionality of MZIs and AWGs is based on interference phenomena that utilize the phase difference of light caused by one or more optical waveguide paths through which light passes. The phase of an optical signal propagating in an optical waveguide is directly affected by the propagation properties of the optical waveguide, which are determined by the equivalent refractive index _ne .

FIG. 5 is a diagram for explaining calculation of an equivalent refractive index in an optical waveguide. FIG. 5 shows the cross-sectional structure of an optical waveguide for obtaining the equivalent refractive index described below. When an optical signal having a wavelength λ propagates through an optical waveguide having a length L and an equivalent refractive index _ne , the phase lead amount φ of the optical signal is expressed by the following equation.

When the cross-sectional shape of the optical waveguide is rectangular as shown in FIG. 5, the equivalent refractive index n _e of the optical waveguide is determined by the width a and height b of the optical waveguide core, the core refractive index n _core and the clad refractive index n _clad determined from For example, in Non-Patent Document 3 and Non-Patent Document 4, the propagation constant β and the equivalent refractive index A procedure for determining n _e is disclosed. The outline of the calculation of the equivalent refractive index n _e is shown below. For the optical waveguide core having the cross-sectional configuration of FIG. 5, only the first quadrant is considered to solve the dispersion equations for the x-axis direction and the y-axis direction, respectively.

For the x-axis direction, the following equation is obtained.

Similarly, the following equation is obtained for the y-axis direction.

In the above equation, p is the number of peaks of the horizontal electromagnetic field component on the x-axis, and q is the number of peaks of the horizontal electromagnetic field component on the y-axis. The propagation constant β is obtained from the above equations (2) to (9), and the equivalent refractive index n _e is obtained as follows.

From the above equation, it can be seen that the equivalent refractive index n _e of the optical waveguide is determined from the core width a and height b, the core refractive index n _core and the clad refractive index n _clad .

The optical characteristics of optical circuits such as _MZI and AWG are strongly influenced by the equivalent refractive index ne, which represents the propagation characteristics of the optical waveguide. Therefore, in the manufacturing process of an optical integrated circuit, it is necessary to stably control the width and height of the optical waveguide core and the refractive index of the core, which determine the equivalent refractive index _ne . On the other hand, as shown in FIGS. 4A and 4B, the film thickness and refractive index of the core film are not uniform but have a distribution at the stage when the deposition process of the core film is completed. It can be understood that this is directly linked to variations in optical characteristics. It is clear that the thickness of the core film, which is the height of the optical waveguide core, and the variation in the refractive index of the core film directly affect the indicators of the optical film thickness and the equivalent refractive index _ne .

The inventors apply the information of the core film obtained at the stage of finishing the deposition process of the core film to any subsequent post-process of forming the optical waveguide, thereby suppressing the influence of the error generated in the core film deposition process. I considered whether I could not do it. Among the indices representing the state of light propagation described above, the optical film thickness of the core film (= refractive index × core film thickness), the refractive index, and the film thickness are the in-plane distribution of the wafer at the stage when the deposition process of the core film is completed. can be measured as Also, the refractive index distribution of the lower clad can be similarly measured. The difference between these set values determined from the target optical characteristics and the measured values corresponds to the amount of deviation from the respective target values, that is, the manufacturing error. In the subsequent fabrication process of the optical waveguide, it was thought that it would be possible to correct the changeable structural value of the optical waveguide so as to offset this amount of deviation.

In the process of fabricating the optical waveguide after the process of depositing the core film, structural values that can be practically changed include the core width of at least a part of the optical circuit configured on the wafer, as described in detail later. Core thickness, physical length of the optical waveguide, etc. By determining the amount of correction to these structural values that can be applied in the optical waveguide manufacturing process that follows the deposition of the core film, the error distribution of the core film within the wafer plane that occurs in the deposition process of the core film can be suppressed. should be possible. Correcting the structural values that can be changed in the manufacturing process of the optical waveguide is substantially the same as correcting the equivalent refractive index n _e in the state in which the optical waveguide is manufactured.

As already explained, the final equivalent refractive index n _e cannot be obtained unless the core film is etched, the core width is determined, and the shape of the core is determined. However, even before the fabrication of the optical waveguide, if, for example, the value of the core width is assumed to be a constant value with respect to the actual measurement data of the optical film thickness, core film thickness, and refractive index, a constant It is also possible to estimate the equivalent refractive index n _e with an accuracy of . Modification of the manufacturing conditions of the post-process so that the equivalent refractive index n e is estimated based on the estimated equivalent refractive index n _e in the pre-process of the optical waveguide manufacturing process, and the correction is added to the equivalent refractive index n _e . can. For example, if the estimated equivalent refractive index _ne becomes larger than the set value, the structural value of the optical waveguide can be corrected so that the equivalent refractive index _ne becomes smaller. Conversely, if the estimated equivalent refractive index _ne is smaller than the set value, the structural value of the optical waveguide should be corrected so that the equivalent refractive index _ne becomes larger. In this way, by applying the feedforward concept of FIG. 3 to the core film deposition process and correcting the manufacturing conditions and processes in the processes after the core film deposition process, variations in the optical characteristics of the optical integrated circuit can be suppressed and the yield can be improved. can be raised to reduce costs.

In the manufacturing method of the optical integrated circuit of the present embodiment, the wafer in-plane process data of the core film obtained in the deposition process of the core film is used to correct the structural value of the optical waveguide in the subsequent manufacturing process of the optical waveguide. The wafer in-plane process data of the core film includes film thickness distribution data of the core film and refractive index distribution data of the core film. In this embodiment, the correction amount distribution is determined for the structural values of the optical waveguide in the direction parallel to the wafer surface (xy plane) before the exposure/etching process for fabricating the optical waveguide core. Thereafter, a mask corresponding to each exposure unit area of the optical integrated circuit in the wafer is selected according to the determined correction amount distribution of the structural value of the optical waveguide. Hereinafter, details of the method for manufacturing the optical integrated circuit of this embodiment will be described with reference to the drawings.

FIG. 6 is a diagram explaining the concept of a method for correcting the structural value of the optical waveguide based on the deposition process data of the core film. The manufacturing process of the optical circuit progresses from the left side to the right side in FIG. 6, and in FIG. 2 corresponds to 2: core deposition process to 6: etching process of the manufacturing method. The upper side of FIG. 6 shows the relationship between the occurrence of an error in the core film deposition process in the manufacturing process of the conventional technology and the result thereof, and is compared with the manufacturing method of the present embodiment on the lower side. Various "values" are indicated by arrows in FIG. 6, and the vertical position of the arrow indicates the magnitude relationship of the values. In the following description, each value is defined as follows.

"Optical characteristic value" is an index value representing the function and performance of an optical circuit, and can be, for example, the center wavelength of MZI or the center wavelength of the transmission band of AWG. Although not shown in FIG. 6, the core film "target value" is the structure and properties of the core film necessary for forming a specific optical waveguide with respect to the determined target optical characteristic value 1010 (specification). It is the target value. Therefore, the core film "target value" remains unchanged as long as the design specifications for the optical properties remain the same. For example, it may be a film thickness target value or a refractive index target value of the core film.

"Core film set value" 1011 shown in FIG. 6 is a value set for an actual core deposition process in order to fabricate an optical waveguide in an actual manufacturing process. In the manufacturing process in the initial state, the core film set value 1011 is the core film target value. The core film actual measurement value 1012 includes wafer in-plane distribution data of the core film thickness and the refractive index of the core film actually measured after the core deposition process is finished. As described with reference to FIG. 4, since the core film characteristics within the wafer surface are unevenly distributed, the core film actual measurement value 1012 in FIG. In FIG. 6, as an example, it is assumed that the core film actual measurement value 1012 deviates in a direction (downward) smaller than the core film set value 1011 at a certain position within the wafer surface. Specifically, assume that the measured value of the core thickness is smaller than the set value of the core thickness. A method for measuring the refractive index of the core film, which is suitable for the manufacturing method of this embodiment, will be described later as a third embodiment.

The deviation of the above-mentioned core film measured value similarly occurs in both the manufacturing method of the prior art and the manufacturing method of the present invention. In the prior art shown in the upper part of FIG. 6, the error generated in the core film deposition process is maintained as it is until the exposure/etching process for fabricating the optical waveguide, and the resulting optical characteristic value 1013 obtained at the end of the process is also the target optical characteristic value. It deviated from 1010. In the prior art, since the core film set value 1011 was fed back after the process was finished, the problems of accuracy and throughput, etc., as already described, occurred.

On the other hand, in the manufacturing method of this embodiment shown in the lower part of FIG. 6, "correction" is performed on the structural value of the optical waveguide in the exposure/etching process after the core film deposition process. Specifically, the “correction amount” is determined with respect to the structural value of the optical waveguide core. By performing the exposure/etching process after correcting the structural value of the optical waveguide, errors and deviations detected in the core film actual measurement values 1012 caused in the core film deposition process are offset. After the end of the process, it is possible to approach a substantially equivalent state in which no error occurs in the core film set value 1011 in the core film deposition process. Applying the determined correction amount of the structural value of the optical waveguide to the exposure/etching process substantially _corrects the equivalent refractive index ne. As described above, in the manufacturing method of the present disclosure, the correction amount of the structural value of the optical waveguide is determined in advance based on the measured value distribution of the thickness and refractive index of the core film in the wafer plane, and the process is changed based on this correction amount. is applied to the exposure/etching process, the error of the resulting optical characteristic value 1016 in the final process is minimized.

In the manufacturing method of the present embodiment, the structural values of the optical waveguide to be corrected based on the measured value distribution of the film thickness and the measured value distribution of the refractive index of the core film are the optical waveguide core width and the core length of the optical waveguide constituting the optical interference circuit. , including the physical length difference between two or more optical waveguides in an optical interferometric circuit. The process change in the exposure/etching process is performed by selecting a mask corresponding to the correction amount distribution of the structural value of the optical waveguide for each exposure unit area in the exposure process of the photoresist film on the wafer. Therefore, a different mask can be selected for each chip depending on the position within the wafer plane.

The exposure unit area means the minimum area that forms an optical circuit with one selected mask. In the manufacturing process of optical integrated circuits, a semiconductor exposure apparatus can expose different transfer circuit patterns at arbitrary positions on a wafer using a projection optical system. An exposure unit area is configured within a wafer by one mask shot (transfer), and can correspond to the entire area of one chip or a partial area within the chip. Also, a plurality of exposure unit areas may exist within one chip. For example, in an optical integrated circuit in which a plurality of optical interference circuits having the same configuration are included in one chip, an area composed of one optical interference circuit is an exposure unit area. A different mask can be selected for each optical interferometer within one chip area.

FIG. 7 is a schematic diagram of a manufacturing system that implements the method for manufacturing an optical integrated circuit according to the second embodiment. The wafer flow in manufacturing system 1000 generally corresponds to the manufacturing process flow of FIGS. The manufacturing system 1000 includes a measurement unit 1001 that acquires the film thickness distribution and refractive index distribution of the core film within the plane of the wafer. Further, a core film correction amount calculation unit 1002 for determining the correction amount distribution of the structural value of the optical waveguide based on the film thickness distribution and the refractive index distribution, and an exposure/etching device 1003 are provided. The exposure/etching apparatus 1003 selects a mask corresponding to the above correction amount distribution for one or more exposure unit areas in the wafer, and uses the selected mask to apply photoresist on the wafer. Operates to expose the membrane.

In the manufacturing system 1000, information and data are exchanged as indicated by arrows. The calculation unit 1002 controls the measurement unit 1001 to measure the film thickness and the refractive index, and acquires the data of the film thickness distribution and the refractive index distribution of the core film from the measurement unit 1001 . The calculation unit 1002 also calculates the core film correction amount and provides the exposure/etching apparatus 1003 with information for determining the mask to be selected.

FIG. 8 is a flowchart showing rough steps of the method for manufacturing an optical integrated circuit according to the second embodiment. The flow of FIG. 8 corresponds to the outline of the manufacturing method of this embodiment shown from left to right in the lower part of FIG. The flow of FIG. 8 starts with a step (S1021) of setting the initial design values of the core film for realizing the target optical characteristics of the optical circuit. Next, a step of depositing a core film is performed using the core film set value, which is the initial value (S1022).

When the deposition process of the core film is finished, the film thickness and refractive index of the core film are measured within the wafer surface (S1023). Through this process, distribution data of the core film thickness and refractive index in the wafer plane as shown in FIGS. 4A and 4B can be obtained. By the step of S1023, the core film actual measurement value 1012 is obtained in FIG. The film thickness and refractive index of the core film on the wafer can be measured by non-contact evaluation of the film thickness and refractive index of the multilayer film based on the light reflection spectrum analysis described in the third embodiment. According to the measurement method of the third embodiment, the lower oxide film is measured before deposition of the core film, and the lower oxide film is fixed and analyzed during the measurement of the core film, so that the characteristics of the core film can be evaluated with high accuracy. can. Details are disclosed in the third embodiment.

Next, a core film thickness and a refractive index map are created based on the actual measurement values of the core film described above (S1024). When evaluating the measured value of the core film in the wafer plane, the throughput of this process can be improved by using the wafer in-plane fitting of the acquired information of Embodiment 4, which will be described later. Throughput can be improved by sparsely acquiring core film distribution data and fitting with a function. Improving the accuracy of film thickness and refractive index maps by acquiring data densely in areas where the change is steep (e.g., near the edge of the wafer) instead of uniformly using distribution data within the wafer surface. can be done. Details are disclosed in the fourth embodiment.

Next, a correction amount for the equivalent refractive index _ne of the waveguide core is determined (S1025). This step corresponds to the correction amount determination step 1014 in FIG. As already mentioned, the equivalent refractive index can be calculated only after the structure of the optical waveguide is determined. . A specific example in this embodiment will be further described in the detailed flow of FIG. 9, which will be described later.

Subsequently, the core set value is corrected based on the correction amount of the equivalent refractive index ne ( _S1026 ). This step corresponds to the mask selection step 1015 of FIG. By selecting a mask corresponding to the amount of correction of the equivalent refractive index n _e determined in preceding S1025, that is, the amount of correction of the structural value of the optical waveguide, the set value of the core of the optical waveguide is corrected, and the estimated equivalent refractive index A _correction of the rate ne is performed. The steps S1025 and S1026 described above correspond to "correction of equivalent refractive index n _e " in FIG.

Finally, the exposure and etching processes are performed with the corrected core set values (S1027). In the process flow of FIG. 8 described above, the general steps of the method for manufacturing an optical integrated circuit have been described from the viewpoint of correction of the estimated equivalent refractive index _ne . The manufacturing method of the optical integrated circuit of this embodiment will be further described with reference to FIG. 9 together with more specific steps and an example of the amount of correction of the structural value of the optical waveguide.

FIG. 9 is a flowchart showing specific steps of the method for manufacturing an optical integrated circuit according to the second embodiment. The flow of FIG. 9 acquires the correction amount distribution ΔW(x, y) of the width of the optical waveguide as a specific example of the correction amount of the equivalent refractive index n _e of the optical waveguide core in the steps of the manufacturing method shown in FIG. This is an example of what to do.

First, as an example of correcting the structural value of the optical waveguide, calculation of the width correction amount when correcting the width of the optical waveguide will be described. The optical waveguide width is a structural value that can be changed after deposition of the core film. is selected, correction of the equivalent refractive index _ne can be realized.

A correction amount distribution .DELTA.W(x, y) of the waveguide width in the wafer plane is obtained below. Let n _set and n(x, y) be the set value and measured value of the refractive index of the core, respectively. Also, the set value and measured value of the waveguide width are W _set and W(x, y), respectively. A correction amount distribution .DELTA.W(x, y) that can be given to the set value of the waveguide width for the core film whose film thickness is actually measured is defined by the following equation.

In equation (12), the constant A can be considered to mean the contribution of the refractive index variation to the required correction amount of the waveguide width in addition to the correction of the waveguide width.

In order to determine the correction amount distribution .DELTA.W(x, y) in Equation (12), the constant A is obtained as follows. In this example, when the measured values of the width W of the optical waveguide and the refractive index n of the core deviate from the set value n _e (W _set , n _set ), the width correction amount for correcting the equivalent refractive index is A distribution quantity ΔW(x,y) is determined. The equivalent refractive index when the waveguide width and core refractive index are slightly changed by dW and dn, respectively, is expressed by the following equation by performing Taylor expansion up to the first order on the premise that the amount of change is minute around the set value. be done.

In order for the slightly changed equivalent refractive index n _e (W _set +dW, n _set +dn) to match the set value n _e (W _set , n _set ) in the expression (13), the second line of the expression (13) , the second term should be 0, so the following equation should be satisfied.

Here, for simplicity, if the refractive index distribution n(x, y) is represented by n and the equivalent refractive index value n _e (W _set ,n _set ) is represented by n _e , dn=(n−n _set ) from the definition. Therefore, equation (14) can be transformed as follows.

From equation (15), the following equation is further obtained.

Referring to the equation (12) for the correction amount distribution ΔW(x, y) of the waveguide width and the equation (16), the constant A is expressed by the following equation.

The expression (6) of the constant A in the waveguide width correction amount distribution ΔW (x, y) means that the constant A is the differential value of the equivalent refractive index with respect to the core refractive index, and the equivalent refractive index of the core waveguide It represents the ratio of the differential value to the width. Therefore, the constant A can be determined by Equation (17) by obtaining the slope of the equivalent refractive index with respect to the change in the optical waveguide width W and the slope of the equivalent refractive index with respect to the change in the core refractive index n. The constant A is an optical circuit in an optical integrated circuit, and if the structure of the optical waveguide, the material properties of the core film, and the preparation conditions are determined, the dependence of the equivalent refractive index on the waveguide width W and the core refractive index under those conditions. should be measured. If the correction amount distribution ΔW(x, y) of the waveguide width can be determined by the equation (12), the manufacturing process of the optical integrated circuit of FIG. 9 can be carried out.

Returning to FIG. 9, an example of correction of the equivalent refractive index by correcting the width of the optical waveguide will be described. First, similarly to FIG. 8, the step of depositing the core film is performed (S1031). Furthermore, the film thickness and refractive index of the core film are measured within the wafer plane (S1032). Here, the non-contact evaluation of the film thickness and refractive index of the multilayer film based on the light reflection spectrum analysis described in the third embodiment can be used.

Subsequently, a map of the actual measured values of the core film thickness and the measured values of the refractive index is created (S1033). In this step, the method of fitting the acquired information within the wafer plane described in the fourth embodiment can be used.

If the actual measured value of the film thickness and the measured value of the refractive index of the core film are obtained, the correction amount distribution ΔW(x, y) of the waveguide width can be obtained using Equation (12) (S1034). In the wafer plane, the set value of the width W (structural value) of the optical waveguide is corrected based on the correction amount distribution ΔW (x, y), so that the equivalent refractive index predicted when the optical waveguide is completed is will approach the target equivalent refractive index. As a result, the resulting optical characteristic value of the optical circuit also approaches the target optical characteristic value.

Specifically, in the next step, the structural value of the optical waveguide is corrected by selecting a mask based on the acquired correction amount distribution ΔW(x, y) (S1035). The above S1033 to S1035 are performed by the calculator 1002 in the manufacturing system shown in FIG.

Finally, using the selected mask, the exposure/etching device 1003 in the manufacturing system of FIG. 7 performs the exposure and etching process (S1036), completing the fabrication of the optical waveguide.

Therefore, the method for manufacturing an optical integrated circuit of the present disclosure includes the step of acquiring the film thickness distribution and the refractive index distribution of the core film in the plane of the wafer (S1032), and based on the film thickness distribution and the refractive index distribution, a step of determining a correction amount distribution of structural values of an optical waveguide (S1034); and a step of selecting a mask corresponding to the correction amount distribution for one or more exposure unit regions in the wafer (S1035). and exposing a photoresist film on the wafer using the selected mask (S1036).

Further, the optical integrated circuit manufacturing system of the present disclosure includes a measurement unit 1001 that acquires the film thickness distribution and the refractive index distribution of the core film in the wafer plane, and the optical guiding unit 1001 based on the film thickness distribution and the refractive index distribution. a calculating unit 1002 for determining the correction amount distribution of the structural value of the wave path; and a mask corresponding to the correction amount distribution for one or more exposure unit regions in the wafer, and applying the selected mask. and an exposure section 1003 that is used to expose the photoresist film on the wafer.

In the manufacturing process flow of FIG. 9 described above, an example of correcting the waveguide width W as the structural value of the optical waveguide has been described. Structural values are not limited to width only, if modulus can be corrected. Even if the equivalent refractive index is different, in the direction parallel to the wafer surface (xy plane), not only the width of the optical waveguide but also the length of the optical waveguide can be corrected for the phase lead amount of the optical signal as a structural value. It is possible. As shown in equation (1), the amount of phase lead of the optical signal in the optical waveguide of the optical interference circuit is greatly affected by the error in the equivalent refractive index _ne .

For example, in-wafer variation in the waveguide length difference between two arm waveguides in an MZI, which is an optical interference circuit, is a problem because it directly affects the interference conditions along with variation in the equivalent refractive index. When correcting the waveguide length difference as a structural value, it is sufficient to prepare a plurality of masks having arm waveguides with different physical length differences. A mask having an appropriate physical length difference can be selected in the wafer plane according to the equivalent refractive index expected from the core film's measured film thickness and refractive index measured value. Thus, the structural value to be corrected based on the measured value of the core film is the core width of the optical waveguide, the length of the optical waveguides forming the optical interference circuit, or the waveguide length between the optical waveguides in the optical interference circuit. You can use the difference.

In the following more specific optical interference circuit, the waveguide length difference (optical path length difference) between the optical waveguides in the optical interference circuit is corrected based on the measured value of the film thickness and the measured value of the refractive index of the core film, and the selection of the mask An embodiment in which the equivalent refractive index is corrected by is described.

In the manufacturing method of the optical integrated circuit of the present embodiment, the waveguide physical length difference of the arm waveguide is used as the “optical waveguide structural value” to correct the optical characteristic, for example, the intra-wafer variation of the AWG center wavelength. can also be applied to

In the AWG, an optical multiplexing/demultiplexing action occurs due to the phase difference caused by the optical path length difference of each arrayed waveguide. When a wavelength-multiplexed optical signal including a plurality of optical signals with different wavelengths is input, the optical signals can be separated for each wavelength and output to different ports. The center wavelength of each output port in the AWG is determined by the phase difference caused by the optical path length difference of the arrayed waveguides. The optical path length difference in an AWG arrayed waveguide is determined by the refractive index of the core and cladding of the optical waveguides constituting the arrayed waveguide, and the shape of the core.

FIG. 10 is a diagram for explaining an example in which the method for manufacturing an optical integrated circuit according to this embodiment is applied to an AWG. Referring to Non-Patent Document 5, the theoretical formula for the central wavelength λ ₀ in the AWG is expressed by the following formula as shown in FIG.

Here, ne is the equivalent refractive index of the core, _ΔL is the waveguide physical length difference between the adjacent arm waveguides, and m is the diffraction order. In this embodiment, the target optical characteristic value is the central wavelength λ ₀ of the AWG. The waveguide physical length difference ΔL between the arms may be selected according to the expected equivalent refractive index _ne from the measured value distribution of the core film in the wafer plane. Specifically, as shown in the table of FIG. 10, if the predicted equivalent refractive index n _e is expected to be slightly smaller than the set value of the equivalent refractive index, then from equation (18), a mask with a larger ΔL should be selected. Conversely, if the expected equivalent refractive index n _e is expected to be slightly larger than the set value of the equivalent refractive index, a mask with a smaller ΔL should be selected from equation (18). If there is no deviation from the set value of the equivalent refractive index and the design value is satisfied, the standard ΔL mask may be used.

In this way, a plurality of AWG masks with different waveguide physical length differences ΔL are prepared in advance, and from the film thickness distribution and refractive index distribution of the core film in the wafer plane, the waveguide structure value is A correction amount for the physical length difference ΔL is determined. By selecting an AWG mask corresponding to the determined correction amount of ΔL from among a plurality of AWG masks, substantially the same effect as correcting the equivalent refractive index n _e can be obtained.

In the manufacturing method of the optical integrated circuit of the present embodiment, the optical path length difference of the arm waveguide is used as the "structural value of the optical waveguide" to correct the optical characteristic, for example, the variation in the center wavelength of the MZI within the wafer. can also be applied.

FIG. 11 is a diagram illustrating an example in which the method for manufacturing an optical integrated circuit of this embodiment is applied to MZI. In the MZI 1040, the output port changes due to the phase difference caused by the optical path length difference between the two

arm waveguides

1041 and 1042 formed between the

directional couplers

1043 and 1044. FIG. By heating one arm waveguide 1042 with a heater 1045, it can also be utilized as a switch. Here, when the optical path length difference ΔS between the two arm waveguides deviates from the set value, the extinction wavelength changes. The optical path length difference between the two arm waveguides in the MZI is determined by the refractive index of the core and clad forming the optical waveguide, and the shape of the core (see Non-Patent Document 5).

Non-Patent Document 5 describes the structure and operating principle of a two-input, two-output MZI switch. In the MZI switch 1040, the interference state at the

directional couplers

1043 and 1044 changes due to the optical path length difference ΔS between the two

arm waveguides

1041 and 1042, and the optical output intensity changes. Assuming that the branching ratio of each of the two directional couplers is 50%, when the optical path length difference between the two arm waveguides is 0±2nπ in terms of phase, all the light input to the input port 1 is output from the output port 2. output. On the other hand, when the optical path length difference is .pi..+-.2n.pi. A theoretical formula for the MZI transmission spectrum is also disclosed in Non-Patent Document 6.

According to Non-Patent Document 6, the output P3 of the output port ₃ when light is input from the input port 1 is expressed by the following equation.

From equation (19), it can be seen that the output P3 of output port ₃ is determined by the equivalent refractive index _ne .

In order to apply the manufacturing method of the optical integrated circuit of this embodiment to _MZI , the physical length difference between the arm waveguides ( L ₁ −L ₂ ) can be selected. Specifically, as shown in the table of FIG. 11, when the predicted equivalent refractive index n _e is slightly smaller than the set value of the equivalent refractive index, the optical path length difference ΔS becomes shorter from equation (19). . Therefore, a mask 1048 with a larger physical length difference (L ₁ -L ₂ ) should be selected. Conversely, if the predicted equivalent refractive index n _e is slightly larger than the set value of the equivalent refractive index, the optical path length difference ΔS will be long from equation (19). In this case, a mask 1046 with a smaller physical length difference (L ₁ -L ₂ ) should be selected. A standard mask 1047 having a standard physical length difference (L ₁ -L ₂ ) can be used if the equivalent refractive index does not deviate from the set value and is as designed.

In this way, a plurality of MZI masks having different physical length differences (L ₁ −L ₂ ) are prepared in advance, and the structure of the optical waveguide is determined from the film thickness distribution and refractive index distribution of the core film within the wafer plane. A correction amount for the physical length difference (L ₁ -L ₂ ) is determined as the value. By selecting a mask corresponding to the correction amount of the determined physical length difference (L ₁ −L ₂ ) from a plurality of masks for MZI, the same effect as substantially correcting the equivalent refractive index n _e is obtained.

As described in detail above, according to the manufacturing method of the optical integrated circuit of the present disclosure, the error distribution data generated in the deposition process of the core film is immediately reflected in the optical waveguide manufacturing process, which is the subsequent process. The feedforward system described in Embodiment 1 together with FIG. 3 is realized. A manufacturing process of an optical integrated circuit that suppresses variations in optical characteristics occurring within a wafer surface is realized. It is possible to realize efficiency improvement and high throughput of the manufacturing process of the optical integrated circuit.

[Embodiment 3]
A non-contact method of measuring the film thickness and refractive index in each step in the feedforward system of this embodiment will be described below.

[3.1 Measuring device]
FIG. 12 is a diagram showing a schematic configuration of a measuring device according to one embodiment of the present invention. The measuring device irradiates an optical device, which is an object to be manufactured, with a laser beam, analyzes the reflected light from the object, and measures the physical properties of the film formed on the optical device, such as film thickness and refractive index, in a non-contact manner. measure the rate. As a principle of measurement, conventional measurement methods such as a spectroscopic reflection film thickness gauge and a spectroscopic ellipsometer can be applied.

The measuring device is composed of an optical measuring system 101 and a tester 102. A silicon wafer 114 formed with an optical device to be measured is fixed to a wafer chuck 113 and moved in three axial directions by a driving mechanism 112 on a base 111 . A test head 121 connected to the tester 102 has a light transmitting optical system 122 , a light receiving optical system 123 and a control circuit 124 . The tester 1 controls the driving mechanism 112 to irradiate the laser beam from the light transmitting optical system 122 onto a desired position of the film constituting the optical device formed on the silicon wafer 114 , and the test head 121 . A command is sent to the control circuit 124 . Reflected light from the object is received by the light receiving optical system 123 , and the control circuit 124 processes the signal from the light receiving optical system 123 and sends back the measurement result to the tester 1 .

As will be described later, the tester 1 analyzes the signal from the light receiving optical system 123 and calculates the physical property values of each film formed on the silicon wafer 114 .

[3.2 Non-contact measurement method of film thickness and refractive index]
For example, a case of measuring the lower clad 12 formed in the lower clad deposition step 1 of FIGS. 1 and 2 by a spectral reflection method will be described. The tester 1 irradiates the surface of the lower clad 12 with a laser beam with a wavelength sweeping range of 450-900 nm at a predetermined incident angle from the light transmission optical system 122 of the test head 121 . From the light intensity of the light received by the light receiving optical system 123, the tester 1 calculates a reflection spectrum represented by reflectance with respect to wavelength.

FIG. 13 is a diagram showing the reflection spectrum of the lower clad obtained by the measurement device of this embodiment. The laser light emitted from the light-transmitting optical system 122 is reflected at the surface of the lower clad 12 and the interface between the substrate 11 and the lower clad 12, and the reflected light resulting from interference between the two is incident on the light-receiving optical system 123. . This light interference is reflected in the reflection spectrum, and if the film is formed uniformly, a wavy spectrum is observed due to the interference. When the film thickness is thin, the width of the waves is large (the number of waves is small), and when the film thickness is thick, the width of the waves is small (the number of waves is large). The tester 1 can calculate the film thickness and refractive index of the lower clad 12 from the wave amplitude and period of the reflection spectrum.

A method of calculating the film thickness and refractive index of the lower clad will be described with reference to FIG. The intensity E ₀ of the incident light from the light-sending optical system 122, the intensity E ₁ of the reflected light reflected from the surface of the lower clad 12, the intensity E ₂ of the reflected light reflected at the interface between the substrate 11 and the lower clad 12, and do. Furthermore, reflected light E ₃ , E _{4 ,} . . . Assuming that the light intensity of the entire reflected light is E, the following relationship holds.

formula (20)

Here, r _ij is the reflectance when incident light from the i-th layer with a refractive index n _i (0 is in the air) is reflected at the interface with the j-th layer with a refractive index n _j , and r _ij =−r _ji . In addition, t _ij is the transmittance when the interface between the i layer and the j layer is transmitted from the i layer to the j layer, and t _ji is the opposite, the film thickness d _i of the i-th layer, the refractive index n _i , the wavelength λ, the phase coefficient Δi when going back and forth in the _i -th layer is given by the following equation.

_At this time, the reflectance _r12 and the transmittance t12 for the TE(P) wave are given by the following equations.

Also, the reflectance _r12 and the transmittance _t12 for the TM(S) wave are given by the following equations.

Rearranging the above equation from the Maclaurin expansion, the reflectance r1 of the entire reflected light from the lower clad 12, which is the _first layer, is given by the following equation.

Next, the film thickness _d1 and the refractive index _n1 of the lower clad ₁₂ are calculated from the wavelength dependence of the reflectance r1, that is, the amplitude and period of the reflection spectrum.

[3.3 Multistep measurement method]
Next, a case of measuring the core 13 formed in the core deposition step 2 of FIGS. 1 and 2 by a spectral reflection method will be described. The laser light emitted from the light-transmitting optical system 122 described above is reflected at the surface of the core 13 , the interface between the core 13 and the lower clad 12 , and the interface between the substrate 11 and the lower clad 12 . When the measurement is performed by the spectral reflection method, the reflected light from the interface between the substrate 11 and the lower clad 12 becomes noise and must be removed when calculating the film thickness and refractive index. On the other hand, in the silica-based planar lightwave circuit, the refractive index difference between the core 13 and the lower clad 12 is as small as about 1%. , the reflectance is higher. In addition, since the film thickness of the core 13 is as thin as several μm, the reflected light component from the interface between the substrate 11 and the lower clad 12 is large, making it difficult to calculate the film thickness and refractive index.

Therefore, in this embodiment, in the manufacturing process for forming a multilayer film, a multi-step measurement method is applied in which the result of measurement in the previous process is reflected in the measurement result of the current process.

FIG. 15 is a diagram showing a method for measuring a multilayer film according to one embodiment of the present invention. A case of measuring the core 13 formed in the core deposition step 2 by a spectral reflection method will be described. A measurement result of the lower clad 12 formed in the lower clad deposition step 1 of the previous step is obtained. The reflection spectrum is as shown in FIG. 13, and the physical property values (n, d) of the lower clad 12 obtained therefrom, here the refractive index and film thickness are obtained (S141). Next, the measurement result of the core 13 formed in the core deposition step 2 is obtained (S142).

FIG. 16 is a diagram showing the reflection spectrum of the core obtained by the measurement device of this embodiment. As described above, the measurement result of the core 13 has a small refractive index difference between the core 13 and the lower clad 12, so that the envelope of the reflection spectrum fluctuates as shown in FIG.

FIG. 17 is a diagram illustrating a method of calculating the film thickness and refractive index of the core. The light intensity E _total of the entire reflected light incident on the light receiving optical system 123 with respect to the incident light from the light transmitting optical system 122 is given by the following equation.

As in Equation (21), the light intensity E _total is the intensity E of the total reflected light emitted from the surface of the core 13 calculated by Equation (20), and the total intensity of the reflected light reflected from the lower clad 12. An intensity E' is added. From the reflection dependency of the reflectance obtained here, that is, the reflection spectrum (FIG. 16), the difference of the reflection spectrum (FIG. 13), which is the measurement result of the previous step, is calculated (S143). The film thickness and refractive index of the core 13 are calculated from this difference, that is, the amplitude and period of the fluctuation (S144).

Furthermore, when forming three or more layers, such as forming the upper clad 19, the reflectance is obtained using the film thickness and the refractive index, which are the measurement results of the previous step, as constants, and the wavelength dependence of the reflectance is calculated. Analyze and solve for film thickness and refractive index for the current process. For example, consider the case of stacking four layers on a substrate and calculating the film thickness and refractive index of the layer formed in the final step. Taking the reflectance r ₄ of the total reflected light from the layer formed in the first step as a constant, the reflectance r _{3 + 4} of the layer formed in the second step is obtained, and the reflectance of the layer formed in the third step Find r ₂₊₃₊₄ . From the reflectance r _1+2+3+4 of the layer formed in the last step, the wavelength dependence of the reflectance is analyzed, and the solution for the film thickness and refractive index of the fourth layer formed in the last step is obtained.

According to the present embodiment, in the manufacturing process for forming a multilayer film, the film thickness and refractive index in each process are measured without contact, and the difference is analyzed to obtain the physical property values of the film formed in each process. can be obtained.

[3.4 Utilization of temperature information]
Refractive index measurements are dependent on the temperature at which they are measured. For example, SiO ₂ as a substrate material has a linear expansion coefficient of approximately 0.5×10 ⁻⁶ (room temperature to 1000° C.) and a refractive index change rate of approximately 10 ⁻⁵ /K. In optical devices that perform optical signal processing, it is generally necessary to control the refractive index with an accuracy of about 10 ⁻⁵ , so correction of measured values using temperature information is beneficial.

In the measuring apparatus of this embodiment, a plurality of thermistors are arranged on the upper surface of the wafer chuck 113 to measure the temperature at arbitrary positions on the silicon wafer 114 . When calculating the refractive index of the lower clad 12 and the core 13, the refractive index change rate with respect to the temperature is taken into account from the measured temperature. Also, when calculating the film thickness, the above linear expansion coefficient is taken into account from the measured temperature.

The temperature measurement points on the wafer may be provided at predetermined intervals, or predetermined measurement points may be provided for each optical circuit chip formed on the wafer. Furthermore, it is preferable to provide a measurement point for each functional component formed in one chip and for each functional component susceptible to refractive index conversion. In this case, the measurement points for the reflectance measurement described above are the same as the measurement points for the temperature measurement.

For example, an arrayed waveguide grating (AWG), which is a typical optical circuit, will be described. The AWG connects an input slab waveguide connected to the input waveguide and an output slab waveguide connected to the output waveguide with a plurality of arrayed waveguides each having a difference in physical waveguide length ΔL. Connected. Among the optical characteristics of the AWG, the center wavelength _λ0 is determined by the following equation as already described.

Here, _nc is the effective refractive index of the arrayed waveguide, and m is the diffraction order. The effective refractive index of an arrayed waveguide is determined by the film thickness of the clad and core, and its uniformity affects the interference characteristics. Therefore, manufacturing errors have a great influence on the accuracy of the center wavelength. Therefore, in order to manufacture a plurality of arrayed waveguides with high accuracy and uniformity, for example, temperature measurement points are provided near the input slab waveguide, the intermediate point of the arrayed waveguides, and the output slab waveguide, and the temperature distribution is derived. Keep By ensuring the measurement accuracy of the refractive index in this way, it is possible to improve the accuracy of adjusting or correcting the manufacturing conditions in the post-process.

[3.5 Characteristic prediction by machine learning]
It is conceivable to store the above-described temperature measurement results as data necessary for prediction of physical property values by machine learning. By calculating the temperature distribution of the silicon wafer 114 fixed on the wafer chuck 113 from the obtained temperature information, it is possible to obtain the temperature at an arbitrary position of the optical circuit where no thermistor is arranged, thereby obtaining more accurate refraction. It is possible to correct the rate and film thickness.

When calculating the temperature distribution, even if the relational expression required for fitting is unknown, the actual measurement data can be used as training data and learned together with the correlated feature values such as the waveguide width and height with respect to the center wavelength. The temperature distribution can be calculated by

This embodiment obtains a first measurement result of a first film obtained in a previous step in a multi-layer non-contact measurement method for measuring physical property values of each film of a multi-layer film formed on a substrate. a step, wherein the first measurement result is the light intensity of reflected light from the surface of the first film of laser light irradiated to the surface; obtaining a second measurement result of the film, wherein the second measurement result is the light intensity of the reflected light from the surface of the laser light irradiated to the surface of the second film; , a step of calculating a difference between the second measurement result and the first measurement result, which is a difference in the reflection spectrum obtained from the light intensity of the reflected light and represented by reflectance with respect to wavelength; and calculating the film thickness and refractive index of the second film from the above.

The above non-contact measurement method further comprises the step of measuring the temperature at an arbitrary position on the substrate, and the step of calculating the film thickness and the refractive index is characterized by further performing correction according to temperature.

[Embodiment 4]
This embodiment presents a method for efficiently performing the non-contact measurement of Embodiment 3 without further reducing the throughput of the manufacturing process.

Multiple identical optical devices formed on a wafer must be able to exhibit uniform characteristics and functions. However, in the manufacturing process, as described above, characteristic fluctuations cannot be avoided due to the temperature distribution in the wafer surface, and variations in the accuracy of processing such as the deposition and etching of the semiconductor material also inevitably cause characteristic fluctuations. do not have. Therefore, in the inspection of characteristics such as optical characteristics, manufacturing errors can be suppressed as the number of measurement points on the wafer increases. However, an increase in the number of measurement points reduces the throughput of the manufacturing process.

Therefore, in the present embodiment, before manufacturing a product lot, a fitting process is performed in the inspection of a prototype lot, the measurement points on the wafer are specified in advance, and the number of measurement points is reduced, thereby reducing the throughput of the manufacturing process. prevent. Details of the fitting process will be described below.

In a prototype lot, the reflectance of the lower clad 12 formed in the lower clad deposition step 1 of FIGS. 1 and 2 is measured at predetermined intervals over the entire wafer using the above-described measuring apparatus. The tester 1 calculates the film thickness and the refractive index from the reflection spectrum of the measurement results. Here again, reference is made to the one-dimensional refractive index distribution of FIG. 4(b). In FIG. 4(b), the center of a wafer with a diameter of 300 mm is set at coordinates x=0, y=0, and the refractive index obtained by measuring the straight line of y=0 at intervals of 2 mm is indicated by white circles. The horizontal axis (x-axis) is the distance from the center and the vertical axis is the refractive index.

Similarly, measurements are made at 2 mm mesh intervals across the wafer by measuring on a plurality of straight lines parallel to the y axis at 2 mm intervals in the y direction. From this measurement result, the refractive index approximation n(x, y) and the film thickness approximation t(x, y) at an arbitrary position on the wafer are generated by the following equations.

In the above two approximation formulas, a _k , b _k , c _k , and d _k are coefficients, which can be obtained using the method of least squares or the like. The solid line shown in FIG. 4(b) is the fitting function represented by this approximation, and represents the fitting result of the refractive index n(x, y) on the straight line of y=0. Although FIG. 4(b) shows the fitting to the refractive index, the same can be done for the film thickness.

Referring again to (a) of FIG. 4, the results of the fitting process are shown. FIG. 3 is a three-dimensional representation of the result of fitting processing over the entire wafer, and represents the optical film thickness (refractive index×film thickness) on a two-dimensional plane where the center of the wafer is x=0 and y=0. . The approximation formula may be obtained from one wafer in a trial lot, or may be obtained by averaging the results obtained from a plurality of wafers. In addition, the measurement points in the substrate surface of the wafer do not need to be set in a mesh pattern, but are arbitrary, but as can be seen from FIG. do. Although FIG. 4A shows an example of three-dimensional fitting to the optical film thickness, the refractive index or film thickness can also be represented three-dimensionally by performing a similar fitting process.

FIG. 18 is a diagram showing the relationship between the number of measurement points and the refractive index measurement accuracy. The measurement accuracy for a single measurement result obtained from only one measurement is shown by the dashed line as normalized to 1. Even if the number of measurement points on the wafer increases, the measurement accuracy remains constant. If the number of measurement points is increased and the fitting process is performed, the same effect as the method of measuring the same measurement point multiple times and calculating the average value is obtained, and the measurement accuracy is improved as shown by the solid line. This is based on the same principle as accuracy improvement by the central limit theorem that the mean in the sample mean distribution approaches the population mean and the sample variance in the sample mean distribution approaches 1/n of the population variance as the number of samples increases. That is, the measurement result at a certain single measurement point can suppress randomly occurring measurement variations based on the results of the surrounding measurement points, thereby improving the measurement accuracy.

Next, we will explain how to identify the measurement points. First, set the desired measurement accuracy and from the results shown in FIG. 18, determine the number of measurement points in the wafer. For example, if an entire wafer with a diameter of 300 mm is measured at mesh intervals of 2 mm as described above, the number of measurement points is about 18,000. Assuming that the predetermined measurement accuracy is a normalized value of 0.25 or less, the measurement accuracy can be satisfied if the number of measurement points is 200 or more, as shown in FIG. At this time, from the upper limit of the number of measurements obtained by dividing the time allocated for measurement in the time allowed in the manufacturing process (for example, the lower clad deposition process 1) by the measurement time required for one measurement point , may determine the number of measurement points. Second, select a measurement point at a specific location.

FIG. 19 is a diagram showing measurement points to be actually measured as a result of fitting processing. Measurement points can be selected regardless of the location of the optical devices formed on the wafer. For example, from the results of the fitting process shown in FIGS. 4A and 4B, it is conceivable to select measurement points that are in good agreement with the approximation formula. It is also conceivable to densely set the measurement points near the inflection point of the fitting function represented by the approximation formula. Furthermore, it can be set in consideration of the temperature distribution on the wafer in the manufacturing process, variations in manufacturing errors, and the like. On the other hand, if the optical functional circuit included in the optical device includes a circuit that requires precision, the position where the optical functional circuit is formed can be selected as the measurement point.

　In the production of product lots, only the measurement points obtained by the fitting process are measured. The measurement result at this measurement point and the result of the fitting process are passed to the next manufacturing process (for example, core deposition process 2), and the manufacturing conditions for the next manufacturing process are obtained. In the post-process, the result of the fitting process is preferentially used. If the distribution of the measurement results in the prototype lot is smooth and the approximate expression is represented by an appropriate function as a result of the fitting process, the result of the fitting process is better than the single measurement result as described above. , because the measurement accuracy is high.

The non-contact measurement method of the present embodiment is a non-contact measurement method for measuring physical property values of a film formed on a substrate, and the substrate surface A fitting step including a first step of obtaining an approximate expression representing the distribution of physical property values in advance, and a second step of selecting a second measurement point at a specific position, which is fewer than the first measurement points. and measuring the physical property values of the film using the measurement result obtained at the second measurement point and the approximate expression in the manufacturing process of the film.

The measurement result is the light intensity of reflected light from the surface of the film of the laser beam irradiated to the surface, and the physical property value is the film thickness and the refractive index of the film.

Another non-contact measurement method of the present embodiment is a multi-layer non-contact measurement method for measuring physical property values of each film of a multi-layer film formed on a substrate, wherein A first step of obtaining an approximate expression representing the distribution of physical property values in the substrate surface from the measurement results at the measurement points; 2 step in advance, and obtaining a first measurement result of the first film obtained in the previous step in the manufacturing process of the multilayer film, wherein the first measurement result is includes a measurement result measured at a second measurement point in the fitting process for the first film and an approximation, and a step and a second measurement result of the second film formed in the current process. and calculating a difference between the second measurement result and the first measurement result, wherein the difference between the second measurement result and the measurement result measured at the second measurement point or obtaining the difference between the second measurement result and the measurement result corresponding to the measurement point of the second measurement result obtained from the approximation formula; and the physical property value of the second film from the difference. and calculating the .

In the above-described another non-contact measurement method, the first and second measurement results are light intensities of reflected light from the surfaces of the first and second films of laser light irradiated to the surfaces. characterized by Further, in the above-mentioned another non-contact measurement method, the difference is a difference of reflection spectra expressed by reflectance with respect to wavelength, which is obtained from the light intensity of the reflected light. The physical property values may be film thicknesses and refractive indices of the first and second films.

In this embodiment, a method of performing fitting processing using one or more wafers in the inspection of a trial lot before manufacturing a product lot has been described. Even after entering the manufacturing process of the product lot, the approximation formula may be changed as needed by comparing the measurement result obtained in the manufacturing of the product lot with the result of the fitting process obtained in advance.

According to this embodiment, by specifying the measurement points on the wafer in advance by the fitting process and reducing the number of measurement points, it is possible to prevent the throughput of the manufacturing process from decreasing without lowering the measurement accuracy.

Claims

obtaining a film thickness distribution and a refractive index distribution of the core film in the plane of the wafer;
determining a correction amount distribution for structural values of the optical waveguide based on the film thickness distribution and the refractive index distribution;
selecting a mask corresponding to the correction amount distribution for one or more exposure unit areas in the wafer;
and exposing a photoresist film on said wafer using said selected mask.
The structural value includes at least one of a core width of the optical waveguide, a length of the optical waveguides forming the optical interference circuit, or a waveguide length difference between the optical waveguides in the optical interference circuit. 2. The method for manufacturing an optical integrated circuit according to claim 1.
The exposure unit area includes an optical interference circuit having two or more optical waveguides,
3. The method of manufacturing an optical integrated circuit according to claim 1, wherein the wafer includes at least one exposure unit area.
4. The method of manufacturing an optical integrated circuit according to claim 3, wherein the exposure unit areas are each included in a chip area of the wafer.
3. The correction amount distribution of the core width is determined based on the core film thickness initial setting value and refractive index initial setting value, the film thickness distribution and the refractive index distribution. 4. The method for manufacturing an optical integrated circuit according to any one of 4.
An optical integrated circuit manufacturing system,
a measurement unit that acquires the film thickness distribution and the refractive index distribution of the core film in the plane of the wafer;
a calculation unit that determines a correction amount distribution of structural values of an optical waveguide based on the film thickness distribution and the refractive index distribution;
Exposure for selecting a mask corresponding to the correction amount distribution for one or more exposure unit areas in the wafer and exposing a photoresist film on the wafer using the selected mask A manufacturing system for an optical integrated circuit, comprising: a unit;
The structural value includes at least one of a core width of the optical waveguide, a length of the optical waveguides forming the optical interference circuit, or a waveguide length difference between the optical waveguides in the optical interference circuit. The manufacturing system according to claim 6, wherein
The exposure unit area includes an optical interference circuit having two or more optical waveguides,
the wafer includes at least one exposure unit area, or
8. A manufacturing system according to claim 6 or 7, wherein each is contained within a chip area of said wafer.