Jump to content

Specific Process Knowledge/Lithography/Aligners/Aligner: Maskless 01 processing: Difference between revisions

← Older edit
VisualWikitext
Taran (talk | contribs)
DCH-Employees-701, NLAB-Employees-701, NLAB-LabmanagerAllUsers, NLAB-Nlab307-labadviser-users-34891
3,496 edits
Taran (talk | contribs)
DCH-Employees-701, NLAB-Employees-701, NLAB-LabmanagerAllUsers, NLAB-Nlab307-labadviser-users-34891
3,496 edits
 
(21 intermediate revisions by the same user not shown)
Line 123: Line 123:


=Alignment=
=Alignment=
The alignment accuracy of the Aligner: Maskless 01 is a combination of the position accuracy of the stage, the accuracy of the alignment mark detection, and the accuracy of the pattern already on the wafer (first print).
<br/> By measuring the stitching accuracy between two layers printed on the same substrate (without unloading the substrate), we can assess the stage accuracy. By aligning to a pattern previously exposed by the Aligner: Maskless 01, we can assess the mark detection accuracy. And finally, by aligning to a pattern exposed on a mask aligner, we can assess the mask-less aligner's ability to compensate for any scaling and orthogonality errors between the two machines.


The results reported here use printed verniers to assess the misalignment along the two axes at different points on the wafer using an optical microscope. Two different designs were used; a ±5µm vernier and a ±1µm vernier. Both consist of a scale of 4µm lines with 10µm pitch, and a vernier scale to enable subdivision of the 5µm or 1µm scale into tenths, i.e. 0.5µm or 0.1µm. During inspection, observation of the symmetry of neighboring lines enables the observer to read the shifts with ±0.25µm or ±0.05µm accuracy.
The alignment accuracy of the Aligner: Maskless 01 is a combination of the position accuracy of the stage, the accuracy of the alignment mark detection (mostly determined by the offset between camera and stage), and the accuracy of the pattern already on the wafer (first print).
<br/>The measurements are used to calculate the misalignment of the second layer with respect to the first print: The Misalignment [µm] is the median of all measurement points in X or Y; the Translation [µm] is the amount by which the image is shifted; the Rotation [ppm] is the angle by which the image is rotated; and the Run-out [ppm] is the amount of gain in the image. The unit of ppm (parts per million) is used as the rotation and run-out are generally small. A rotation of 1ppm corresponds to an angle of 0.2" (arcseconds) or a shift of 100nm across an entire 4" wafer, while a run-out of 1ppm corresponds to a shift of 50nm at the edge of a 4" wafer compared to the center. For comparison, the pixel size at the wafer surface is 500nm X 500nm, and the address grid size is 50nm.
<br/> By measuring the stitching accuracy between two layers printed on the same substrate (without unloading the substrate), we can assess the stage accuracy. By aligning to a pattern previously exposed by the Aligner: Maskless 01, we can assess the mark detection accuracy. And finally, by aligning to a pattern exposed on a different aligner, we can assess the mask-less aligner's ability to compensate for any scaling and orthogonality errors between the two machines.  
<br/>The deviations (±) given for the results here are calculated as half the range of measurements. If the range is small, the measurement uncertainty is used in stead.
<br/>The samples used for these tests are 100mm Si wafers coated with a 1.5µm layer of the positive tone resist AZ 5214E.
 
The conclusion to the tests are that the stitching accuracy of the Aligner: Maskless 01 is ±0.1µm. The machine-to-self overlay accuracy is ±0.5µm. The machine-to-machine overlay accuracy could not be determined.  


==Important note about correction options==
==Important note about correction options==
Line 151: Line 145:
<span style="color:red">If four marks are used, but scaling and shearing is not applied, ''significant'' misalignment will be observed, even on chips. On a 4" wafer the shift in Y can be several hundred µm.</span>
<span style="color:red">If four marks are used, but scaling and shearing is not applied, ''significant'' misalignment will be observed, even on chips. On a 4" wafer the shift in Y can be several hundred µm.</span>


==Stitching==
==Alignment tests==
In the stitching test, the design consists of ±5µm and ±1µm verniers along the X and Y axis placed in a 3 by 3 matrix covering a 60mm by 60mm area centered on the wafer. The sample is loaded, and the first layer (the linear scales) is printed. Without unloading, the second layer (the vernier scales) is printed on top of the first, and then the sample is developed.
After installation, multiple tests were conducted in order to assess the overlay accuracy of Aligner: Maskless 01. The conclusion to the early tests were that the stage accuracy is ±0.1µm, and the machine-to-self overlay accuracy is ±0.5µm. The machine-to-machine overlay accuracy was not determined (due to the lack of a suitable mask for the mask aligners). In 2019, efforts to establish regular QC of the equipment were started, and the accuracy of the alignment mark detection has been measured regularly since 2020. While both the average and the spread of the alignment errors for the x-axis (measured in 3x3 positions covering a 60x60mm<sup>2</sup> area) has consistently been within the ±1µm specification of the machine, the spread of the alignment errors for the y-axis is typically 3±1µm, despite the average error being in spec, due to negative offsets on the upper half of the wafer and positive offsets on the lower. In 2025, it was decided to investigate this problem further, in order to determine whether a specific alignment protocol could remedy the alignment error, or whether the acceptance limits for the QC would have to be changed.  
 
The result of these tests suggest that when aligning to a pattern exposed using MLA1, only 2 alignment marks on the X-axis should be used. If the first pattern was exposed using a different tool, 4 alignment marks must be used (with all corrections applied), but the alignment accuracy in Y-direction suffers. Most likely, the Y-shift will grow linearly with the distance from the center, so small samples will be less affected, while full wafers will experience shifts in Y that far exceed the ±1µm specification. But in general, larger alignment error in Y must be accepted when aligning to a pattern exposed on a different tool.
 


The results in the table below show that the errors are at or below the measurement uncertainty for the stitching tests using no flat alignment. In other words, the stage is accurate to within 100nm.
In the MLA1-MLA1 alignment tests, the design consists of ±5µm verniers with 0.1µm resolution along the X and Y axis placed in a 3 by 3 matrix covering a 60mm by 60mm area centered on the wafer. The sample is loaded, and the first layer with linear scales is printed (without global angle). Without unloading, the second layer with vernier scales is printed on top of the first, and then the sample is developed. The deviations (±) given for the results here are calculated as half the range of measurements. If the range is small, the measurement uncertainty is used instead.
<br/>When flat alignment is used, a rotation error of ~1.5ppm appears, along with a ~0.3µm misalignment and a -6ppm scaling of the Y axis. This level of accuracy (which corresponds to approximately half a pixel) is what we can expect when the rotation compensation is applied, i.e. when aligning to a previously printed layer.


{|border="1" cellspacing="0" cellpadding="3" style="text-align:left;"  
{|border="1" cellspacing="0" cellpadding="3" style="text-align:center;"  
|-
|-


|-
|-
|-style="background:silver; color:black"
|-style="background:silver; color:black"
!colspan="2" align="center"|
!colspan="2"|MLA1-MLA1
!Misalignment [µm]
!Mark positions
!Translation [µm]
!Rotation [mRad]
!Run-out [ppm]
!Scaling [a.u.]
!Rotation [ppm]
!Shearing [mRad]
!Average error [µm]
!Deviation [µm]
|-
|-


|-
|-
|-style="background:WhiteSmoke; color:black"
|-style="background:WhiteSmoke; color:black"
|rowspan="2" align="center"|No flat alignment
|rowspan="2"|No alignment
(two samples)
(stage accuracy test)
|'''X'''
|'''X'''
| -0.03±0.08
|rowspan="2" align="center"| -
| 0.00±0.05
|rowspan="2" align="center"| -
| -0.97±1.67
| -
| 0.42±1.18
|rowspan="2" align="center"| -
| 0.03
| ±0.05


|-
|-
|-style="background:WhiteSmoke; color:black"
|-style="background:WhiteSmoke; color:black"
|'''Y'''
|'''Y'''
| -0.05±0.05
| -
| -0.07±0.05
| 0.10
| -0.42±1.67
| ±0.225
| 0.00±1.18


|-
|-
|-style="background:WhiteSmoke; color:black"
|-style="background:WhiteSmoke; color:black"
|rowspan="2" align="center"|With flat alignment
|rowspan="2"|Field alignment
(after development,<br>global alignment to 2 marks on X-axis)
|'''X'''
|'''X'''
| 0.05±0.10
|rowspan="2" align="center"|1: -37500; 0<br>2: 37500; 0
| 0.05±0.05
|rowspan="2" align="center"|8.414
| 0.56±1.67
| -
| -1.67±1.18
|rowspan="2" align="center"| -
| 0,04
| ±0,05


|-
|-
|-style="background:WhiteSmoke; color:black"
|-style="background:WhiteSmoke; color:black"
|'''Y'''
|'''Y'''
| -0.25±0.35
| -
| -0.29±0.16
| -0,22
| -6.39±1.67
| ±0,075
| -1.39±1.18


|}
|-
|-style="background:WhiteSmoke; color:black"
|rowspan="2"|4 alignment marks
(like QC)
|'''X'''
|rowspan="2" align="center"|1: -37500; 0<br>2: 37500; 0<br>3: 0; 35000<br>4: 0; -35000
|rowspan="2" align="center"|0.000
| 1.000001
|rowspan="2" align="center"|0.0002
| 0.06
| ±0.05
 
|-
|-style="background:WhiteSmoke; color:black"
|'''Y'''
| 0.999960
| -0.31
| <span style="color:red">±1.30</span>


==Overlay==
|-
In the overlay test, two alignment accuracies are assessed: The machine-to-self overlay accuracy (MLA-MLA), and the machine-to-machine (MA6-MLA) overlay accuracy. Because alignment is possible using two marks or four marks, both are tested in each case. Exposing the first print with or without flat alignment was also tested, but no significant effect was observed.
|-style="background:WhiteSmoke; color:black"
|rowspan="2"|2 alignment marks
(on X-axis)
|'''X'''
|rowspan="2" align="center"|1: -37500; 0<br>2: 37500; 0
|rowspan="2" align="center"|0.003
| -
|rowspan="2" align="center"| -
| 0.01
| ±0.15


In the MLA-MLA overlay test, the design is the same as for stitching; ±5µm and ±1µm verniers along the X and Y axis placed in a 3 by 3 matrix covering a 60mm by 60mm area centered on the wafer. The sample is loaded, and the first layer (the linear scales) is printed. The sample is unloaded and developed. The second layer (the vernier scales) is aligned to marks contained in the first layer, and then the sample is developed again. The alignment marks used for 2 mark alignment are placed 60mm apart on the X axis, while the marks used for 4 mark alignment are placed at the corners of a 60mm by 30mm rectangle.
|-
|-style="background:WhiteSmoke; color:black"
|'''Y'''
| -
| -0.26
| ±0.20


The results in the table below show an alignment error of -2µm in X. They also show that including the scaling and shearing of the axes calculated by the machine from alignment to four marks in the exposure only seems to introduce a gain error in Y. However, if the wafer has significant bow or other distortions due to processing of the first layer, scaling and/or shearing correction may be necessary.
|-
<br/>Since the alignment error in X seems so consistent, the machine was calibrated by correcting the X-axis beam off-set 2µm in the machine configuration files. As seen below, this removed the alignment error, and suggests that the machine-to-self overlay accuracy of the Aligner: Maskless 01 is ±0.5µm or better.
|-style="background:WhiteSmoke; color:black"
|rowspan="2"|2 alignment marks
(on Y-axis)
|'''X'''
|rowspan="2" align="center"|1: 0; 35000<br>2: 0; -35000
|rowspan="2" align="center"|0.000
| -
|rowspan="2" align="center"| -
| 0,11
| ±0,275


{|border="1" cellspacing="0" cellpadding="3" style="text-align:center;"
|-
|-
|-style="background:WhiteSmoke; color:black"
|'''Y'''
| -
| <span style="color:red">-2,69</span>
| ±0,2


|-
|-
|-style="background:silver; color:black"
|-style="background:WhiteSmoke; color:black"
!colspan="2"|MLA-MLA
|rowspan="2"|2 alignment marks
!Scaling [ppm]
(top half of wafer)
!Shearing [mRad]
|'''X'''
!Misalignment [µm]
|rowspan="2" align="center"|1: -30000; 30000<br>2: 30000; 30000
!Translation [µm]
|rowspan="2" align="center"|-0.003
!Run-out [ppm]
| -
!Rotation [ppm]
|rowspan="2" align="center"| -
| 0,08
| ±0,1
 
|-
|-
|-style="background:WhiteSmoke; color:black"
|'''Y'''
| -
| <span style="color:red">-2,16</span>
| ±0,15


|-
|-
|-style="background:WhiteSmoke; color:black"
|-style="background:WhiteSmoke; color:black"
|rowspan="2"|2 alignment marks
|rowspan="2"|2 alignment marks
(three samples)
(bottom half of wafer)
|'''X'''
|'''X'''
|NA
|rowspan="2" align="center"|1: -30000; -30000<br>2: 30000; -30000
|rowspan="2"|NA
|rowspan="2" align="center"|0.005
| -2.08±0.25
| -
| -2.07±0.25
|rowspan="2" align="center"| -
| -2.31±8.33
| -0,26
| 1.02±1.67
| ±0,25


|-
|-
|-style="background:WhiteSmoke; color:black"
|-style="background:WhiteSmoke; color:black"
|'''Y'''
|'''Y'''
| NA
| -
| -0.33±0.18
| -0,01
| -0.34±0.19
| ±0,15
| 1.20±1.67
| -0.46±5.89


|-
|-
|-style="background:WhiteSmoke; color:black"
|-style="background:WhiteSmoke; color:black"
|rowspan="2"|4 alignment marks
|rowspan="2"|2 alignment marks
(three samples)
(on Y-axis, starting at bottom)
|'''X'''
|'''X'''
| 1±2
|rowspan="2" align="center"|1: 0; -35000<br>2: 0; 35000
|rowspan="2" align="center"|0.002±0.001
|rowspan="2" align="center"|0.004
| -2.08±0.25
| -
| -2.08±0.25
|rowspan="2" align="center"| -
| -3.70±8.33
| 0,02
| 0.56±1.67
| ±0,1


|-
|-
|-style="background:WhiteSmoke; color:black"
|-style="background:WhiteSmoke; color:black"
|'''Y'''
|'''Y'''
| -12±1
| -
| -0.22±0.35
| 0,02
| -0.21±0.08
| ±0,2
| -6.30±1.67
| -0.93±5.89


|-
|-
|-style="background:WhiteSmoke; color:black"
|-style="background:WhiteSmoke; color:black"
|rowspan="2"|2 alignment marks
|rowspan="2"|5 alignment marks
after calibration
(0;0 + QC marks)
(three samples)
|'''X'''
|'''X'''
|NA
|rowspan="2" align="center"|1: 0; 0<br>2: -37500; 0<br>3: 37500; 0<br>4: 0; 35000<br>5: 0; -35000
|rowspan="2"|NA
|rowspan="2" align="center"|-0.001
| -0.22±0.25
| 1.000003
| -0.19±0.19
|rowspan="2" align="center"|0.0000
| 1.85±2.08
| 0.16
| -0.09±2.92
| ±0.05


|-
|-
|-style="background:WhiteSmoke; color:black"
|-style="background:WhiteSmoke; color:black"
|'''Y'''
|'''Y'''
| NA
| 0.999963
| -0.28±0.38
| -1.11
| -0.28±0.13
| <span style="color:red">±1.175</span>
| 4.35±5.97
| -0.09±2.64


|}
|}


The stage alignment test shows a relatively good repeatability of the stage. The X-axis is clearly more accurate than the Y-axis, as evidenced by the relatively large deviation on the Y-axis values. The raw data shows that all positional errors are within ±0.3µm, mainly due to the Y-shifts on the Y-axis being ~0.2µm larger than in the other positions.
<br>The field alignment test shows much tighter values, and the errors represent the true error on the camera offset, i.e. the shift that can be corrected in the machine configuration.
<br>The alignment test with 4 alignment marks mimics the shift from the field alignment test, but the deviation on the Y-axis is very large, probably due to the surprising -40ppm scaling measured by the alignment routine. Keep in mind that the wafer has not been unloaded between the two exposures. Something is going on with the Y-axis.
<br>Aligning with 2 marks on the X-axis seems to fix this problem, and shows an average error similar to the camera offset, with a tight distribution across the wafer. However, aligning using 2 marks on the Y-axis introduces a large shift in Y. This shift is repeated if 2 alignment marks along the X-axis on the top half of the wafer is used, but it is fixed if 2 marks along the X-axis on the bottom half are used, or if 2 marks on the Y-axis is used with the first mark on the bottom half of the wafer. Again, there seems to be something strange going on with the Y-axis.
<br>In an attempt to fix the large deviation on the Y-axis when using 4 alignment marks, a test was made adding an alignment mark in 0;0 as the first mark during alignment. This did not have any beneficial effect, as the deviation on the Y-axis values is similar to the deviation from the 4 mark test.


In the MA6-MLA overlay test, a mask from an existing design (GreenBelt METAL v2, dark field) is reused to print the first layer using the Aligner: MA6-2, and the sample is developed. The second layer is aligned and printed in Aligner: Maskless 01, using a design containing four sets of ±5 vernier scales located at the corners of a ~90mm by ~3mm rectangle, before being developed again. For 2 mark alignment, the original alignment marks at X±~43mm are used. Since only two alignment marks exist in the original design, four corners (positioned in a ~60mm by ~60mm square pattern) are used for manual alignment in the 4 mark test. Using corners rather than crosses reduces the accuracy of the alignment, as the positions of corners are subject to shifts due to bias in the first print lithography.


The results in the table below reproduce the alignment error in X seen in the machine-to-self overlay tests (should be better now that the machine has been calibrated). Due to the close proximity of the verniers in Y, run-out and rotation have not been calculated for that axis. But the X axis shows some run-out, consistent with alignment between two stages (the mask writer and the maskless aligner) without scaling compensation. The fact that using scaling and shearing compensation doesn't remove this gain, and seems to add a significant rotation error, is probably more due to the alignment "marks" used than to the machine itself. More tests using a purpose-designed mask would have to be conducted in order to estimate the machine-to-machine overlay accuracy of the Aligner: Maskless 01.
In the MLA3-MLA1 alignment tests, the design consists of ±5µm verniers with 0.25µm resolution along the X and Y axis placed in a 3 by 3 matrix covering a 60mm by 60mm area centered on the wafer. The first layer with linear scales was printed in MLA3 as QC test wafers a long time ago and subsequently patterned using lift-off of gold. These wafers are coated with resist, the second layer with vernier scales is printed in MLA1, and then the sample is developed. The deviations (±) given for the results here are calculated as half the range of measurements. If the range is small, the measurement uncertainty is used instead.


{|border="1" cellspacing="0" cellpadding="3" style="text-align:center;"  
{|border="1" cellspacing="0" cellpadding="3" style="text-align:center;"  
Line 304: Line 354:
|-
|-
|-style="background:silver; color:black"
|-style="background:silver; color:black"
!colspan="2"|MA6-MLA
!colspan="2"|MLA3-MLA1
!Scaling [ppm]
!Mark positions
!Rotation [mRad]
!Scaling [a.u.]
!Shearing [mRad]
!Shearing [mRad]
!Misalignment [µm]
!Average error [µm]
!Translation [µm]
!Deviation [µm]
!Run-out [ppm]
|-
!Rotation [ppm]
 
|-
|-style="background:WhiteSmoke; color:black"
|rowspan="2"|4 alignment marks
(old QC wafer)
|'''X'''
|rowspan="2" align="center"|1: -35000; -25000<br>2: 35000; -25000<br>3: -35000; 25000<br>4: 35000; 25000
|rowspan="2" align="center"|7.687
| 1.000040
|rowspan="2" align="center"|-0.108
| -0,03
| ±0,625
 
|-
|-
|-style="background:WhiteSmoke; color:black"
|'''Y'''
| 0.999979
| -1.06
| <span style="color:red">±1,375</span>


|-
|-
|-style="background:WhiteSmoke; color:black"
|-style="background:WhiteSmoke; color:black"
|rowspan="2"|2 alignment marks
|rowspan="2"|2 alignment marks
(old QC wafer)
|'''X'''
|'''X'''
|NA
|rowspan="2" align="center"|1: -30000; 0<br>2: 30000; 0
|rowspan="2"|NA
|rowspan="2" align="center"|7.951
| -1.50±0.75
| -
| -1.50±0.25
|rowspan="2" align="center"| -
| 6.94±5.56
| -0.94
| 2.78±3.89
| <span style="color:red">±7,375</span>


|-
|-
|-style="background:WhiteSmoke; color:black"
|-style="background:WhiteSmoke; color:black"
|'''Y'''
|'''Y'''
| NA
| -
| 0.00±0.25
| -0,72
| 0.00±0.25
| ±0.5
| NA
 
| NA
|-
|-style="background:WhiteSmoke; color:black"
|rowspan="2"|5 alignment marks
(0;0 + old QC marks)
|'''X'''
|rowspan="2" align="center"|1: 0; 0<br>2: -35000; -25000<br>3: 35000; -25000<br>4: -35000; 25000<br>5: 35000; 25000
|rowspan="2" align="center"|9.462
| 1.000040
|rowspan="2" align="center"|-0.108
| -0.22
| ±0.25


|-
|-
|-style="background:WhiteSmoke; color:black"
|-style="background:WhiteSmoke; color:black"
|rowspan="2"|4 alignment marks
|'''Y'''
| 0.999979
| -0.94
| <span style="color:red">±1.5</span>
 
|-
|-style="background:WhiteSmoke; color:black"
|rowspan="2"|MLA1-MLA3
(4 alignment marks)
|'''X'''
|'''X'''
| -8
|rowspan="2" align="center"|1: -37500; 0<br>2: 37500; 0<br>3: 0; 35000<br>4: 0; -35000
|rowspan="2" align="center"|0.005
|rowspan="2" align="center"|-16.298
| -1.75±0.63
| 0.999978
| -1.69±0.25
|rowspan="2" align="center"|0.106
| 6.25±5.56
| -0.09
| -12.5±3.89
| ±0.15


|-
|-
|-style="background:WhiteSmoke; color:black"
|-style="background:WhiteSmoke; color:black"
|'''Y'''
|'''Y'''
| 13
| 1.000004
| 0.25±0.63
| 1.02
| 0.19±0.56
| ±0.25
| NA
| NA


|}
|}
The alignment test with 4 alignment marks shows a +40ppm scaling on the X-axis, as well as a 0.1mRad shearing of the axes. The result is a decent alignment in X, but a shift in Y as well as a relatively large deviation. The raw data shows the deviation in Y is due to a -40ppm scaling along the Y-axis, as seen in the MLA1-MLA1 test with 4 marks, suggesting that the scaling in Y is consistently overestimated.
<br>Aligning using only 2 marks yields acceptable shifts in the center of the wafer, but very large shifts in X towards the edges, as evidenced by the 7.4µm deviation in X. The raw data suggests that this deviation is mainly due to a 0.2mRad tilt in the Y-axis, which corresponds well with the 0.1mRad shearing measured using 4 marks. There is also a (-)40ppm scaling along the X-axis, again similar to what was measured during 4 mark alignment. Even a 5mm chip would be affected by the 0.2mRad tilt, so clearly 4 mark alignment is needed when aligning to a pattern that was not exposed using MLA1.
<br>Attempting to fix the shift in Y when using 4 alignment marks by adding 0;0 as the first mark unfortunately makes no difference. However, when MLA3 is used to align to a pattern printed using MLA1, the resulting spread of alignment errors is quite small, suggesting that MAL3 is somehow better at compensating for the differences between the two machines than MLA1.


=Optimal use of the maskless aligner=
=Optimal use of the maskless aligner=
Retrieved from "https://labadviser.nanolab.dtu.dk//index.php?title=Specific_Process_Knowledge/Lithography/Aligners/Aligner:_Maskless_01_processing"