LEO SEM Lithography Guide

Table of Contents

• Chip Prep & Mounting
• LEO 440 SEM Lithography Checklist
• Deposit Gold
• Liftoff

Chip Prep & Mounting

Preliminary Cleaning

• Scintillation vials or small beakers
• Three pipets
• One set of tweezers
• Enough copper backings and SEM mounts for your samples
• Electronic grade (or better) 2-propanol.
• Acetone (straight from the source bottle is best)

PMMA Deposition

Mounting and Final Prep for SEM

1. Place a clean SEM mount in a vise
2. Affix copper/chip assembly to SEM mount with carbon paint by applying a thin, uniform layer to the SEM mount, then press down very firmly to seat the copper well against the mount surface (when it stops slipping around). The copper and mount should be concentric.
3. Now you will need to place landmarks for position and focus. For this you will need silver paint that has precipitated in the bottle, do not shake it up.
4. Break a thin wooden rod, e.g. a cotton swab stick to get a nice sharp tip. Using this tip, place two tiny silver paint dots on opposite sides of the chip, trying to center them along the edges. The dots go on the edges perpendicular to the edges aligned on the edge of the copper, as in the image on the right.
5. Dry (10 minutes under 75W incandescent lamp works fine).

LEO 440 SEM Lithography Checklist

Assumptions & Conventions

• This process assumes you have a sample with silver paint dots centered on the outer edges of the sample to use as landmarks for positioning and focus adjustment. These paint regions will also be used for adjusting wobble and astigmatism in this process.
• The "Gold standard" is actually a sample with silver paint on it
• The mouse button functions can change readily, especially when you are adjusting values other than magnification and focus. Be very careful to click on the magnification icon when you are finished adjusting a parameter with the mouse to return it to magnification mode as a habit. Otherwise, you may risk causing serious harm to your sample and/or the machine.
• The mouse has three buttons, which will be referred to as left, middle, and right.
• The beam blanker is a box near the middle of the actual SEM itself. You will only turn it on or off throughout this procedure
• Jarring the table on which the SEM optics sit can reduce imaging/lithography quality

Sample insertion

Warning: The time to re-acquire a good vacuum depends heavily on how long the sample chamber is open. Try to minimize this time so as to keep the chamber clean and insure a timely pump down.

Note: The SEM will normally be in standby mode.  The beam is on in this mode, but at a lower than normal power.  You do not need to turn off the beam to successfully vent the system and change out samples.

1. Insure stage is in home position of x = 50 mm and y = 50 mm (check display on joystick or the onscreen status window). If it is not, then go to Stage/Vac® Stage move® right mouse button on window® Record ® Labels: HOME (a single click will do).
2. Venting

2.1. Establish N₂ flow (Nitrogen of non-specified purity, passing through a dessicator column)

2.1.1. Using canned air, blow out connectors, and connect hose to tank
2.1.2. Adjust flow to two to three lbs.

2.2. Gently close the gate between the sample chamber and the column chamber (swivel the slim metal rod). This isolates the column chamber from the sample chamber. An incorrect position prompt should appear, notifying you that the gate is closed, but you want it that way, so don’t worry about it and just click OK.
2.3. Unlatch the sample chamber door (silver latch on back right panel), otherwise you can over pressurize system.  This could lead to death and dismemberment which can distract you from research.
2.4. Stage/Vac® Specimen Change: Vent
2.5. Monitor column chamber pressure to insure it is not also venting (if so, then probably the gate it compromised and you should stop venting and contact the Rimberg Lab.
2.6. When venting is complete, there should be an audible hiss of escaping Nitrogen from the edge of the sample chamber door, which will now be slightly ajar.

3. Using tweezers, carefully place sample in mount align with axis of hexagonal sample mount holder and the gold standard. It is very easy to break the carbon paint seal and delay the process by fifteen or more minutes while you remount the sample. There is a set screw with which to secure your sample, but be careful as it will likely induce a rotation of your sample as you tighten it.

Load State

35kvlab6.app – sets up for run at 35 kV; used for fine features.
20kvlab6.app – sets up for run at 20 kV; used for larger features.
Lab6_sta.app – this is the "standby" state; you will normally load this state when you are finished with the SEM.

Saturation and Gun Alignment

Initial Aperture Change

20.1. Stage/Vac® Stage move® right mouse button on window® Record ® Labels: gold_standard_7mm
20.2. Beam® Aperture
20.3. Note the x and y values of the micrometers for your aperture:
• 10 micron: 6.01 mm, -0.31 mm (Note: this would correspond to dialing to zero and then rotating it to "19" on the micrometer).
• 20 micron: 5.99 mm, 11.72 mm (Note: the mark for the "10" is actually centered under the "0" so be careful that you go far enough; the mark is larger than the other, intermediate marks).
• 50 micron: 5.99 mm, 5.68 mm
20.4. Select the desired aperture, click "OK", and close the window. Note the following guidelines:
1. 10 microns for ~10 pA
2. 20 microns for ~100-1000 pA
3. 50 microns for ~3000pA)
20.5. Set the probe current to the lowest setting (e.g. 137 pA for 20 microns).
20.6. Manually adjust the micrometers to the corresponding coordinates to get the desired aperture.
20.7 Go to about 1 k to adjust the wobble:
20.7.1. Select the icon for wobble (the icon looks like a probe at 45 degrees above a red sine wave pattern).
20.7.2. Manually adjust the aperture via the two micrometers to decrease the wobble. Just tweak the micrometers until the image in the middle box stops moving around (it may pulse, but you want to stop it from translating). You can increase the magnification and further adjust the wobble, but you’ll do this soon enough, so don’t bother.

Initial Adjustment of Probe Current

Note: Disconnect ammeter and reconnect stage wire as soon as you are done as this configuration has additonal safeties against crashing the probe.  Because of this, you should generally keep the extrenal ammeter unconnected except when taking a reading.

Warning: be sure to select the magnification icon to reset the mouse controls back to magnification and focus.

29. Scanning® Normal.
30. Check your watch, you can not expose the sample until one hour from now. While we are waiting, we will continue to set up the system (read on).

Wobble, Stigmatism, and Focus, the three Furies* J

Now we will adjust the wobble and stigmatism in the system to give the best results when exposing our sample. The finer your desired features are, the more crucial this step is. However, if you want to generate large features, rough adjustment may work (ask someone that knows). If in doubt, be meticulous as it certainly won’t hurt the sample if you tune the SEM extremely well.

Note: the speed of translation will decrease with higher magnifications.

31. Stage/Vac® Stage move® right mouse button on window® Record ® Labels: gold_standard_sample. This will move you to the "gold standard" position (which is actually silver). Normally, this will mean your sample is down and to the right.

Note: To avoid unwanted exposure of the sample, you need to avoid scanning the region where we you plan to have the pattern. To do so, we will work at the edges of the sample and also keep it at the edges of the screen. In preparation, you may want to move or minimize the Status window to facilitate this.

32. Move to the sample and line up the left edge of the sample on the far right edge of the screen, being careful to keep the chip edge at the edge of the display, i.e. do not scan across the chip as you might expose the region of interest and compromise the sample.
33. Keeping the sample at the edge of the display, find the silver paint region.
34. We want to keep the sample at a working distance such that it is in focus when the focus is at ~7 mm:

34.1. Focus on the paint.
34.2. Double click on focus to see what the current focal length is and make note of it and the current z value the stage is at.
34.3. Calculate the change in stage z.  Realize that the focal length is the absolute magnitude of the distance from the lens, so if the focal length is less than 7 mm, you will want to decrease the stage z to lower the stage, thus moving the stage farther away from the lens.  Also, do not adjust it by more than a 1.0 mm change at a time.  Example: if focal length is 5.8 mm and the stage z is 24.5 mm, then you will want to decrease the stage z by 1.0 mm to 23.5 mm and continue through this instruction loop before adjusting farther.
34.4. Change the stage z by double-clicking on stage goto z and typing in the new z value desired.  Check your numbers and be very careful.  The stage z should be in the 23 mm to 25 mm range usually.  If your value is near or above 30 mm, get someone to help you.
34.5. Immediately after entering the new z value, click on the magnifying glass icon to restore the mouse controls to magnification and focus.  Otherwise, a quick mouse stroke will crash the sample into the probe.
34.6. Now, double-click on focus and type in the focal length you expect to be able to image with now.  If you do not see the sample, you may have gone in the wrong direction.  If you went in the wrong direction, then re-enter the starting values for stage z and focal length, and then start over again with step
34.7.  If the focal length is approximately equal to 7 mm now, then continue on, otherwise repeat 34.1- 34.6 until the focal length is set properly.

35. If needed (and normally it should not be), adjust the rotation.
36. Adjust the focus and increase the magnification to 25 k. The closer to the sample center the piece is, the more like the central region of interest it will be.
37. Select the icon for wobble (the icon looks like a probe at 45 degrees above a red sine wave pattern).
38. Manually adjust the aperture via the two micrometers to decrease the wobble. Simply enough, just tweak the micrometers until the image in the middle box stops moving around (it may pulse, but you want to stop it from translating). It usually will not be necessary, but you can increase the magnification and then further adjust the wobble (to get it more precise, but it should not be necessary normally).
Increase the magnification to 100 k, centering on a region of the silver paint where two edges are very close together, e.g.:

39. Select the icon for the stigmatism (a red arrow pointing from a yellow ellipse to a yellow circle) and adjust until the edges of the gap are sharp.

Note: Another method is to find a round piece of paint and vary the focus. If the piece does not expand uniformly, then the beam has a stigmatism. If it evenly expands, there is no stigmatism. Adjust as needed. The gap is the more sensitive method.

OK, now we want to pick some focus points to use as landmarks.  We do not want to prematurely expose our resist, thus destroying it.  So, we want to have some way to extrapolate the proper focus for the points at which we wish to "draw" our pattern.  I simply take two points along the x-axis of the sample, centered on the inner sides (the edges near the center of the sample) of the silver paint spots I have created.  Then I select my point(s) along this line.  I usually try to keep a good 250 microns or so between patterns, but that is an overly cautious margin for most purposes.

Realize that the farther you stray from this line, the less appropriate your extrapolated focus value(s) will be.  This has some consequences:

• patterns that are long along their y-axis are progressively more likely to have problems with fine features
• an array of points spaced along the y-axis as well, say a 2 x 2 matrix of points, really calls for extrapolation of the shift in focus for changes in y as well, so focus points should be taken at oppposite corners of the piece.

40. Pick a piece of silver that is well-centered w.r.t. the edge of the sample and farthest from the edge (towards the center).  The stage y value accuracy is greater than what is displayed, so I record the value displayed and type it back into the stage goto y value so that I can reproduce the same y value with my second focus point.  Save it as Focus1 by choosing Stage/Vac® Stage move® right mouse button on window® Record ® Labels® Add: Focus1.  Don’t worry about overwriting the old Focus1.  Record the position. The only reasons I see for  using the inner edge are that the spots are more isolated and the lesser concern that the outer edge may have more random junk that might affect charging and also create a more rugged surface for focusing on.
41. Being careful to keep the sample at the edges of the display, move to the other side of the sample and repeat step #41 for a second, focus point, saving it as Focus2. Personally, I make note of the width of my chip before starting and then I blank the beam, type in a stage goto x that is safely off to the side of the chip, unblank the beam and move over to find a paint spot.  Again, don’t worry about overwriting the old Focus2. Record the position. Realize that the approximation of focus will be more accurate if you choose focus points at the same y value.
42. Now you have two points
Go back to the faraday cup, Stage/Vac® Stage move® right mouse button on window® Record ® Labels: faraday_cup, and wait until the sixty minutes after the probe current adjustment have expired (from step 29).

Wait for the sixty minutes to expire

Pre-exposure Steps

43. Clear the lenses by hitting the F2 key a few times (this takes a few seconds, so wait until the machine is ready before hitting F2 again).

44. Examine the gun’s shift and tilt again (open Gun Alignment)

44.1. Select Emission and Gun tilt:
44.2. Center the patch of area in the crosshairs
44.3. Turn off Optbeam conjugate (in Status window)
44.4. Select Optlbeam Mode Depth
44.5. Select Gun Shift and then center the oval in the cross hairs.
44.6. Re-center the oval if necessary.
44.7. Unselect Optlbeam Mode Depth and then select Gun Tilt.
44.8. Turn on Optbeam conjugate (in Status window).
44.9. Click "normal" and close the window.

45. You should still be over the Faraday cup. You now need to check the probe current once again:

45.1 Decrease magnification to 200-300 and you should see the hole in the cup off to the bottom left
45.2 Center the hole in the screen and increase to magnification of about 9000 (until hole ~fills the screen but you can still see the edges of the hole).
45.3 Scanning® Spot. A cross-hair should appear.
45.4 Drag the cross-hair to the center of the hole.
45.5 Check on the external ammeter that the current read is the current set in the MRF files for your pattern.
45.6 Adjust the IProbe until the external ammeter gives the desired reading.
45.7 Scanning® Normal.

54.1 Select line average and increase to ~52.
54.2 Select Scan – and increase cycle time to max (1.5 days).
54.3 Close Image window

Changing the aperture between patterns

63. Stage/Vac® Stage move® right mouse button on window® Record ® Labels: gold_standard_7mm
64. Set the noise reduction back to normal values:

64.1 Select Frame Average
64.2 Select Scan – and decrease cycle time to 931 ms.

• 10 micron: 6.01 mm, -0.31 mm (Note: this would correspond to dialing to zero and then rotating it to "19" on the micrometer).
• 20 micron: 5.99 mm, 11.72 mm (Note: the mark for the "10" is actually centered under the "0" so be careful that you go far enough; the mark is larger than the other, intermediate marks).
• 50 micron: 5.99 mm, 5.68 mm

• 10 microns for ~10 pA
• 20 microns for ~500-1000 pA
• 50 microns for ~3000pA)

72.1 Select the icon for wobble (the icon looks like a probe at 45 degrees above a red sine wave pattern).
72.2 Manually adjust the aperture via the two micrometers to decrease the wobble. Simply enough, just tweak the micrometers until the image in the middle box stops moving around (it may pulse, but you want to stop it from translating). It usually will not be necessary, but you can increase the magnification and then further adjust the wobble (to get it more precise, but it should not be necessary normally).

74.1 Insure that the Optbeam conjugate is on.
74.2 Decrease magnification to 200-300 and you should see the hole in the cup off to the bottom left
74.3 Center the hole in the screen and increase to magnification of about 9000 (until hole ~fills the screen but you can still see the edges of the hole).
74.4 Scanning® Spot. A cross-hair should appear.
74.5 Drag the cross-hair to the center of the hole.
74.6 Check on the external ammeter that the current read is the current set in the MRF files for your pattern.
74.7 Adjust the IProbe until the external ammeter gives the desired reading.
74.8 Turn on the beam blanker before turning the scan back to normal.
74.9 Scanning® Normal.
74.10 Double check that the beam is blanked.

Running the Next Pattern

75. Stage/Vac® Stage move® right mouse button on window® Record ® Labels: Focus1. This will reset the stage parameters to the values they were before you made adjustments above.
76. Move to the center with the averaged coordinates by manually entering the x and y values for Stage goto x and Stage goto y in the Status window. Note that because we went to Focus1 first, then moved to the center, the value for focus should still be correct.
77. Image® Noise reduction.

77.1 Select line average and increase to ~52.
77.2 Select Scan – and increase cycle time to max (1.5 days).

Repeat steps as needed until all patterns are run.

Wrapping Up

86. Stage/Vac® Stage move® right mouse button on window® Record ® Labels: gold_standard_7mm
87. Set the noise reduction back to normal values:

87.1 Select Frame Average
87.2 Select Scan – and decrease cycle time to 931 ms.

94.1 Select the icon for wobble (the icon looks like a probe at 45 degrees above a red sine wave pattern).
94.2 Manually adjust the aperture via the two micrometers to decrease the wobble. Simply enough, just tweak the micrometers until the image in the middle box stops moving around (it may pulse, but you want to stop it from translating). It usually will not be necessary, but you can increase the magnification and then further adjust the wobble (to get it more precise, but it should not be necessary normally).

Removing Your Sample

1.1 Establish N₂ flow (Nitrogen of non-specified purity, passing through a dessicator column).
1.1.1 Using canned air, blow out connectors, and connect hose to tank
1.1.2 Adjust flow to two to three lbs.
1.2 Gently close the gate between the sample chamber and the column chamber (swivel the slim metal rod). This isolates the column chamber from the sample chamber. An incorrect position prompt should appear, notifying you that the gate is closed, but you want it that way, so don’t worry about it and just click OK.
1.3 Unlatch the sample chamber door (silver latch on back right panel), otherwise you can over pressurize system
1.4 Stage/Vac® Specimen Change: Vent
1.5 Monitor column chamber pressure to insure it is not also venting (if so, then probably the gate it compromised)
1.6 When venting is complete, there should be an audible hiss of escaping Nitrogen from the edge of the sample chamber door.
1.7 Warning: unplug the N₂ after venting, or else oil may back flow from the pump in the case of a power failure

2. Loosen the setscrew holding your sample in place. Using tweezers, carefully remove your.
3. Close the sample chamber, latch it, and then hit Change Specimen: Pump to start the pump down.

DO NOT LEAVE YET

4. Gently open the gate and click on OK in the prompt window. Now the system needs to pump down to 3.0 x 10^-7 Torr (3.5 might work if you are only going to use 20 kV), and this will take a few minutes.

You are finished!

Other stuff

• The Status Window, What’s in it?
• FilI the filament current
• EHT the accelerating voltage
• Beam Current the total current from the filament
• Iprobe current which hits the sample (we have to set the right value using the faraday cup)
• Collector Bias bias on the lens which is used to collect the secondary electrons, which is always set at 250 V.
• Cycle Time time used to make a whole scan across the sample.
• Pixel average for each pixel SEM will sample that point a number of times and average the view (favorite is 4).
• External Scan Control used to give control to an external source (used to give control to the NPGS in our case)
• Optbeam automatically optimizes currents in the lens so it should always be on.
• Optbeam conjugate should be on during lithography and adjusting the gun tilt, off only during gun shift adjustment
• Backlash takes care of backlash so it should always be on.

Deposit Gold

Starting point: Sample fresh out of the SEM. A pattern has been generated on PMMA, dosing/destroying selected regions.

A. Separate the chip from the copper backing.
B. Develop pattern (remove the destroyed PMMA, leaving the surface below exposed):

1. Establish a good, secure tweezer grip.
2. Agitate in developer solution of 1:3 ratio of 4-methyl-2-pentone (a.k.a. "MIBK") and isopropanol solution x 60 seconds
3. Agitate in isopropanol x 30 seconds
4. Rinse well with squirt bottle of isopropanol (squirt about 25 mL of isopropanol across the sample, focussing primarily on the central region of interest)

1. Initial configuration:
• Rough pump closed to bell jar
• Gate (bell jar to diffusion pump) closed
• Rough pump open to diffusion pump
2. Vent with N₂ @ ~5 psi
3. Immediately shut off N₂ after seal broken on vacuum.
4. Don powderless gloves
5. Raise jar. Note: the longer the jar is open, the more contaminated (mostly water vapor) it becomes and thus the longer it will take to pump down
6. Insert sample in mount or rest on plate, as appropriate
7. Remove shields from electrode sources. They should fit snugly around the posts nearby, do not lay them down outside of the vacuum chamber as they will become contaminated
8. Check to insure Chromium source is adequate (look at a new one to get an idea, they should be changed about every 3-4 times), change if necessary
9. Check gold "boat" for patency, replace if cracked, damaged
10. Insert gold in boat
11. Replace shields on sources. Make sure they do not touch anything but the sources, including each other and the base plate of the chamber.
12. Close jar
13. Close rough pump to diffusion pump
14. Slightly and slowly open rough pump to bell jar only and hold the bell jar in place with both hands while a seal forms (until ~600 Torr). Then open rough pump the rest of the way
15. Turn on liquid N₂ to cold trap and shut it off when it spills out of the back (~ a minute or two)
16. Pump down to about 80 milliTorr (do not leave pumping overnight or you can get oil back flow into chamber)
17. Close rough pump to bell jar
18. Open rough pump to diffusion pump
19. Open gate very slowly
20. Turn on ion gauge when pressure ~ 2-3 mTorr. As of December 1998, the thermocouple inside the ion gauge is broken, so you will have to rely on the separate thermocouple, which will bottom out and no longer change at 2 mTorr.
21. Wait until pressure ~ 5 x 10^-6 (will only take about five or ten more minutes)
22. Position shutter between samples and source.
23. Turn on Infinicon, set for density and z-ratio of Chromium.
24. PROG toggles between parameters and film
25. The "3" key, also labeled "ZERO" will zero the thickness.  Note: tooling factor is 100 for the broad round panel with the wide slot in it, but 47.6 for the sample mount higher up
26. Turn the knob for the "B" current source all the way CCW (zero position)
27. Move the shutter into position between the sample and the sources.
28. Warning: You will typically only be depositing for about ten seconds or less when putting down the Chromium, so you need to be on your toes when you start it.
29. Turn on current source (dial "B" for Chromium) and adjust until ~0.5 Å/sec. (approximately 2 full rotations)
30. Move the shutter smoothly and quickly out from between the samples and the source while pressing the "3" key, also labeled "ZERO".
31. Deposit ~20 Å. Note: the display is in kiloAngstroms. Also, recall that 10 Angstroms = 1 nanometer.
32. Close the shutter when the desired amount is reached and shut off the source, i.e. switch off "B" and return knob to zero position (all the way CCW).
33. Adjust Infinicon, set for density and z-ratio of Gold as with Chromium (see 3.17.1-3.17.3)
34. Turn the knob for the "A" current source all the way CCW (zero position)
35. Turn on current source (dial "A" for Gold) until ~1-3 Å/sec. Note: will spike at first.
36. Move the shutter smoothly and quickly out from between the samples and the source while pressing the "3" key, also labeled "ZERO".
37. Deposit desired amount of gold (typically 100 to 500 Å).
38. Close the shutter when desired amount reached and shut off source, i.e. switch off "A" and return knob to zero position (all the way CCW).
39. Turn off the Infinicon
40. Turn off the ion gauge
41. Close the gate between the diffusion pump and the bell jar
42. Turn on N₂ to ~5 psi
43. After system vented shut off N₂ immediately
44. Open bell jar
45. Remove sample
46. Replace sample holder in bell jar (if used)
47. Close rough pump to diffusion pump
48. Slightly and slowly open rough pump to bell jar only and hold the bell jar in place while a seal forms (until ~600 Torr). Then open rough pump the rest of the way and pump down to d 350 mTorr.
49. Close the rough pump to bell jar
50. Open rough pump to diffusion pump
51. Double check that N₂ is off, pressure d 350 mTorr only diffusion pump to rough pump valve is open, ion gauge off, no mess remaining.

Liftoff

1. Rinse a vial with spectraanalyzed acetone
2. Fill it about 1/3 full with spectraanalyzed acetone
3. Place the sample in the vial
4. Wait four hours
5. Rinse the sample with acetone for ~ 15 seconds to remove all of the loose gold (I direct the flow normal to the sample plane)
6. Rinse the sample with electronic grade isopropanol for ~ 15 seconds, being sure to rinse the tweezer tips as well to remove remaining isopropanol
7. Blow dry, keeping the sample pointed up so that trace contaminants flow away from it