TEA Systems White-Paper http://www.TEAsystems.com
TEA Systems White Paper - Sources, Detection & Control of Reticle Haze
Reticle Haze Process Control using Weir PW
In the previous section we discussed the derivation of Dose Uniformity and how the method can be used to detect the onset of reticle haze or lens degradation due to film deposition. This section illustrates how Weir Daily Monitor (Weir DM) can be used to automate the analysis in a production environment for a "two-click" selection of analysis scripts and data. The Weir DM scripts can also be used for Weir DMA analysis; a complete analysis tool allowing Weir DM to be called by external software and have the results submitted to external data files for complete automation .
The Weir DM interface, shown on the right, provides two user fields; one for the Weir DM Script name and the second field lists the data file to be analyzed. The user interface has a check box that allows the output to be put into the data log file. The File menu also contains a command that allows batch-files to be defined for the consecutive analysis of multiple data files.
Weir DM provides the macro-encapsulation of Weir PW & PSFM analyses in the form of scripts stored in Microsoft Excel Templates (".xlt"). Each template defines the analysis plus the user's selection of graphics, statistics for the current dataset, variable trend chart tracking and output of variables for export to other user programs.
Begin the setup of a Weir DM macro by first loading a data file with the focus-dose matrix data that will be used for a Dose Uniformity analysis. It's best to use a data file uses the same metrology recipe. Weir DM will use this data workbook as a template for the Weir DM script workbook.
From the Weir Main or Weir Setup interface, select the "Tools/Weir DM Setup" menu option from the menus and specify the script name and data source directory if t they are different from the name used as default. In the "Analysis Sequence" tab, select "Process Window" and, from the list presented, the variables and their target values. Insure that either "All" field sites or the site number to be selected are entered into the "Field Sites" menu. In the example shown, a BCD variable targeted to 80% +/-10% will be fitted to a process window dose analysis. The MSE or measurement error variable will be used to cull poor metrology data points from this OCD data set.
The Graphics Display tab will be used to select the type of graphics and statistics to be used for the current dataset analysis. These will be the graphics that show the performance results of the current data.
Since we want to test for Dose Uniformity, select the "Dose @ Best Focus" displays along with a contour plot of the field data.
The statistics to add to the trend charts will next be selected using the "Raw Statistics" and "Model Statistics" tabs. Raw statistics reflect the raw data as measured for the analysis. They will NOT include any data removed by the culling methods of the Analysis Sequence tab.
Model Statistics change depending upon the type of analysis as shown below.
Selecting the "Dose at Best Focus" tab allows the dose statistics needed for Dose Uniformity to be characterized while removing the significant errors cause by defocus across the field of exposure. Two sets of statistics are selected with this chart. The "push button" selection title with each variable sets the statistic to display for the current data. The check-box entry will create and add the variable to a long term, multi-dataset plot of the statistic's trend. One variable trend-chart is created for each check-mark selected.
Save the template and exit template setup.
Start the Weir DM interface and, from the interface shown in the first figure of this section, select the proper template and a data file. Next press the "Run Daily Monitor" control button. The data will be loaded and modeled according to the selections made during setup. After the analysis, the graphics showing the response of the current data will be displayed:
The displays contain both report summaries, that are also saved in the workbook spreadsheets, and the graphics selected from the sequence above.
If any raw statistics were selected from the trend charts, each graph can be displayed. More critical to our needs is a display of the dose-sensitive variables that can show the onset and growth of reticle haze as measured on FEM wafers for the example shown below.
In this example the IsoFocal Dose was analyzed across the reticle. Haze growth begins to appear with the "54AO" lot and continues until the reticle is finally cleaned. Early warning of such variance allows the reticle clean operation to be scheduled for any convenient time rather than waiting for a yield or process "bust". An even more critical analysis would be to track and key on the IsoFocal DoseRange variable that looks at the range of dose values across the reticle as opposed to the average dose shown here. Dates or data-lot names can be selected for display on the graphic along with optional statistics reflecting the overall trend data.
Weir DMA call:
-t "C:\Program\Weir Wavefront\DS1_FF_DMtemplate.XLT" -d 040615210805.txt -o D:\TEMP -x D:\TEMP
Weir DMA running options, such as the location of the analysis log file, can be found in the "Tools/Options" menu of the Weir DM interface. Since the interface runs independent of operator control, a log file is maintained and it's size controlled by Weir DM. The entries to the log file can be viewed from any Weir DM interface menu and also reside in the log-file directory, specified in the options file, to be viewed with any text editor. This screen shows the analysis of two data files.
The log file contains the data call and the summary of the analysis. A series of "line switches" are used, such as "-t" to specify the template to use, the directory and name of the data set as well as ("-o") the output directory etc.
Outputs variables can be stored as either comma-delimited ".csv" data files or in Excel Workbooks.
ã Copyright 2007 TEA Systems Corporation, All rights reserved. Legal
Corp. | Tel: +1 610 682 4146