Raith 100 E-beam Lithography (EBL)

The Raith 100 is a direct-write machine used for maskless lithography. A 30 keV electron beam is emitted from a field-emitter, and is focused through electromagnetic lenses. Using a deflection system the beam can be moved in the xy-plane. Lithography is crucial for the planar process (i.e. fabrication of devices), since it is one of the few techniques¹⁾ that allows local changes, whereas deposition and etching typically is a global step.

If you would like to use this system, contact Luc Wigbout.

The Raith 100 EBL system is located in the Cleanroom. Before you can use this system, you should have had a Cleanroom Safety Training.

1. Spincoat your resist of choice, see resist recipes
   • If sample is non-conducting: spincoat (multiple) layers of conductive polymer (e.g. PEDOT or Electra92).
2. Check the sample under the optical microscope. Look for clean spots where you can write your pattern. Write down the coordinates of the good spots (Global Coordinate System (CS))
3. Vent system.
4. Mount the substrate on the sample holder as close as possible to the faraday cup.
5. Make a scratch in the lower left corner of your substrate, about 1 mm in length, start on your sample and scratch outwards.
6. Clean sample and kinematic mount by blowing N2.
7. Mount the sample holder in the system, be careful; the sample holder is fixed with ceramic parts
8. Pump down.

Try to follow these 8 steps to develop a nice working routine.

1. Preparations
2. Origin
3. Focus
4. Writefield Alignment
   • When doing overlay: first 3-point alignment in Local CS then follow with WF alignment)
5. Beam current
6. Exposure parameters
7. Position list
8. Scan

Making sure everything behaves normally before really starting.

• run all the required tools: VegaTC, E-line and Raith Protocol Tool.
• open your design in e-line
• open wafermap if not already open (file/open wafermap
  c:/eline/allusers/wafer/100mm_ush.wlo
• determine size of writefield (2nd tab ‘set WF’) this is usually 100 µm with MAG 2300.
• check voltage (should be 30kV)
• turn on HV when pressure is below 2.5 E-4 mbar
• emission current should be around 60-90 uA
• choose PC (Probe Current) you will be using
  • high PC lower beam current, low PC higher beam current
  • PC 1 for contact pads
  • PC 10 everyday work
  • PC 14 highest resolution (at higher PC's focusing becomes really difficult due to noise)
• check in the image the yellow cross is available (extensions/measurements/radial grid with large radius)
• check in e-line that the PAT/IMG button is working.
• Make sure beam blanker is working.
• go to the faraday cup holder, (if you see nothing, zoom out). In stage control select ‘faraday cup on holder’ and press GO
• The Faraday cup has to be centered. Do this by moving the stage. The faraday cup is used to measure the beam current, if the faraday isn't centered correctly all beam current measurements will be off.
• Click ‘faraday cup on holder/edit/read, this will save the current position of the stage to the the faraday cup
• Measure the beam current once. PC10: ~0.150nA. PC1: ~8nA.
• In the table below you can find typical values for the beam current. If the beam current is low you can perform a AUTO HEAT procedure.

Define the origin of the global coordinate system on your sample. Typically this is the lower left corner.

• Go to e-line and CTRL-right click on the wafermap where you expect the origin of the sample to be, aim a little bit below to make sure you don't expose your sample.
• Zoom out and look for the origin.
• Keep sample origin in upper right corner of image window to avoid exposing the sample
• Keep zooming in and use the joystick to move the stage, keep the origin in the upper right corner.
• Go to >1000 magnification.
• Move corner of the sample to the center of the image this will be defined as the origin of your sample.
  If your process involves multiple steps you need to make sure you define the same origin next time.
• in e-line; go to 'xy uv' (3rd tab), click global button (or make sure it is set to global)
• do an origin correction (1st tab), click adjust.

Focusing will determine the resolution of your patterning. Do this well.

• Zoom out to look for scratch around the origin.
• Click ‘gain/black’ and then ‘auto’
• Use joystick to move around (change MAG if necessary)
• Find a particle which is <1 µm. The particle is used for focusing and later for automatic alignment procedures. A particle with sharp edges works best.
  Find the beginning of your scratch and move along the scratch to find a small particle.
  Good particle: small (<1µm, alone, bright, not in the scratch (height difference). When no good particles are present you can create a carbon needle by focusing the beam to one spot. In VegaTC, panels/analysis&measurement
• Focus (by changing the working distance WD). Use a reduced area, by double clicking in the image window.
• When you found focus; double click 'defocus'. This will set defocus to zero, and will allow you to go back easily to your focus.
• Safe the particle position in e-line (stage control/particle/edit/read). This will allow you to back to the particle easily.

when doing an overlay: first perform a 3-POINT ALIGNMENT in local CS and then perform a WF ALIGNMENT

With a writefield alignment the beam coordinate system is aligned with the stage coordinate system. Without a proper writefield alignment you will get stitch errors at writefield boundaries.

• e-line/2nd tab writefield control/scan manager
• start with a large scan area. Automatic writefield with Images/100 µm WF, 5 µm marks (right click to execute).
  Make sure the selected procedure correspond with the current set WF.
• open raith protocol tool
• check WF alignment tab and check if parameters for the zoom u, zoom v, shift u and shift v are in the proper range.
  0,9999> zoom u,v <1,0001 (accurate to last digit)
  u,v shift < ±0,030 µm
• repeat the alignment procedure until the parameters are in the correct range. After the large scan area, you can use the Automatic writefield with Images procedure/100 µm WF, 1 µm marks

The beam current is used to calculate the exposure time to reach the exposure dose of your resist.

• Measure beam current. It is necessary to measure the beam current again since the filament has to heat up to reach a stable beam current. Optimal filament is achieved after about 60 min. It is recommended to wait about 30 min after starting HV and starting patterning, especially when patterning high resolution structures.

Based on the measured beam current and the sensitivity of your resist you can now calculate the exposure parameters.

• e-line/4th tab/click calculator tab
• The goal is to set a specific dose, depending on your design choices and resist type and to have a beam speed that is lower than 15 mm/s.
• put area dose at 300 µC/cm2 for PMMA 950K. This depends on you resist and you design choices.
• when the tabs are red these parameters are not consistent, this can be resolved by using the calculator buttons next to the parameters
• set area step size to about 1/3 of your spot size.
• click ‘calc’ next to area dwell time. If red, add one min area step size. (area size = N * min area step size).
• Line dose should be 3 times area dose (not looking at the units)
• Do not set the beam speed too high, less than 15 mm/s
• Always click ‘calc’ next to dwell time last and make sure the beam speed is not too high

In the position list you decide what and where to pattern. You can choose the order of patterning and add all sorts of automation scripts.

• e-line/file/new position list, or the button in the toolbar
• go to GDSII database, go to your design and drag to position list, select at least one layer, this can be changed later
• right click 'pattern properties', click 'layer button' and select the layers you want to write in this step
• set U,V to at least 2,2 (due to imhomogenities in the resist). This will place you pattern at 2,2 mm on your sample.
• Create position matrix if you want to pattern more copies of your device
• Add beamshut down script to the bottom of your positionlist!
• save your positionlist, especially when doing a proces with multiple steps

• check if you didn't forget anything
• You can calculate the time it will take to pattern your device. Patterning parameters/drop down/times.
• Click on scan all in the 'filter' menu to start!

• e-line/3rd tab/exchange position and GO
• Check if HV is off (beamshut script should have done this)
• vent
• Unload your sample (GLOVES!)
• Put sample holder in box
• pump down and check vacuum before you leave

Each resist typically has a dedicated developer. PMMA type resist can in general be developed by MIBK or a MIBK/IPA solution. See resist recipes for the specific developing process for a specific resist.

As for developing, a resist typically has a dedicated remover. Most resists can be removed -lifted off- with aceton. See resist recipes for the specific lift off process for a specific resist.

• Make sure all previous flags are cleared by unchecking the checkboxes next to the blue flags. (Do not use the blue flags, they have different meaning!)
• 3-point alignment is done to perform a local coordinate system correction. The global markers on the design will be used to line up the local coordinate system of the design with the stage coordinate system.
• Go to faraday cup.
• Open you design. Make sure you are in the viewer and not in the editor, else the flags and stage position will not be shown on your design.
• Open toolbox (T)
• Beware of the meaning of the flags!
• Edit design grid to 1 µm to place the flags accurately, show grid
• Go to the first entry in the (saved!) position list (while your still in global CS), small black triangle in the position list toolbar/go to position.
• switch to local CS. Xy – uv/local button/3 point tab
• Set the magnification on the SEM to 2300x (Same value as the magnification for the writefield)
• check with the SEM that you're at the position of the lower left global mark.
• center the image, make sure the yellow cross sits at the center of the global mark.
• Place 1st flag on the center of the global mark of the lower left corner
• Then click read and then adjust in ‘Adjust UVW (local)’ for P1
• move to the second global mark, use CTRL-right click in your design to move to the center of the mark.
• Check position in SEM image
• adjust with the joystick to the center of the global mark
• Set 2nd flag in e-line press read and then adjust
• Do the last flag. It should almost be perfect
• ctrl-riight in your design, check SEM image, adjust with joystick to center of global mark
• place 3rd flag
• Press read and then adjust
• Now you have coupled the design coordinates (local coordinates) to the stage coordinates

• After stage alignment, comes e-beam alignment.
• Select an entry in positionlist, right-click to go to properties
• Select layers, only automatic marks, layer 61
• Select as working area WF calibration.
• Click ‘position button’ (Since we are now in local CS (design CS) you do not select the coordinates where your device will be patterned) the coordinates will be calculated from the working area you selected.
• go to your position and click execute (F9)
• Run the WF calibration a few times until the parameters are correct in Raith Protocol Tool.
• When you want to start the patterning, select layers you want to pattern, choose the working area and click the position button.
• SCAN all to start the patterning.

Move the stage to a empty part of the waferholder. Turn on the beam and make sure focus is good. Press 'AUTO' next to 'HEAT' in VegaTC. The system will find the optimal saturation point for heating of the filament and will perform an auto gun alignment. Check if the gun tilt and gun shift values are between -20% and 20%. If the values are 0% this means the procedure failed. To high values need be adjusted with a mechanical alignment of the filament, find a technician to do this for you.

Often it is necessary to switch to a lower PC to write large structures. PC-1 has a 700nm spot size, compared to a 70nm spot size for PC-10. Writing large structures >1µm, can be done with a high PC, but will require long patterning times.
Since the spot size of PC1 is large, the resolution to do a proper writefield alignment is low. For this reason the PC is changed after the writefield alignment is performed at PC-10, but before the beam current is measured. Switching between PC generally introduces a small shift in the pattern. When you take into account this error during the design of your pattern you can easily correct for the shift without going through the trouble of a proper writefield alignment for a low PC.

CTRL-right click on the wafermap will move the stage to the position. This is used when you want to navigate close to the origin and when doing 3-point alignment.
CTRL-left click will point to a feature to do an alignment. This is used during manual writefield alignment.
L – shows the layer information of your design
W - shows working area information of your design
H – selects the handtool to move around on the design
T – opens the toolbox.

Double click to create a reduced area in the SEM image. When reduced area is active you can change the size of the area using the right mouse button.

A failed WF alignment typically happens when a particle isn't anymore in the field of view (FoV) of the WF alignment procedure. This indicates the parameters have drifted from the optimal settings after a -few- bad WF alignments. A bad WF aligment can be due to: bad focus, WF alignment at low PCs (low resolution), particle not centred, etc…
In the second tab 'writefield' control, open 'writefield manager', reset the WF alignment. Choose a WF alignment with large FoV (> 10µm). After this rough WF alignment reset the shift values in the writefield manager. Then perform the WF alignment you need till you're happy with the correction values in Raith Protocol Tool.

fundamentals_of_electron_beam_exposure_and_development.pdf

The Raith-100 e-beam lithography machine can be used to pattern nanoscale structures using an electron beam. The system is, in essence, an scanning electron microscope (SEM) with additional features such as a pattern generator that can write the desired patterns. The e-beam is generated by thermionic emission from a hot cathode emitter at an energy of 30 keV, after which it is formed by the Wehnelt cylinder and several lenses (a.o. the stigmator and objective lens). Within the lens-column, there is also a deflector housed, which can change the incident angle of the electron beams. The amount of power necessary the deflect the 30 keV e-beam, increases with the beam angle. Thus, deflections at larger angles need bigger amplification. In the Raith-100, the large amplifiers tend to be noisy, so when writing, we will keep to the smaller amplification, so that the maximum angle corresponds to a field of view of 400×400 um.

The system can still be used as a SEM, were the electron beam is scanned across an area. The scanning movement is ‘locked-in’ with a detector such that an image can be created. The area that is imaged, is the field of view (FOV) at that specific magnification, e.g., at 575x mag., the FOV is 400×400 um and at 2300x mag., the FOV is 100×100 um.

When writing, we typically use 100×100 um as the area of our writing fields (WF). Thus, the WF is the area onto which can be written by purely changing the deflection of the beam.

Since most patterns tend to be a bit larger than 100×100 um, we have to come up with creative solutions as not to write a noisy pattern. One way of overcoming this issue, is by stitching several WF's in two dimensions. This can be done by moving the sample with high accuracy.

The sample is put onto a sample holder which can be mounted inside the Raith-100 onto a stage with sapphire ball spacers to ensure that it is held in place with extreme accuracy. Furthermore, the sample stage can be moved in X and Y directions. The movement is measured with a laser-interferometer.

Be aware that when you mount the sample, the stage should be in the exchange position. Otherwise the mirrors of the interferometer could be exposed to the outer environment (and potentially get dusty).

By moving the stage 100 um along the X or Y axis, one could switch to another part on the sample. In order to make sure that all the WF’s are aligned–connect properly–a WF-alignment procedure has to be done. This procedure makes sure that all the writing fields are correctly stitched to each other.

WF alignment with images is done using an automated function, which does the following. After you searched for a particle or clear feature with a size of about 5 um, the procedure takes an image.