Machine Learning in Metrology Measurements

Abstract

Metrology methods and targets are provided, that expand metrological procedures beyond current technologies into multi-layered targets, quasi-periodic targets and device-like targets, without having to introduce offsets along the critical direction of the device design. Machine learning algorithm application to measurements and/or simulations of metrology measurements of metrology targets are disclosed for deriving metrology data such as overlays from multi-layered target and corresponding configurations of targets are provided to enable such measurements. Quasi-periodic targets which are based on device patterns are shown to improve the similarity between target and device designs. Offsets are introduced only in non-critical direction and/or sensitivity is calibrated to enable, together with the solutions for multi-layer measurements and quasi-periodic target measurements, direct device optical metrology measurements.

What is claimed is: 
     
         1 . A metrology target comprising:
 at least one portion of a device design having N>2 overlapping layers, which is selected to have a plurality of irregularly repeating units, having specified features, along at least one direction of the portion, and   a plurality of additional cells comprising at least multi-layer cells and sensitivity calibration cells,
 wherein the multi-layer cells comprise at least one of: N cells with selected intended offsets; N or fewer cells with selected intended offsets configured to utilize pupil information; and N-cell calibration targets alongside between N−1 and two overlay targets, 
 wherein the cells are configured according to parameters of at least one machine learning algorithm applied on measurements and/or simulations of the metrology target. 
   
     
     
         2 . The metrology target of  claim 1 , wherein the sensitivity calibration cells comprise at least two target cells with introduced offsets that are adjacent to the device portion. 
     
     
         3 . The metrology target of  claim 2 , wherein the introduced offsets are orthogonal to a critical direction of the portion of the device and wherein the device portion has no intended offset along the critical direction thereof. 
     
     
         4 . The metrology target of  claim 2 , wherein the parameters of at least one of the adjacent target cells and the sensitivity calibration targets are selected, according to the parameters of the at least one machine learning algorithm, to reduce inaccuracy according to a model of the inaccuracy. 
     
     
         5 . The metrology target of  claim 1 , wherein the sensitivity calibration cells are on scribe lines. 
     
     
         6 . The metrology target of  claim 1 , wherein the at least one portion comprises a plurality of device design portions selected to yield a derived pupil plane image from respective pupil images of the portions, which satisfies a specified criterion. 
     
     
         7 . The metrology target of  claim 1 , wherein the specified features comprise at least one set of lines and cuts. 
     
     
         8 . A target design file of the metrology target of any one of  claim 1 . 
     
     
         9 . A method comprising:
 configuring a multi-layered metrology target to have a plurality, M, of target cells over at least three, N≤M, target layers, each cell having at least one periodic structure in each layer, and configuring the periodic structures of each cell to be offset with respect to each other by specified offsets,   measuring, scatterometrically, at least M differential signals from the multi-layered metrology target, and   applying at least one machine learning algorithm to the differential signals and to the specified offsets, to calculate scatterometry overlay (SCOL) metrology parameters from the M measurements of the multi-layered metrology target by solving a set of M equations that relate the SCOL metrology parameters to the differential signals and to the specified offsets.   
     
     
         10 . The method of  claim 9 , wherein the multi-layered metrology target comprises M<2N cells, and wherein the applying at least one machine learning algorithm is further configured to extract overlay information from the M<2N cells. 
     
     
         11 . The method of  claim 9 , further comprising training the at least one machine learning algorithm on target designs of the multi-layered metrology target which are based on metrology simulations, to match a behavior of the target designs to a specified device patterns behavior. 
     
     
         12 . The method of  claim 9 , wherein the at least one machine learning algorithm is derived during setup and/or training and applied in runtime. 
     
     
         13 . The method of  claim 9 , wherein the multi-layered metrology target comprises a single cell per target, and wherein the at least one machine learning algorithm is further configured to enable model-free on-the-fly optical overlay measurements of the single cell. 
     
     
         14 . The method of  claim 9 , wherein the SCOL metrology parameters are overlays between the N layers. 
     
     
         15 . The method of  claim 9 , wherein the application of the at least one machine learning algorithm to calculate the SCOL metrology parameters is carried out sequentially for consecutive layers. 
     
     
         16 . The method of  claim 9 , wherein the application of the at least one machine learning algorithm to calculate the SCOL metrology parameters is carried out simultaneously for the layers, by carrying out the measuring at a pupil plane with respect to the target and using measurements of a plurality of pixel positions at the pupil plane. 
     
     
         17 . The method of  claim 9 , carried out at least partially by at least one computer processor. 
     
     
         18 . A computer program product comprising a non-transitory computer readable storage medium having computer readable program embodied therewith and configured to carry out at least partially the method of  claim 9 . 
     
     
         19 . A metrology module comprising the computer program product of  claim 18 . 
     
     
         20 . A target design file of targets designed according to the method of  claim 9 . 
     
     
         21 . A multi-layered metrology target comprising a plurality of target cells over at least three target layers, each cell having at least one periodic structure in each layer, wherein the periodic structures of each cell are offset with respect to each other by specified offsets, wherein the cells are configured according to parameters of at least one machine learning algorithm applied on measurements and/or simulations of the metrology target. 
     
     
         22 . A method comprising measuring at least one metrology parameter in at least one target cell without introducing an intended offset along a critical measurement direction into the at least one target cell by applying at least one machine learning algorithm to calibrate at least one sensitivity parameter using offsets in at least one of: (i) an orthogonal, non-critical measurement direction and (ii) at least one additional target cell other than the at least one target cell. 
     
     
         23 . The method of  claim 22 , further comprising training the at least one machine learning algorithm on target designs which are based on metrology simulations, to match a behavior of target designs to a specified device patterns behavior. 
     
     
         24 . The method of  claim 22 , wherein the offsets in the orthogonal direction are introduced into at least one additional target cell other than the at least one target cell. 
     
     
         25 . The method of  claim 24 , wherein the at least one additional target cell is adjacent to the at least one target cell. 
     
     
         26 . The method of  claim 24 , wherein the at least one additional target cell is a calibration target positioned on scribe lines. 
     
     
         27 . The method of  claim 24 , further comprising selecting, according to parameters of the at least one machine learning algorithm, parameters of the at least one additional target cell to reduce inaccuracy according to a model of the inaccuracy. 
     
     
         28 . The method of  claim 22 , wherein the at least one target cell comprises at least a part of a device design. 
     
     
         29 . The method of  claim 28 , further comprising introducing the offsets in the orthogonal direction into at least one additional target cell adjacent to the at least one target cell, according to parameters of the at least one machine learning algorithm. 
     
     
         30 . The method of  claim 28 , wherein the offsets are introduced along the orthogonal, non-critical measurement direction of the device design, according to parameters of the at least one machine learning algorithm. 
     
     
         31 . The method of  claim 28 , further comprising introducing the offsets in at least one calibration target positioned on scribe lines. 
     
     
         32 . A computer program product comprising a non-transitory computer readable storage medium having computer readable program embodied therewith and configured to carry out at least partially the method of  claim 22 . 
     
     
         33 . A metrology target comprising:
 at least one target cell without an intended offset along a critical measurement direction of the at least one target cell, and   at least two additional cells having intended offsets along the critical measurement direction of the at least one target cell,   wherein the intended offsets are derived according to the parameters of at least one machine learning algorithm applied to measurements and/or simulations of metrology measurements of the metrology target.   
     
     
         34 . The metrology target of  claim 33 , wherein the at least two additional cells have an orthogonal critical measurement direction with respect to the at least one target cell. 
     
     
         35 . The metrology target of  claim 33 , wherein the at least two additional cells are adjacent to the at least one target cell. 
     
     
         36 . The metrology target of  claim 33 , wherein the at least two additional cells are calibration targets on scribe lines. 
     
     
         37 . The metrology target of  claim 33 , wherein the at least one target cell comprises at least a part of a device design. 
     
     
         38 . A metrology target comprising at least one target cell without an intended offset along a critical measurement direction of the at least one target cell, and having intended offsets along a non-critical measurement direction of the at least one target cell, wherein the intended offsets are derived according to the parameters of at least one machine learning algorithm applied to measurements and/or simulations of metrology measurements of the metrology target.