Method for manufacturing a photonic waveguide master based on a dynamic mask of a dmd without splicing

Abstract

The application relates to the technical field of optical waveguide preparation, and discloses an optical waveguide master plate non-splicing preparation method based on a DMD dynamic mask, layout data of an optical waveguide master plate is acquired as initial design information, a mask pattern sequence is generated, and the mask pattern sequence comprises a plurality of partition mask patterns; a pixel mapping relationship between micro-mirror unit switching states and the partition mask patterns is established; a transparent substrate is driven to perform partition positioning movement on a platform, a current exposure area of a photosensitive material layer is controlled to be aligned with a projection area of the partition mask patterns; based on the pixel mapping relationship and a partition positioning movement result, the mask patterns are activated, non-interval continuous exposure is carried out, and a whole-page latent image pattern is formed; the whole-page latent image pattern is developed, a surface relief structure of the optical waveguide master plate is processed, and the surface relief structure is transferred to the inside of the transparent substrate to form a waveguide core layer pattern, the effect of non-interval continuous exposure and no segmentation splicing is realized, the preparation period of the surface relief optical waveguide master plate is shortened, and the processing efficiency and the batch preparation yield are improved.

Description

Technical Field

[0001] This invention relates to the field of optical waveguide fabrication technology, and more specifically, to a method for fabricating optical waveguide master plates without splicing based on DMD dynamic masks. Background Technology

[0002] Optical waveguide master fabrication refers to the technology of forming waveguide core layer patterns on a substrate through micro-nano fabrication methods, which is a key step in the manufacturing of integrated optoelectronic devices.

[0003] In existing technologies, large-area master images typically employ a partitioned exposure and stitching process. This involves dividing the master image into multiple sub-regions, exposing each region sequentially using digital micromirror devices or masks, and then mechanically stitching the sub-regions together to form a complete image. However, this stitching process suffers from problems such as accumulated alignment errors, uneven light intensity at the stitching seams, and the need for segmented stitching, preventing single-pass molding. These issues lead to breakpoints in the waveguide circuitry at the stitching points or decreased coupling efficiency, making it difficult to guarantee the continuity of the overall image and the yield rate of the components. Summary of the Invention

[0004] This invention provides a method for fabricating optical waveguide master plates without splicing based on DMD dynamic masks. This method solves the problems of pattern breakage and deterioration of coupling efficiency caused by the accumulation of positioning errors and timing asynchrony in existing partitioned splicing exposure. The design pattern is automatically divided into a dynamic mask sequence, achieving continuous exposure without gaps and eliminating the need for segmented splicing. This significantly shortens the fabrication cycle of the surface-embossed optical waveguide master plate and greatly improves processing efficiency and batch production yield.

[0005] To achieve the above objectives, this invention provides a method for splicing-free fabrication of optical waveguide master plates based on DMD dynamic masks, comprising:

[0006] The layout data of the optical waveguide master is obtained as the initial design information, and the initial design information is divided into grids to generate a mask pattern sequence, wherein the mask pattern sequence contains multiple partitioned mask patterns arranged in the exposure sequence.

[0007] The mask pattern sequence is loaded onto the surface of the micromirror array of the digital micromirror device, and a pixel mapping relationship between the micromirror unit switching state and the partitioned mask pattern is established.

[0008] A photosensitive material layer is coated on the surface of a transparent substrate, and the transparent substrate is driven to perform partitioned positioning movement on a platform to control the alignment of the current exposure area of ​​the photosensitive material layer with the projection area of ​​the partitioned mask pattern;

[0009] Based on the pixel mapping relationship and the partition positioning and movement results, each partition mask pattern in the mask pattern sequence is activated sequentially, and continuous exposure without interval is performed on the surface of the photosensitive material layer to form a whole latent image pattern.

[0010] The whole plate latent image pattern is developed to remove the non-cured area of the photosensitive material layer to obtain a waveguide master relief structure, and the waveguide master relief structure is subjected to etching process to transfer the processed waveguide master relief structure to the inside of the transparent substrate to form a waveguide core layer pattern.

[0011] Further, when obtaining the layout data of the waveguide master as initial design information, it includes:

[0012] Extracting waveguide line path vector data from optical design software, wherein the waveguide line path vector data includes straight line path segments, curved line path segments and cross-coupling segments;

[0013] Converting the waveguide line path vector data into a coordinate point array set, wherein each unit of the coordinate point array set records path line coordinates and corresponding path width parameters;

[0014] Identifying the functional partition boundary in the waveguide line path topology to generate a partition boundary coordinate sequence;

[0015] Inserting reference mark coordinates at the end point positions of the partition boundary coordinate sequence, wherein the reference mark coordinates include alignment cross mark and coded identification mark;

[0016] Integrating the coordinate point array set, the partition boundary coordinate sequence and the reference mark coordinates into a data file of the initial design information.

[0017] Further, when the initial design information is subjected to grid segmentation processing to generate a mask pattern sequence, it includes:

[0018] Determining a single exposure field of view range parameter according to the physical size of the micro-mirror array of the digital micro-mirror device;

[0019] Based on the single exposure field of view range parameter, the complete layout corresponding to the initial design information is divided into a plurality of rectangular grid units, wherein each rectangular grid unit contains independent grid number and grid boundary coordinates;

[0020] Iterating through each rectangular grid unit to determine whether the current grid unit contains a complete waveguide line path segment.

[0021] Further, after determining whether the current grid unit contains a complete waveguide line path segment, it further includes:

[0022] If the current grid unit contains a complete waveguide line path segment, mark the current grid unit as an independent mask area, and extract the path pattern in the current grid unit as a partition mask pattern;

[0023] If the current grid cell contains a waveguide line path segment crossing the grid, the current grid cell and the adjacent grid cells are merged into an extended mask region, and the merged path pattern is extracted as a partition mask pattern;

[0024] All partition mask patterns are arranged in the order of grid numbers to generate the mask pattern sequence.

[0025] Further, when the mask pattern sequence is loaded on the surface of the digital micro-mirror device micro-mirror array and the pixel mapping relationship between the micro-mirror unit switching state and the partition mask pattern is established, it comprises:

[0026] Each partition mask pattern is converted into a binary bitmap, wherein the pixel size of the binary bitmap is consistent with the row and column number of the micro-mirror array;

[0027] A bidirectional lookup table is established between the pixel coordinates of the binary bitmap and the physical address of the micro-mirror unit;

[0028] According to the bidirectional lookup table, the light-transmitting pixel region in the binary bitmap is mapped to the open state of the micro-mirror unit, and the light-blocking pixel region is mapped to the closed state of the micro-mirror unit;

[0029] The switching state configuration data of the micro-mirror unit is cached into the frame memory of the digital micro-mirror device to obtain a plurality of frame buffer areas corresponding to the mask pattern sequence;

[0030] A frame switching trigger signal interface is set to make the frame memory activate each frame buffer area in turn in response to an external synchronization pulse.

[0031] Further, when driving the transparent substrate to perform partition positioning movement on the platform, it comprises:

[0032] The actual fiducial mark image on the surface of the photosensitive material layer is captured by an optical alignment sensor;

[0033] The center coordinates of the actual fiducial mark image are extracted, and the alignment deviation between the center coordinates and the theoretical coordinates is calculated;

[0034] According to the alignment deviation, a platform compensation movement instruction is generated, wherein the platform compensation movement instruction contains an X-axis compensation value, a Y-axis compensation value, and a rotation angle compensation value;

[0035] The platform executes the platform compensation movement instruction;

[0036] After completing the current partition exposure, a platform stepping movement instruction is generated according to the grid boundary coordinates corresponding to the next partition mask pattern, so that the next exposure area of the photosensitive material layer enters the projection field of view range.

[0037] Further, in generating the platform compensation movement instruction according to the alignment deviation amount, comprising:

[0038] decomposing the alignment deviation amount into a translation component and a rotation component, wherein the translation component comprises an X-axis deviation value and a Y-axis deviation value, and the rotation component comprises an angle deviation value;

[0039] obtaining a historical compensation record of the platform in a current process cycle, and constructing a deviation fluctuation statistical matrix recording the distribution dispersion degree of the exposure alignment deviation of each partition;

[0040] determining an environmental interference factor based on the proportional relationship between the translation component and the rotation component of the alignment deviation amount, and combining the dispersion characteristics of the deviation fluctuation statistical matrix;

[0041] determining a compensation weight coefficient according to the grid boundary overlap rate of the current exposure area and the adjacent exposed area, wherein the compensation weight coefficient is negatively correlated with the grid boundary overlap rate and the value range is 0 to 1;

[0042] taking the product value of the environmental interference factor and the compensation weight coefficient as a comprehensive compensation correction coefficient;

[0043] correcting the translation component and the rotation component based on the comprehensive compensation correction coefficient respectively, and generating a platform compensation movement instruction.

[0044] Further, in generating the platform step movement instruction according to the grid boundary coordinates corresponding to the next partition mask pattern, comprising:

[0045] obtaining a set of grid boundary coordinates corresponding to all uncompleted exposure partition mask patterns, and constructing a topological connection graph of the to-be-exposed area, wherein the topological connection graph records the adjacent connection relationship between each grid unit to form a connection network;

[0046] calculating a fiducial mark coverage rate based on the distribution density of the fiducial mark coordinates in the to-be-exposed area;

[0047] traversing each grid unit in the topological connection graph, comparing the calculated fiducial mark coverage rate with a preset coverage threshold, retaining the grid units with a coverage rate greater than the preset coverage threshold to form a set of exposable areas;

[0048] calculating the distance from one grid unit to another grid unit inside the set of exposable areas, and adding up the distances between all adjacent grid units to obtain a total movement distance as a path total length parameter;

[0049] comprehensively evaluating the path total length parameter and the fiducial mark coverage rate according to a preset weight to construct a path optimization index;

[0050] Comparing the path optimization indexes corresponding to all grid cells in the set of light-exposure areas, selecting a grid cell with the minimum index, extracting grid boundary coordinates, and generating platform step movement instructions.

[0051] Further, when sequentially activating each partition mask pattern in the mask pattern sequence based on the pixel mapping relationship and the partition positioning movement result, performing non-interval continuous exposure on the surface of the photosensitive material layer, and forming a whole-plate latent image pattern, the method comprises the following steps.

[0052] Based on the pixel mapping relationship, loading micro-mirror cell switch state configuration data corresponding to the current partition mask pattern from the frame memory to the micro-mirror array;

[0053] Turning on the shutter of the ultraviolet light source, so that ultraviolet light is irradiated to the micro-mirror array and reflected to the current exposure area of the photosensitive material layer;

[0054] According to the photosensitive sensitivity parameter of the photosensitive material layer, the exposure time length is set;

[0055] Turning off the shutter of the ultraviolet light source, and completing the exposure process of the current partition mask pattern;

[0056] Triggering the frame switching trigger signal interface, loading the configuration data of the next partition mask pattern based on the pixel mapping relationship, and synchronously executing the step movement of the platform;

[0057] Repeating the above steps until all partition mask patterns in the mask pattern sequence are exposed, and a continuous whole-plate latent image pattern is formed on the surface of the photosensitive material layer.

[0058] Further, when the optical waveguide master plate relief structure is processed by etching, the method comprises the following steps.

[0059] Placing the developed transparent substrate in the reaction chamber of the etching equipment;

[0060] Passing etching gas into the reaction chamber;

[0061] Using the optical waveguide master plate relief structure as a mask layer, performing plasma etching on the exposed area of the transparent substrate to form the waveguide core layer pattern.

[0062] Compared with the prior art, the method has the following advantages:

[0063] The application discloses a light waveguide master plate splicing-free preparation method based on a DMD dynamic mask, layout data of the light waveguide master plate is obtained as initial design information, a mask pattern sequence is generated, and the mask pattern sequence comprises a plurality of partition mask patterns; a pixel mapping relationship between micro-mirror unit switching states and the partition mask patterns is established; the transparent substrate is driven to perform partition positioning movement on the platform, and a current exposure area of the photosensitive material layer is aligned with a projection area of the partition mask pattern; based on the pixel mapping relationship and the partition positioning movement result, the mask pattern is activated, continuous exposure without interval is performed, and a whole-plate latent image pattern is formed; the whole-plate latent image pattern is developed, the light waveguide master plate relief structure is processed, and the light waveguide master plate relief structure is transferred to the inside of the transparent substrate to form a waveguide core layer pattern, the effect of continuous exposure without interval and without segmentation splicing is realized, the preparation period of the surface relief light waveguide master plate is shortened, and the processing efficiency and the batch preparation yield are improved. BRIEF DESCRIPTION OF DRAWINGS

[0064] Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The detailed description is made with reference to the accompanying drawings.

[0065] Figure 1 A flowchart of the light waveguide master plate splicing-free preparation method based on the DMD dynamic mask in the embodiment of the application is shown. DETAILED DESCRIPTION

[0066] The specific embodiments of the application will be further described in detail below with reference to the accompanying drawings and examples. The following examples are used to illustrate the application, but are not used to limit the scope of the application.

[0067] In the description of the present application, it should be understood that the terms "center", "upper", "lower", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, and are only used to facilitate the description of the present application and simplify the description, and therefore cannot be understood as indicating or implying that the device or element referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as limiting the present application.

[0068] The terms "first", "second", "third", etc. are only used for description purpose, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the technical features indicated. Therefore, the features defined with "first", "second", etc. can explicitly or implicitly include one or more of the features. In the description of the present application, unless otherwise specified, the meaning of "a plurality of" is two or more.

[0069] In the description of the present application, it should be noted that unless otherwise explicitly specified and limited, the terms "mounting", "connection", "connecting" should be understood in a broad sense, for example, can be fixed connection, can also be detachable connection, or integral connection; can be mechanical connection, can also be electrical connection; can be direct connection, can also be indirect connection through an intermediate medium, can be internal communication of two elements. For those skilled in the art, the specific meaning of the above terms in the present application can be understood according to the specific circumstances.

[0070] The following is a description of the preferred embodiments of the present application in conjunction with the accompanying drawings.

[0071] As Figure 1 shown, the embodiments of the present application disclose a light waveguide master plate splicing-free preparation method based on DMD dynamic mask, comprising:

[0072] S110: Obtain the layout data of the light waveguide master plate as initial design information, and perform grid segmentation processing on the initial design information to generate a mask pattern sequence, wherein the mask pattern sequence contains a plurality of partition mask patterns arranged in exposure timing;

[0073] S120: Load the mask pattern sequence on the micromirror array surface of the digital micromirror device, and establish a pixel mapping relationship between the switching state of the micromirror unit and the partition mask pattern;

[0074] S130: Coating a photosensitive material layer on the surface of the transparent substrate, driving the transparent substrate to perform partition positioning movement on the platform, and controlling the current exposure area of the photosensitive material layer to align with the projection area of the partition mask pattern;

[0075] S140: Based on the pixel mapping relationship and the partition positioning movement result, sequentially activate each partition mask pattern in the mask pattern sequence, and perform continuous exposure without interval on the surface of the photosensitive material layer to form an entire plate latent image pattern;

[0076] S150: Develop the entire plate latent image pattern, remove the non-cured area of the photosensitive material layer, obtain the light waveguide master plate relief structure, and perform etching process on the light waveguide master plate relief structure, and transfer the processed light waveguide master plate relief structure to the inside of the transparent substrate to form a waveguide core layer pattern.

[0077] In some embodiments of the present application, when obtaining the layout data of the light waveguide master plate as initial design information, comprising:

[0078] Extracting waveguide line path vector data from optical design software, wherein the waveguide line path vector data contains straight line path segments, curved line path segments and cross-coupling segments;

[0079] convert the waveguide path vector data into a coordinate point array set, wherein each unit of the coordinate point array set records path line coordinates and corresponding path width parameters;

[0080] identify functional partition boundaries in the waveguide path topology, and generate a partition boundary coordinate sequence;

[0081] insert reference mark coordinates at the end point positions of the partition boundary coordinate sequence, wherein the reference mark coordinates include alignment cross marks and coded identification marks;

[0082] integrate the coordinate point array set, the partition boundary coordinate sequence, and the reference mark coordinates into a data file of the initial design information.

[0083] In this embodiment, the waveguide path vector data is derived from OptiBPM software, containing 500 to 2000 path segments, each defined by start point coordinates (x1, y1) and end point coordinates (x2, y2), and a curvature radius parameter for curved path segments.

[0084] In this embodiment, when converting the coordinate point array set, each path is sampled at 0.5 micrometer intervals to generate a discrete coordinate sequence, with each unit recording path line coordinates (xi, yi) and path width parameter wi at that point (usually 8 micrometers to 12 micrometers). The functional partition boundary identification is based on the functional module division of the waveguide device, such as the input coupling area, beam splitting area, and output coupling area, to generate a partition boundary coordinate sequence, with each partition boundary consisting of 4 vertex coordinates forming a rectangular area. Reference mark coordinates are inserted at the end point positions of the 8 partition boundaries, with the alignment cross mark being a 100 micrometer x 100 micrometer cross line pattern, and the coded identification mark being a 7-bit binary coded pattern for partition number identification. The data file is stored in JSON format, containing three main keys: "waveguide_paths", "partition_boundaries", and "alignment_marks", with the file size controlled within 5 megabytes.

[0085] The above technical solution has the beneficial effects of: through the standardized data extraction and coordinate conversion process, the integrity and accuracy of the initial design information are ensured, providing a reliable data basis for subsequent grid segmentation and mask generation, and reducing the error risk caused by incompatible data formats.

[0086] In some embodiments of the present application, when the initial design information is subjected to grid segmentation processing to generate a mask pattern sequence, it includes:

[0087] determine a single exposure field of view range parameter according to the physical size of the digital micromirror device's micromirror array;

[0088] According to the single-exposure field range parameter, the complete layout corresponding to the initial design information is divided into a plurality of rectangular grid units, wherein each rectangular grid unit contains an independent grid number and a grid boundary coordinate;

[0089] Each rectangular grid unit is traversed to determine whether a complete waveguide line path segment is contained in the current grid unit.

[0090] In some embodiments of the present application, after determining whether a complete waveguide line path segment is contained in the current grid unit, the method further comprises:

[0091] If the current grid unit contains a complete waveguide line path segment, the current grid unit is marked as an independent mask region, and the path pattern in the current grid unit is extracted as a partition mask pattern;

[0092] If the current grid unit contains a waveguide line path segment across the grid, the current grid unit and the adjacent grid unit are merged into an extended mask region, and the merged path pattern is extracted as a partition mask pattern;

[0093] All partition mask patterns are arranged in order of grid number to generate the mask pattern sequence.

[0094] In this embodiment, the physical size of the micro-mirror array is 20.736 mm x 11.664 mm, and considering the 1:5 demagnification of the projection objective, the single-exposure field range parameter is determined to be 4.0 mm x 2.2 mm. The complete layout is divided into 20 x 15 rectangular grid units according to this size, and the grid number format of each grid unit is G_i_j (i is the row number and j is the column number), and the grid boundary coordinates contain two points, the lower left corner (x0, y0) and the upper right corner (x1, y1).

[0095] In this embodiment, if all the sample point coordinates of a path segment fall within the same grid unit, it is determined to be a complete path segment; otherwise, it is a cross-grid path segment. For the cross-grid case, an eight-neighborhood merging strategy is adopted to merge the current grid unit and its eight adjacent grid units above, below, left, right, and diagonally, forming an extended mask region of up to 12 mm x 6.6 mm. When extracting the partition mask pattern, the circumscribed rectangle of the merged region is taken as the boundary, and all path patterns located in the region are cropped to generate an independent bitmap file. The mask pattern sequence is arranged in a zigzag scanning order from left to right and from top to bottom, generating about 280 partition mask patterns to ensure no path cutting.

[0096] The above technical solution has the beneficial effects of: through the adaptive extended mask region mechanism, the contradiction between the fixed field of view of the DMD and the complex waveguide topology is solved, the microstructure defects caused by forced segmentation are avoided, the pattern integrity of long line segment waveguides and dense intersection nodes is ensured, and the waveguide transmission loss consistency is improved.

[0097] In some embodiments of the present application, when the mask pattern sequence is loaded on the surface of the digital micro-mirror device micro-mirror array, and the mapping relationship between the micro-mirror unit switching state and the pixel of the partitioned mask pattern is established, it includes:

[0098] Converting each partitioned mask pattern into a binary bitmap, wherein the pixel size of the binary bitmap is consistent with the row and column number of the micro-mirror array;

[0099] Establishing a bidirectional lookup table between the pixel coordinates of the binary bitmap and the physical address of the micro-mirror unit;

[0100] According to the bidirectional lookup table, the light-transmitting pixel area in the binary bitmap is mapped to the open state of the micro-mirror unit, and the light-blocking pixel area is mapped to the closed state of the micro-mirror unit;

[0101] Caching the switching state configuration data of the micro-mirror unit to the frame memory of the digital micro-mirror device to obtain a plurality of frame buffer areas corresponding to the mask pattern sequence;

[0102] Setting a frame switching trigger signal interface to enable the frame memory to activate each frame buffer area in turn in response to an external synchronization pulse.

[0103] In this embodiment, when the partitioned mask pattern is converted into a binary bitmap, the path width area is assigned a value of 255 (light-transmitting), and the background area is assigned a value of 0 (light-blocking). The bitmap resolution is fixed at 1920x1080 pixels, which is consistent with the resolution of the DLP9500 chip. When the bidirectional lookup table is constructed, the pixel coordinates (u, v) are mapped to the physical address (s, d) of the micro-mirror unit, and the mapping formula is s=u, d=v, forming a direct coordinate correspondence relationship. The lookup table is stored as a two-dimensional array structure.

[0104] In this embodiment, the light-transmitting pixel area determination criterion is that the bitmap pixel value is greater than 128, and the corresponding micro-mirror unit is tilted by +12 degrees to be in the open state, and the reflected light enters the projection light path; the light-blocking pixel area corresponds to the micro-mirror unit tilted by -12 degrees to be in the closed state, and the reflected light deviates from the projection light path. The frame memory is configured as 32 frame buffer areas, each buffer area stores 1920x1080x1 bits of data, about 2.6 megabits, and the total cache capacity is 83.2 megabits. The frame switching trigger signal interface uses TTL level signal, the rising edge triggers the switching, and the signal frequency is set to 10-20 Hz, which is synchronized with the exposure beat.

[0105] The beneficial effects of the above technical solution are: through pixel-level mapping and frame buffer preloading mechanism, the mask pattern is switched at nanosecond level, and the uneven exposure energy caused by pattern switching delay is avoided.

[0106] In some embodiments of the present application, when driving the transparent substrate to perform partition positioning movement on the platform, the method comprises:

[0107] capturing actual fiducial mark images of the photosensitive material layer surface by the optical alignment sensor;

[0108] extracting the center coordinates of the actual fiducial mark images, and calculating the alignment deviation between the center coordinates and the theoretical coordinates;

[0109] generating platform compensation movement instructions according to the alignment deviation, wherein the platform compensation movement instructions include X-axis compensation value, Y-axis compensation value and rotation angle compensation value;

[0110] driving the platform to perform the platform compensation movement instructions;

[0111] After completing the current partition exposure, generating platform step movement instructions according to the grid boundary coordinates corresponding to the next partition mask pattern, so that the next exposure area of the photosensitive material layer enters the projection field of view range.

[0112] In the present embodiment, the platform refers to a precise displacement platform.

[0113] In the present embodiment, the fiducial mark images captured by the optical alignment sensor are 100 microns x 100 microns cross lines, the image processing algorithm uses sub-pixel edge detection to extract the center coordinates of the cross lines, the positioning accuracy is 0.1 pixels, and the spatial resolution is 0.2 microns.

[0114] In the present embodiment, the alignment deviation includes a translation component and a rotation component. When determining the translation component, the horizontal coordinate value of the center point of the actual fiducial mark image is extracted, the value is subtracted from the corresponding theoretical horizontal coordinate value, the absolute value of the subtraction result is taken to obtain the X-axis deviation value, that is, dx; the actual center point vertical coordinate value is subtracted from the theoretical vertical coordinate value, and the absolute value is taken to obtain the Y-axis deviation value, that is, dy. When determining the rotation component, the actual center point coordinates of two fiducial mark images are captured, the slope of the line connecting the two actual center points is calculated, the slope is converted into an angle value (the controller has a built-in arctangent function calculation module, the obtained slope value is input into the module, and the module outputs the corresponding angle value), the angle of the actual connecting line is obtained, the angle of the line connecting the two theoretical mark points is extracted, and the actual connecting line angle is subtracted from the theoretical connecting line angle to obtain the rotation deviation value, that is, dθ.

[0115] The beneficial effects of the above technical solution are: through the closed-loop alignment compensation mechanism, the alignment error between partition exposures is controlled within 0.5 microns, the waveguide line splicing misplacement is avoided, the optical performance consistency of the entire waveguide is ensured, and the transmission loss and crosstalk are reduced.

[0116] In some embodiments of the present application, when generating the stage compensation movement instruction according to the alignment deviation amount, the method comprises:

[0117] decomposing the alignment deviation amount into a translation component and a rotation component, wherein the translation component comprises an X-axis deviation value and a Y-axis deviation value, and the rotation component comprises an angle deviation value;

[0118] obtaining a historical compensation record of the stage in a current process cycle, and constructing a deviation fluctuation statistical matrix, wherein the deviation fluctuation statistical matrix records a distribution dispersion degree of the exposure alignment deviation of each partition;

[0119] determining an environmental interference factor based on a proportional relationship between the translation component and the rotation component of the alignment deviation amount, and in combination with a dispersion feature of the deviation fluctuation statistical matrix;

[0120] determining a compensation weight coefficient according to a grid boundary overlap rate of a current exposure area and an adjacent exposed area, wherein the compensation weight coefficient is negatively correlated with the grid boundary overlap rate and has a value range of 0 to 1;

[0121] taking a product value of the environmental interference factor and the compensation weight coefficient as a comprehensive compensation correction coefficient;

[0122] correcting the translation component and the rotation component respectively based on the comprehensive compensation correction coefficient, and generating a stage compensation movement instruction.

[0123] In the present embodiment, the pixel deviation in the image coordinate system is converted into a physical space deviation during decomposition, the X-axis deviation value dx and the Y-axis deviation value dy in the translation component are in units of microns, the rotation component dθ is obtained by calculating the angle change of the two reference mark connecting lines, and is in units of degrees, and the angle calculation accuracy is 0.01 degrees.

[0124] In the present embodiment, the historical compensation record is stored in a non-volatile memory of the displacement stage controller, records the alignment deviation data of the last 50 partitions, the constructed deviation fluctuation statistical matrix is an N×3 matrix, N is 50, three columns respectively record the historical values of dx, dy and dθ, and the matrix element is in units of microns and degrees. When calculating the distribution dispersion degree, the standard deviations σ_x, σ_y and σ_θ of the three columns of data dx, dy and dθ are respectively calculated, and the dispersion feature is taken as the three-tuple (σ_x, σ_y, σ_θ).

[0125] In the present embodiment, the proportional relationship between the translation component and the rotation component is calculated as , and r is the proportional relationship between the translation component and the rotation component, in units of degrees per micron.

[0126] In this embodiment, the environmental interference factor is determined by hierarchical evaluation and table mapping. First, the proportional relationship between the translation component and the rotation component is divided into three grades: low coupling grade, medium coupling grade and high coupling grade. The low coupling grade corresponds to a proportional relationship value less than 0.05 degrees per micrometer, the medium coupling grade corresponds to 0.05 to 0.1 degrees per micrometer, and the high coupling grade corresponds to more than 0.1 degrees per micrometer. Second, the stability level of the displacement platform is evaluated according to the deviation fluctuation statistical matrix: if the standard deviation of the three columns of data in the matrix is less than 80% of the historical average level, it is determined to be stable; if any standard deviation exceeds 120% of the historical average level, it is determined to be fluctuant; otherwise, it is determined to be general. An environmental interference factor mapping table is established, which records nine combinations of corresponding relationships: low coupling grade + stable level corresponds to environmental interference factor 1, low coupling grade + general level corresponds to environmental interference factor 1.05, low coupling grade + fluctuation level corresponds to environmental interference factor 1.1, medium coupling grade + stable level corresponds to environmental interference factor 1.02, medium coupling grade + general level corresponds to environmental interference factor 1.08, medium coupling grade + fluctuation level corresponds to environmental interference factor 1.15, high coupling grade + stable level corresponds to environmental interference factor 1.05, high coupling grade + general level corresponds to environmental interference factor 1.12, and high coupling grade + fluctuation level corresponds to environmental interference factor 1.2. According to the current calculated proportional relationship and the stability level evaluated by the matrix, the environmental interference factor is obtained by querying the mapping table.

[0127] In this embodiment, the grid boundary overlap rate η is defined as the ratio of the overlap area of the current exposure area and the adjacent exposed area to the single area, which is calculated through the grid coordinates. The compensation weight coefficient q = 1 - η / 20%. The product value of the environmental interference factor and the compensation weight coefficient is taken as the comprehensive compensation correction coefficient p. When weighting correction, the X-axis compensation value dx' = p × dx, the Y-axis compensation value dy' = p × dy, the rotation angle compensation value dθ' = p × dθ, the unit is degree, the linear proportional relationship is maintained, and the finally generated platform compensation movement instruction contains the target position value and the target angle value of the three coordinate axes. Through the EtherCAT bus, it is sent to the displacement platform driver.

[0128] The beneficial effects of the above technical solution are: through adaptive correction of the compensation amount based on historical data, environmental fluctuation interference is eliminated, and overcompensation of the overlap area is avoided, ensuring the setting accuracy of the platform compensation movement instruction.

[0129] In some embodiments of the present application, when generating the platform step movement instruction according to the grid boundary coordinates corresponding to the next partition mask pattern, it includes:

[0130] The grid boundary coordinate set corresponding to all uncompleted exposure partition mask patterns is obtained, and a topological connection graph of the to-be-exposed area is constructed, which records the adjacent connection relationship between each grid unit to form a connection network.

[0131] calculating fiducial mark coverage based on a distribution density of the fiducial mark coordinates in the to-be-exposed region;

[0132] traversing each grid cell in the topological connection graph, comparing the calculated fiducial mark coverage with a preset coverage threshold, retaining the grid cells with a coverage greater than the preset coverage threshold to form a set of exposable regions;

[0133] within the set of exposable regions, calculating the distance from one grid cell to another, accumulating the distance between all adjacent grid cells to obtain a total moving distance as a path total length parameter;

[0134] comprehensively evaluating the path total length parameter and the fiducial mark coverage according to a preset weight to construct a path optimization index;

[0135] comparing the path optimization indices corresponding to all grid cells in the set of exposable regions, selecting the grid cell with the smallest index, extracting grid boundary coordinates, and generating a platform step moving instruction.

[0136] In this embodiment, the set of grid boundary coordinates is read from a pattern segmentation module of a digital micromirror device, containing position information of two hundred and eighty unexposed partition mask patterns, and each grid cell records the coordinates of its lower left corner and upper right corner. When constructing the topological connection graph, an adjacency list method is used for storage, which automatically identifies the adjacent cell numbers in the up, down, left and right directions of each grid cell to form an undirected connection network. When calculating the fiducial mark coverage, the number of marks per square millimeter in the to-be-exposed region is counted, and the actual statistical value is divided by the preset ideal mark number to obtain the fiducial mark coverage. The coverage threshold is set to 35%.

[0137] In this embodiment, the path total length parameter is calculated using the accumulation method, starting from the current grid cell and sequentially accumulating the distance to the next adjacent grid cell. The distance value is the interval between the center points of the two grid cells. When performing weighted comprehensive evaluation, the weight of the path total length parameter is set to 0.6, and the weight of the fiducial mark coverage is set to 0.4. The path total length parameter is normalized and weighted summed with the coverage to obtain the path optimization index.

[0138] In this embodiment, the grid cell with the minimum index is selected as the next exposure target, and the platform step moving instruction is calculated according to the difference between the grid boundary coordinates, for example, the difference between the lower right corner coordinates (x1, y1) of the current grid G_i_j and the lower left corner coordinates (x0_next, y0_next) of the grid boundary coordinates G_i_(j+1) of the next exposure target is the step amount, and the platform is driven to move at a speed of 10 mm / s. For example, the lower right corner coordinates of the current grid are 30 mm and 20 mm, and the lower left corner coordinates of the target grid are 32 mm and 20 mm, so the X-axis step amount is 0.2 mm and the Y-axis step amount is 0 mm.

[0139] The beneficial effects of the above technical solution are: by combining topological path planning and coverage rate screening, the optimal exposure moving path is selected, the platform idle stroke is reduced, and at the same time, it is ensured that each exposure area has enough reference marks for alignment, and the alignment success rate is improved.

[0140] In some embodiments of the present application, when each partition mask pattern in the mask pattern sequence is activated in turn based on the pixel mapping relationship and the partition positioning movement result, and continuous exposure without interval is performed on the surface of the photosensitive material layer to form an entire plate latent image pattern, it includes:

[0141] Based on the pixel mapping relationship, the micro-mirror unit switch state configuration data corresponding to the current partition mask pattern is loaded from the frame memory to the micro-mirror array;

[0142] The shutter of the ultraviolet light source is opened, so that ultraviolet light is irradiated to the micro-mirror array and reflected to the current exposure area of the photosensitive material layer;

[0143] According to the photosensitive sensitivity parameter of the photosensitive material layer, the exposure time length is set;

[0144] The shutter of the ultraviolet light source is closed, and the exposure process of the current partition mask pattern is completed;

[0145] The frame switching trigger signal interface is triggered, the configuration data of the next partition mask pattern is loaded based on the pixel mapping relationship, and the step movement of the platform is executed synchronously;

[0146] The above steps are repeated until all partition mask patterns in the mask pattern sequence are exposed, and a continuous entire plate latent image pattern is formed on the surface of the photosensitive material layer.

[0147] In this embodiment, the ultraviolet light source shutter is a high-speed electromagnetic shutter, and the opening / closing response time is 5 ms. The shutter aperture is 25 mm. The exposure time length is set according to the exposure dose curve of SU-8 photoresist. When the thickness of the photoresist is 30 microns, the exposure time is 150 milliseconds, and when the thickness of the photoresist is 50 microns, the exposure time is 350 milliseconds.

[0148] In the embodiment, the whole-plate latent image pattern is formed by seamlessly splicing 280 sub-area patterns, and the overlapping area of the edges of adjacent sub-areas is controlled to be 10-20 microns to avoid splicing gaps.

[0149] The beneficial effects of the above technical solution are: through millisecond mask refresh synchronization with platform movement, continuous exposure without interval is realized, the performance change of the resist layer caused by stoppage waiting is eliminated, the whole-plate exposure energy uniformity is ensured, the latent image pattern continuity is good, and splicing is avoided.

[0150] In some embodiments of the present application, when the optical waveguide master relief structure is processed by etching process, the following steps are included:

[0151] The developed transparent substrate is placed in the reaction chamber of the etching equipment;

[0152] Etching gas is introduced into the reaction chamber;

[0153] The exposed area of the transparent substrate is plasma etched through the optical waveguide master relief structure as a mask layer to form the waveguide core layer pattern.

[0154] In the embodiment, the etching equipment is an inductively coupled plasma etching machine, and the temperature of the reaction chamber is controlled to be 20-25 degrees Celsius. The etching gas includes a mixed gas of carbon tetrafluoride and trifluoromethane. The exposed area of the transparent substrate is plasma etched at an etching rate of 0.5-1.5 microns per minute to form the waveguide core layer pattern. The specific implementation process is mature, and will not be described here.

[0155] The beneficial effects of the above technical solution are: the mode field matching requirements of single-mode or multi-mode waveguides are met, the insertion loss performance of waveguide devices is improved, and the optical waveguide master is prepared without splicing.

[0156] In the description of the above embodiments, specific features, structures, materials or characteristics can be combined in any one or more embodiments or examples in a suitable manner.

[0157] Although the present application has been described in the foregoing with reference to embodiments, various modifications can be made to it and components thereof can be replaced with equivalents without departing from the scope of the present application. In particular, the features of the embodiments disclosed in the present specification can be combined with each other in any manner as long as there is no structural conflict, and all combinations of these features are not described in the present specification only for the purpose of omitting the length and saving resources.

[0158] Those skilled in the art can understand that the above are only preferred embodiments of the present application, and are not used to limit the present application, and although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacements to some of the technical features. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for spliceless fabrication of photonic waveguide masters based on DMD dynamic mask, characterized in that, The method comprises the following steps: acquiring layout data of a light waveguide master as initial design information, and performing grid segmentation processing on the initial design information to generate a mask pattern sequence, wherein the mask pattern sequence comprises a plurality of partition mask patterns arranged in exposure time sequence; loading the mask pattern sequence on the micro-mirror array surface of a digital micro-mirror device, and establishing a pixel mapping relationship between the switching state of the micro-mirror unit and the partition mask pattern; coating a photosensitive material layer on the surface of a transparent substrate, driving the transparent substrate to perform partition positioning movement on the platform, and controlling the current exposure area of the photosensitive material layer to align with the projection area of the partition mask pattern; based on the pixel mapping relationship and the partition positioning movement result, sequentially activating each partition mask pattern in the mask pattern sequence to perform continuous exposure without interval on the surface of the photosensitive material layer, and forming an entire plate latent image pattern; developing the entire plate latent image pattern, removing the non-cured area of the photosensitive material layer to obtain a light waveguide master relief structure, and performing etching process on the light waveguide master relief structure, and transferring the processed light waveguide master relief structure to the inside of the transparent substrate to form a waveguide core layer pattern; when driving the transparent substrate to perform partition positioning movement on the platform, the method comprises the following steps: capturing an actual reference mark image on the surface of the photosensitive material layer by an optical alignment sensor; extracting the center coordinates of the actual reference mark image, and calculating the alignment deviation between the center coordinates and the theoretical coordinates; generating a platform compensation movement instruction according to the alignment deviation, wherein the platform compensation movement instruction comprises an X-axis compensation value, a Y-axis compensation value and a rotation angle compensation value; driving the platform to execute the platform compensation movement instruction; after completing the current partition exposure, generating a platform step movement instruction according to the grid boundary coordinates corresponding to the next partition mask pattern, so that the next exposure area of the photosensitive material layer enters the projection field of view range; when generating the platform compensation movement instruction according to the alignment deviation, the method comprises the following steps: decomposing the alignment deviation into a translation component and a rotation component, wherein the translation component comprises an X-axis deviation value and a Y-axis deviation value, and the rotation component comprises an angle deviation value; acquiring the historical compensation record of the platform in the current process cycle, constructing a deviation fluctuation statistical matrix, and recording the distribution dispersion degree of the alignment deviation of each partition exposure in the deviation fluctuation statistical matrix; determining an environmental interference factor based on the proportional relationship between the translation component and the rotation component of the alignment deviation, and combining the dispersion characteristics of the deviation fluctuation statistical matrix; determining a compensation weight coefficient according to the grid boundary overlap rate of the current exposure area and the adjacent exposed area, wherein the compensation weight coefficient is negatively correlated with the grid boundary overlap rate and the value range is 0 to 1; taking the product value of the environmental interference factor and the compensation weight coefficient as a comprehensive compensation correction coefficient; based on the comprehensive compensation correction coefficient, respectively correcting the translation component and the rotation component to generate a platform compensation movement instruction; when generating the platform step movement instruction according to the grid boundary coordinates corresponding to the next partition mask pattern, the method comprises the following steps: Obtaining a set of grid boundary coordinates corresponding to all uncompleted exposure partition mask patterns, constructing a topological connection graph of the to-be-exposed region, and recording the adjacent connection relationship between each grid unit to form a connection network; Calculating a fiducial mark coverage rate based on the distribution density of the fiducial mark coordinates in the to-be-exposed region; Traversing each grid unit in the topological connection graph, comparing the calculated fiducial mark coverage rate with a preset coverage threshold, retaining the grid units with a coverage rate greater than the preset coverage threshold to form a set of exposable regions; Within the set of exposable regions, calculating the distance from one grid unit to another, accumulating the distance between all adjacent grid units to obtain a total moving distance as a path total length parameter; Comprehensively evaluating the path total length parameter and the fiducial mark coverage rate according to a preset weight to construct a path optimization index; Comparing the path optimization index corresponding to all grid units in the set of exposable regions, selecting the grid unit with the smallest index, extracting the grid boundary coordinates, and generating a platform step movement instruction.

2. The DMD dynamic mask based photonic waveguide master template splicing-free fabrication method according to claim 1, wherein, When obtaining layout data of an optical waveguide master as initial design information, comprising: Extracting waveguide line path vector data from optical design software, wherein the waveguide line path vector data includes straight path segments, curved path segments, and cross-coupling segments; Converting the waveguide line path vector data into a set of coordinate point arrays, wherein each unit of the set of coordinate point arrays records path line coordinates and corresponding path width parameters; Identifying functional partition boundaries in the waveguide line path topological structure to generate a sequence of partition boundary coordinates; Inserting fiducial mark coordinates at the endpoint positions of the sequence of partition boundary coordinates, wherein the fiducial mark coordinates include alignment cross marks and code identification marks; Integrating the set of coordinate point arrays, the sequence of partition boundary coordinates, and the fiducial mark coordinates into a data file of the initial design information.

3. The DMD dynamic mask based photonic waveguide master template splicing-free fabrication method according to claim 1, wherein, When performing grid segmentation processing on the initial design information to generate a mask pattern sequence, comprising: Determining a single exposure field of view range parameter according to the physical size of the micro-mirror array of the digital micro-mirror device; Based on the single exposure field of view range parameter, dividing the complete layout corresponding to the initial design information into multiple rectangular grid units, wherein each rectangular grid unit contains an independent grid number and grid boundary coordinates; Traversing each rectangular grid unit to determine whether the current grid unit contains a complete waveguide line path segment.

4. The DMD dynamic mask based photonic waveguide master template splicing-free fabrication method according to claim 3, wherein, After determining whether the current grid unit contains a complete waveguide line path segment, further comprising: If the current grid unit contains a complete waveguide line path segment, marking the current grid unit as an independent mask region and extracting the path pattern within the current grid unit as a partition mask pattern; If the current grid unit contains a waveguide line path segment that crosses the grid, merging the current grid unit and adjacent grid units into an extended mask region and extracting the merged path pattern as a partition mask pattern; Arranging all partition mask patterns in grid number order to generate the mask pattern sequence.

5. The DMD dynamic mask based photonic waveguide master template splicing-free fabrication method according to claim 1, wherein, The method comprises the following steps: Converting each partitioned mask pattern into a binary bitmap, wherein the pixel size of the binary bitmap is consistent with the row and column number of the micromirror array; Establishing a bidirectional lookup table between the pixel coordinates of the binary bitmap and the physical address of the micromirror unit; According to the bidirectional lookup table, the light-transmitting pixel area in the binary bitmap is mapped to the open state of the micromirror unit, and the light-blocking pixel area is mapped to the closed state of the micromirror unit; Caching the switching state configuration data of the micromirror unit into the frame memory of the digital micromirror device to obtain a plurality of frame buffer areas corresponding to the mask pattern sequence; Setting a frame switching trigger signal interface to enable the frame memory to activate each frame buffer area in turn in response to an external synchronization pulse.

6. The DMD dynamic mask based photonic waveguide master template splicing-free fabrication method according to claim 5, wherein, When each partitioned mask pattern in the mask pattern sequence is activated in turn based on the pixel mapping relationship and the partitioned positioning movement result, and continuous exposure without interval is performed on the surface of the photosensitive material layer to form a whole-plate latent image pattern, the method comprises the following steps: Based on the pixel mapping relationship, the switching state configuration data of the micromirror unit corresponding to the current partitioned mask pattern is loaded from the frame memory to the micromirror array; Opening the shutter of the ultraviolet light source to make the ultraviolet light irradiate to the micromirror array and reflect to the current exposure area of the photosensitive material layer; According to the photosensitive sensitivity parameter of the photosensitive material layer, the exposure time length is set; Closing the shutter of the ultraviolet light source to complete the exposure process of the current partitioned mask pattern; Triggering the frame switching trigger signal interface to load the configuration data of the next partitioned mask pattern based on the pixel mapping relationship, and synchronously executing the step-by-step movement of the platform; Repeating the above steps until all the partitioned mask patterns in the mask pattern sequence are exposed to form a continuous whole-plate latent image pattern on the surface of the photosensitive material layer.

7. The DMD dynamic mask based photonic waveguide master template splicing-free fabrication method according to claim 1, wherein, When the etching process is performed on the optical waveguide master plate relief structure, the method comprises the following steps: Placing the developed transparent substrate in the reaction chamber of the etching equipment; Introducing etching gas into the reaction chamber; Using the optical waveguide master plate relief structure as a masking layer, performing plasma etching on the exposed area of the transparent substrate to form the waveguide core layer pattern.

Images