Alignment

The JBX-5500FS electron beam writer (EBW) uses alignment marks to establish a coordinate system that it can use to expose patterns relative to a pattern on the substrate. The placement accuracy depends on the tool used to expose the alignment marks and the most accurate alignment is always the tool to itself.

Alignment Marks

There are two kinds of alignment marks. Global marks are large marks placed at the outer edges of the wafer used to correct for offset, rotation and expansion/contraction of the substrate. Chip marks are placed in the four corners of the write field and are used to adjust the deflection gain and rotation for the main and sub deflector. Chip marks are used to compensate for scan errors that arise from local variations of the sample surface. An example of how alignment marks are positioned on a substrate is shown in Fig. 1.
Figure 1: Example of alignment marks on a wafer

Figure 1: Example of alignment marks on a wafer

Global marks should be placed on the outer edges of the substrate along the X and Y axis. Four global marks per substrate is standard; however, only a pair of global marks along either the X or Y axis is required.

Standard global marks are crosses that are 2 mm long and 3 μm wide. Marks should be dry etched or wet etched at least 1 to 2 μm deep into the substrate. Metal lift-off can also be used as alignment marks, but it is not as reliable due to the chances of mark damage from subsequent processing steps such as high temperature annealing. For best results, use 100 to 200nm of dense materials such as Au on Cr, W on Ti, and Pt on Ti. In each case, 10 to 20 nm of Ti or Cr provides improved adhesion for the high-Z material to the substrate.

Chip marks must be placed at the corners of a single field. In general, a set of chip marks is usually used to covers no more than a 20mm x 20mm region. Each time a chip mark is used for alignment it is essentially rendered unusable for future alignments. If multiple levels of alignment need to be performed, then multiple sets of chip marks needs to be created. Each subsequent chip mark level should be offset toward the center of the chip by at least 100 microns in both the X and Y directions. Typically, chip marks are not needed if the alignment tolerance is greater than 500 nm. Standard chip marks are crosses that are 60 μm long and 3 μm wide.

Sample placement on Cassette

Placement and rotation errors can be large if wafer is not properly loaded onto the cassette. Jeol recommends that the rotation error should be less than 1 degree. There are a couple features designed into the 3” and 4” cassette that assists the user to load the wafer roughly aligned to the writing axis of the instrument. The first feature is a set of alignment pins that should be touching the edges of the wafer. The second feature is a spring-loaded grounding pin that pushes the wafer towards the alignment pins. Proper alignment of the wafer on the cassette will facilitate finding the global alignment mark in the instrument.

Smaller sized samples must be loaded into the piece holder cassette. The piece holder cassette has alignment pins for samples that are approximately 1 inch x 1 inch. Smaller samples must be aligned by eye. Nonetheless, a careful user can often align a sample to within 0.5 degrees. Once the sample is on the cassette, user a ruler to measure the position of the alignment mark from a reference point on the cassette. This will significantly facilitate finding the mark using the SEM mode of the EBW.

Locating the Alignment Mark

The JBX-5500FS has two coordinate systems called the Expose coordinate and the Stage coordinate. The Expose coordinate’s origin is located at the center of cassette, the positive x axis is directed towards the right and the positive y axis is directed towards the top of the wafer. The Stage coordinate’s origin is located at the top left corner of the stage, the positive x axis is directed towards the right and the positive y axis is directed towards the bottom of the stage. There is a finite offset of 62.5,37.5 mm between the Expose origin and the Stage origin. In addition, the positive y axis among the two coordinate systems are pointing in the opposite direction. The Ebeam Writer Helper program can be used to convert between the two coordinate systems.

The Expose coordinate system is the natural coordinate system for identifying the position of the alignment marks since it uses the cartesian coordinate system with its origin located at the center of the wafer. Thus, the position of the alignment mark is identified as a displacement from the center of the wafer or cassette. Normally, the position of the global alignment marks with respect to the wafer center is known. In the case that the positions of the global alignment marks are not known, roughly determine the displacement of the mark from the center of the cassette with a ruler.

The EBW has an SEM mode that an operator can use to find the global mark. The SEM mode as a field of view that is up to 1 x 1 mm square in Mode 2 and up to 0.1 x 0.1 mm square in Mode 4. A new user should operate in Mode 2 until they are comfortable with locating the alignment mark since the larger field of view makes it significantly easier to locate the mark. Global alignment marks can be up to a few millimeters off from the expected position depending on variables such as actual pattern position on sample, sample position on cassette and cassette position on the stage. If using Mode 4, it is often easier to find the alignment mark using Mode 2 first before switching to Mode 4.

User Mark Detection

The user mark is used to store the position of a single global mark. The user mark is used to determine the optimal scan and gain conditions for automatic detection of the global and chip marks. The scan and gain specifications should be aimed towards achieving robust detection of the specified alignment mark. The scan conditions depends on the mark geometry. The gain conditions depends on the beam current, mark material and resist thickness.

The scan and gain conditions can be accessed through the Calibration GUI -> Standard Mark Detection Tab -> User Mark Detection Tab. Typically, the scan parameters do not need to be changed unless the marks are different than what is specified in this document. The scan conditions consists of the following parameters:

Scan Parameters Recommended Value Description
Mark Shape Cross Specifies the shape of the mark as either cross or L-shaped. The L-shaped mark is not supported.
Scan Position X [μm] 20 Specifies the X scan position from the center of the mark
Scan Position Y [μm] 20 Specifies the Y scan position from the center of the mark
Rough scan width [μm] 400 Specifies the rough scan range used to find the mark. A larger range makes it easier to find the mark. The range is limited by the EOS mode -- 1000 for Mode 2 and 100 for Mode 4.
Fine scan width [μm] 100 Specifies the fine scan rage used to determine the position of the mark
Number of scans [times] 10 Number of times to scan the mark and average the signals to improve the signal quality.
Scan clock [nsec] 20000 Specifies the scan duration of a single scan

The gain conditions can be accessed through the Calibration GUI -> Standard Mark Detection Tab -> User Mark Detection Tab -> AGC conditions button. Typically the gain conditions must be defined once for a new combination of beam current, mark material and resist thickness. The gain conditions consists of the following parameters:

Gain Parameters Typical Values Description
Signal mode Add signal (first derivative) Specifies the signal mode of the waveform.
Coarse Gain 60 Specifies the coarse gain value.
Middle Gain 0 to 15 Specifies the middle gain value.
Fine Gain 100 Specifies the fine gain value.
Rough Offset 1000 to 2000 Specifies the rough offset value.
Fine Offset 100 Specifies the fine offset value.

A systematic process used to determine the gain conditions is summarized below:

  1. Input the typical values for the conditions. Set Middle Gain to 7 and Rough Offset to 1500.
  2. Execute User Mark Detection and stop the calibration as soon as a waveform is visible
  3. Analyze the waveform and adjust one gain parameter. Then execute the User Mark Detection again. Repeat this until you acquire a U-shaped or inverted U-shaped signal with a base level at 1000 and a maximum at 3000.
    1. Signal is high and flat
      1. Reduce Middle Gain
      2. Reduce Coarse Gain if Middle Gain is 0
    2. Signal is high and not flat
      1. Reduce Rough Offset
    3. Signal is low and flat
      1. Increase Middle Gain
      2. Increase Coarse Gain if Middle Gain is 15
    4. Signal is low and not flat
      1. Increase Rough Offset
    5. Adjust Middle Gain until signal amplitude is approximately 2000
    6. Adjust Rough Offset until signal base level is 1000 and signal maximum is 3000
    7. Adjust Fine Gain and Fine Offset if necessary
  4. Record the gain settings in a notebook since other users can change these settings for their marks

RG Mark Detection

The RG Mark Detection is a subprogram of the Calibration GUI that is used to configure and test the automatic alignment process. The RG Mark Detection must run successfully before one can perform alignment during exposure.

The RG Mark Detection can be accessed through Calibration GUI -> RG Mark Detection tab. The scan conditions for the global marks can be configured in the Global mark detection settings tab and the scan conditions for chip marks can be configured in the Chip mark detection settings tab. The RG Mark Detection uses the same gain conditions as defined by the User Mark Detection settings.

The global mark settings are:

Global Mark Parameters Typical Values Description
Measurement Mode Semi auto Specify the measurement mode. Semi auto mode will put the program into manual mode when it fails to detect the mark automatically.
Material Size 3 Size of the wafer.
AGC Disabled AGC stands for automatic gain control and it will attempt to automatically determine the gain conditions if enabled. AGC is not reliable.
P point mark position Specify the position of the P mark in Expose coordinates.
Q point mark position Specify the position of the Q mark in Expose coordinates.
P point mark offset position Auto update enabled Specify the offset of the P mark. When auto update is enabled, the offset of the P mark will be determined from manually finding the mark.
Scan settings The scan settings are described in the User Mark Detection section. The settings in this tab applies specifically to the global marks.

The chip mark settings are:

Chip Mark Parameters Typical Values Description
Measurement Mode No mark Specify the measurement mode. 1-point mark will use M1 to correct for offset. 3-point mark will use marks M1, M2 and M3 to correct for offset, rotation and scale. No mark will disable chip marks.
AGC Disabled AGC stands for automatic gain control and it will attempt to automatically determine the gain conditions if enabled. AGC is not reliable.
M1 mark position Specify the position of M1 with respect to the mark center.
M2 mark position Specify the position of M2 with respect to the mark center.
M3 mark position Specify the position of M3 with respect to the mark center.
Mark field center position (material coordinate) Specify the mark center offset with respect to the chip center.
Scan settings The scan settings are described int he User Detection section. The settings in this tab applies specifically to the chip marks.

RG Mark Detection is performed after User Mark Detection has been optimized. A typical RG Mark Detection process is:

  1. Locate the P mark and store it as a User Mark
  2. Specify the P and Q mark position in Expose coordinates
  3. Specify the appropriate Scan settings
  4. Select the Semi auto measurement mode
  5. Select No mark for the chip mark measurement mode
  6. Execute RG Mark Detection
    1. The program will try to automatically detect the P mark
      1. If the program fails to automatically detect the mark, it will allow the user to manually find the mark. In this case, use the SEM mode and stage commands to locate and center the mark on the display. Alternatively, move to the User Mark.
      2. Once the P mark is detected it will update the P point mark offset position
    2. The program will automatically detect the Q mark

Exposure Configuration

Global mark detection can be enabled in the Exposure GUI by checking the Global mark box in the Material Information window. Then specify the position of the P mark and Q mark. Specifying the size of the mark will draw the P mark and Q mark on the wafer. This is just a visual indicator, so the JBX-5500FS will not expose those marks.

The global mark settings are:

Global Mark Settings Description
Global Mark Enable/Disable global mark detection
P Position [μm] Specifies the expected X and Y position of the P global mark in Expose coordinate.
Q Position [μm] Specifies the expected X and Y position of the Q global mark in Expose coordinate.
Width [μm] Specify the width of the pair of cross marks to be drawn on the Expose layout.
Length [μm] Specify the length of the pair of cross marks to be drawn on the Expose layout.

The P and Q mark positions should be the same values entered in the RG Mark Detection configuration tab. When the program goes to the P mark position defined in the Expose GUI, it will automatically apply the offset defined in RG Mark Detection.

Chip mark detection can be enabled in the Job Property window in the Exposure GUI. This feature needs to be tested at UHNF.