FIB









FIB basic introduction booklet
FIB ion milling introduction booklet
FIB auto-TEM introduction booklet (additional introduction mandatory, please contact admin)

Electron optics

  • Non-immersion column
  • High-stability Schottky field emission gun at 20 eV-30 keV
  • 0.9 nm  resolution at 30 keV STEM, 1.2 nm at 1 keV with beam deceleration

Ion optics

  • Sidewinder ion column with acceleration voltage: 500 V – 30 kV
  • Beam current range: 1.5 pA – 65 nA
  • Ion beam resolution: 3.0 nm at 30kV

Detectors

Up to four simultaneously detected signals

  • Trinity Detection System (in-lens and in-column)
  • T1 segmented lower in-lens detector
  • T2 upper in-lens detector
  • Everhart-Thornley secondarz electron detector
  • Retractable low-voltage, high-contrast, segmented solid-state
    backscatter electron detector
  • Retractable STEM 3+ detector with Brightfield/ Darkfield/ high angle annular darkfield

Stage

Flexible 5-axis motorized stage with compucentric rotation and tilt:

  •  XY range: 110 mm,  Z range: 65 mm, endless rotation (360°)
  • Tilt range: -15° to +60°
  • Max sample height: Clearance 85 mm to eucentric point
  • Max sample weight at 0° tilt: 2 kg (including sample holder)
  • Max sample size: 110 mm with full rotation (larger samples
    possible with limited rotation)

Request for introduction

FIB introductions are organized about twice per year. The prerequisites to enter the introduction are:
The next FIB intro is scheduled from:

Monday April 27nd until Friday April 7th, 2020.

  1. The participant needs an expertise level at  SEM. Practically, this means you must bring a minimum of 20 hours of hands-on experience in SEM. A heads-up intro on the Tescan SEM can be offered in the weeks ahead of the FIBSEM introduction.
  2. A thorough discussion on the project – and especially how the FIB comes in – before the introduction. Only projects that can theoretically be matched to the functionality of the FIB can be considered.
  3. There is a maximum number of 4 (four) participants per intro.
  4. Administration. You need to:
    1. Fill out the request form below.
    2. Fill out this request for introduction form in parallel, filled out and signed
    3. If you are not a member of the Adolphe Merkle Institute, you must upload  Agreement for employees and students from UniFR and external guests. Fill it out and sign it where appropriate.
    4. The intro is only available to academic users.

However, please note that COVID restrictions may cause changes to these dates. Worst case, the course may be skipped completely. Furthermore, note an absolute limit of maximum four participants, using the first come first serve basis.


Resources (PDF)

SCIOS manual

Auto Slice & view (3D tomography)

AutoTEM (thin lamellae making)

Maps (correlative microscopy)

Nanobuilder (patterning software)

Pathfinder (EDS software)

Shortcuts List

System overview

  • The E-beam column is a NiCol, a non-immersion column and is positioned vertically (pointing down).
  • The Focused ion beam (HT sidewinder) is positioned at a 52° angle compared to the horizon.
  • A platinum gas injection system needle can be inserted to allow precise and localised platinum deposition. The platinum precursor is a gas, that is deposited by either the electron beam or the Ga-ion beam.
  • An on-board plasma cleaner allows to clean the sample and the chamber
  • A manipulator needle can be insrted to pickup and move very small pieces that were cut out
  • A number of detectors.

002 - system overview

The FIB imaging resolution is about 5nm, whereas the e-beam resolution is better (about 1 nm). These values are also depending on the sample type.

Detectors

  • Trinity Detection System (in-lens and in-column)
    • T1 segmented lower in-lens detector
    • T2 upper in-lens detector
  • Everhart-Thornley SE Detector (ETD)
  • Retractable low-voltage, high-contrast, segmented solid-state backscatter electron detector (DBS)
  • Retractable STEM 3+ detector with BF/ DF/ HAADF segments
  • IR camera for viewing sample and chamber
  • In-chamber Nav-Cam sample navigation camera
  • Integrated beam current measurement
  • Up to four simultaneously detected signals

003-DETECTORS

Loading a sample

  • Load the highest samples on the right side (when standing in front of the microscope) of the stage. In any case, do not load samples that differ strongly in height.

001 - multisatge samplepositions

  • Keep the stage clean. Use basic household cleaning products.

Notes on sample preparation

  • Use long stub pins. Short stub pins will not work well.
  • Avoid using C-tape to mount samples on a stub. It “sets” over time and may cause therefore drift during ablation. Use colloidal silver instead.

NavCam

To use the NavCam

– Do not take a NavCam image during pumping. It may be imprecise because o-ring movements during pumping. Better to wait until pumping is finished.

– Move the stage to its 7 mm working distance: start the e-beam, focus, link and set Z= 7.0 mm.

– take the image (stage > take nav-cam photo or CTRL+SHIFT+Z).

– in the stage camera window (not the nav cam window), double click. This will set the 7 mm overlay line to the correct height of 7 mm

Electron Column

Overview
  • The SEM beam and Ga beam are 38° apart (the FIB beam is at 52°, the e-beam at 90°). The working distance (WD) of the SEM is set at 7 mm by the factury, the Ga beam 19 mm. CTRL+F will reset the WD of both beams to these values. The WD of the ion beam MUST be 19 mm. If the sample is tilted to 52° and at 19mm but not in focus, the ion-beam alignments must be repeated.
  • The beams are meeting at the Beam Coincidence Point (BCP). See below. Everything is aligned to the BCP.
  • FIB resolution < SEM resolution. Typically, a FIB can produce 5 nm resolution at 30 kV, SEM about 1 nm at 30 kV. Due to the higher mass of the Ga compared to electrons, the Ga is more prone to produce surface information.
Alignments

All alignments will be made by the supervisor

  • Chamber cleaning: every month, less than 1 hour, including 3-4 cycles. Damage on the EDS windows is expected at 700 hours of plasma cleaning
  • Ion beam alignments: every 2 months
  • Electron beam alignment: every 2 weeks
Column modes

There are three column modes: standard, OptiPlan and OptiTilt

Standard mode
  • Is used for aligning (e.g. lens modulation, crossover, …), finding BCP, basic low magnification navigation…
  • Typical detector: ETD
  • Advantage: largest field of view, all currents available
  • Disadvantage: not suitable for high resolution (no beam deceleration)
OptiPlan mode

OptiPlan and OptiTilt mode will apply a stage bias, typically 20V, which pushes/helps the escaping electrons back up. This increases the electron yield on the in-beam detectors T1 and T2. When active, note the small arrow after the KV value.

  • Is used for high resolution imaging at 0° tilt.
  • Typical detector: T1 & T2 (note that the ETD detector signal will drop)
  • WD can be set to < 5 mm (you anyway will not tilt, it is only useful at 0°)
  • Advantage: high resolution, more signal
  • Disadvantage: smaller field of view, not all currents available,
  • Pushing back the electrons does not make sense for STEM
OptiTilt mode
  • As Optiplan, but corrected for a tilt to 52°, is used for high resolution imaging but  at 52° tilt.
  • will apply a default 20 V stage bias (note the small arrow after the KV value)
  • Typical detector: T1 & T2 (note that the ETD detector signal will drop)
  • WD should be at 7 mm (or about there, depending on BCP)
  • Advantage: high resolution
  • Disadvantage: smaller field of view, not all currents available,
  • does not make sense for STEM
Beam Deceleration

During OptiPlan and OptiTilt modes, the beam in the column will be accelerated by 8 kV (this reduces chromatic aberration in the lenses) and decelerated before the beam exits the pole piece.

FIB  Column

Overview
  • known as HT sidewinder.
  • Using the Ga ion beam will:
  • Always induce Ga implants
  • Always amorphorize your sample (outer layers)
  • CTRL + F will set the WD of the Ga ion beam to 19 mm (at which it is in focus)
Aligning SEM and FIB

This is done by setting up the beam coincidence point. Finetuning it done by an image shift of the FIB image.

Beam Coincidence point Protocol

  • Reset the beam shifts (beam shift > right click > reset)
  • Set the Beam coincidence point
    • Search for a landmark in your SEM image and center it at the yellow cross (if there is none, get one in the overlay)
    • Link the SEM image
    • CTRL + F (or move WD of SEM to 7 mm)
  • Rotate to about 10°, use the z-stage (not X!!) to move the landmark back to the middle. Hold the middle mouse button and move the mouse down to move the landmark down.
  • Go in steps to 52°. At 52° the same object (landmark) should be exactly at the cross again.
  • Image with the Ga ion beam:
  • use a low beam current (10 pA or about), 30 kV
  • Zoom out if needed and find the same landmark.
  • If the same object is not in the middle, use beam shift XY to put it in the middle of the Ion beam image. The shift XY buttons are on the physical control panel below the central screen.
  • switch off the imaging as soon everything is fine (you are milling away stuff)
  • the couple magnifications: OFF
Lifetime maximizing concepts

If the Ion beam is switched off, the current is read by a Faraday cup and shown in the software bottom left (ion Beam current)

This value should be about the same as the beam current setting of the ion beam.

If this is not the case:

– If this is the case for only 1 current setting, but not for others: the Automated variable aperture strip wore out.

– If this is the case for all currents settings: the beam acceptance aperture wore out.

If the FIB image is not in focus at WD=19mm (i.e. after you hit CTRL+F), then the alignments are off (ADMIN)

Beam settings

 

The FIB beam diameter communicated is always FWHM (full width at half maximum): the FIB beam has the form of a guassian curve in 3D.

The beam scans at distinct positions. At 50% overlap, the FWHM will exactly coincide with the center of the next position. At 0% overlap, the beam at two neighboring positions will touch at FWHM.

Basic milling concepts / preferential milling

The milling efficiency is a function of the local curvature of the sample. Milling works most efficiently between 75-85°. Hence, do not mill samples that are not flat: you will end up with a preferentially milled object.

Crystalline structures will cause channeling of the ions depending on the Bragg conditions. Hence, crystalline structures will not mill to a flat surface,

Applications

Cross sections protocol

1. Focus, link, go to 7 mm WD (at 0 degree)

2.  insert the GIS needle: it might drop a shadow on your image. Hit F9

3. Place a Pt marker at 0° with e-beam deposition

– set e.g. 20µm x 2µm , deposit using the Pt application

– Patterning settings:

* 20µm x 2µm x 500 nm (z),

* 15 µs dwell time

* Application Pt_EBID (E beam induced deposition)

– Microscope settings (might slightly vary on your sample):

* Standard mode

* 2000 X

* 2 KV

* 1.6 nA beam current

(total time is 4:13 patterning time)

– hit start patterning

– e- beam deposition is soft and slow, iSPI is not possible

– retract the GIS needle. Hit F9. remove the pattern.

4. Set the BCP. You may use the deposition from step 3.

CTRL + i (will tilt to 52° – ion beam), CTRL + e (will tilt to 0° – electron beam)

5. Pt deposition with the Ga ion beam (in the Ga ion image)

– Don’t make an image with the Ga ion beam!

– draw a rectangle in the Ga ion image with the patterning tool (e.g. 20µm x 2µm)

– z= about 1 µm

– Dwell time: 200 ns

– select Pt dep (not Si) in the application

– calculate the Ga current required using the magic number 6 (pA/µm2).

* 20 µm x 2µm x 6 = 240 pA

* use this value and chose the closest current for the Ga beam

* this is important! too much current and you will mill instead of deposit

* Too less current you you will destroy your vacuum

* You should get a time round 3-5 minutes

– Insert the Pt GIS

– Press F9 in the ion image (this will contrast/brightness correct and take a snapshot). Make sure you have the ETD selected

– Check the position of the rectangle, overlay the e-beam deposited marker.

– Run the deposition

– retract the GIS needle

6. Bulk mill

– Use the regular cross section pattern. Position it just below the Pt deposition you just made with the Ga

– Place the RCS pattern a bit wider than the Pt deposition marker (about 10-20%) and not exactly touching the Pt above it

– Application: Si multipass

– determine / decide on the depth (e.g. 5 µm)

– calculate the Y, with Y> 2 times Z

– Pick a Ga ion beam current to mill between 2-5 minutes (rule of thumb, no calculation needed)

– iSPI is possible. Use the brightnes contrast buttons on the physical control panel to adjust B/C, not F9

7. Polishing

– set a tilt angle to correct for the beam shape: somewhere back 0.5-1.5° (i.e. between 50.5° and 51.5° absolute angle)

– refresh the ion beam image (F9)

– place a cleaning cross section between the Pt deposition and the edge (or a little bit over it) of the hole the step before made. place is just a little bit into to Pt (is this correct?). Width of the section: about as wide as the Pt deposition

– go two steps back in beam current (in the list)

– Depth (currently, we think this is Z): 1/4 of the previous setting. If it was 1000 nm in the regular CS, set it to 250 nm now.

– You can use the iSPY: this will stop the patterning temporarily, make a SEM image and continue

8. Imaging

 – Go to a very low ion beam current (10 pA)

– Press F9

– Curtaining issues: Do not use the ETD, since curtaining is the strongest in that detector. Switch to OptiTilt and use T1 and T2.

– Lower beam currents: more focused Beam, but more curtaining.

 – To image the object with the FIB:

* Go to 0° tilt

* rotate the stage 180°

* scan rotate 180° (use SHIFT F12)

* ise really low currents and do a focussing (could be needed) outside the ROI.

9. EDX on a cross section

– adjust the formula for Y:  Y > 3 x Z (or even 4 times). Because BSEs could produce X rays in the gap.

– make a second bulk milling on the right side of your ROI to avoid shadowing. You will end up with an L-shaped gap (below and on the right of the ROI).

– Be careful with the interpretation: the imaging is from under an angle, which means that shallow layers may overlap

Slice and view protocol

 1. Find a ROI using the SEM

2. Make an e-beam Pt deposition if needed

3. Setup the BCP

4. Move to 52° and make a Pt layer (1 µm thick) with the Ga beam. Use the magic formula to calculate the Ga ion beam current = Width x Height x 6  (in pA)

5. Draw a regular cross section pattern. Make it really big, 2x the width of the Pt deposition (so material can leave during the Slice and view). Place it a bit away from the Pt (the higher the beam current of the Ga ion beam, the further away).

* Use Si-multipass overlap 50%

* 30-50nA

6. Mill the side trenches, bottom to top (both trenches), same settings: thick line of the yellow square should face to the inside

* Left trench: left to right

* Right trench: right to left

7. Remove redepositioned material (redep) by repeating left to right (on the left trench) and right to left (on the right trench)

8. Open Auto Slice and View. Make sure you save your project and all data on drive D, not on drive C.

7. Setup Auto Slice and View milling

– X resolution: consider Nyquist (1/2 the size of the smallest object)

 – Y = 1.27 x Z

– don’t make elongated cubes. Try to have a more or less cubic cutout

– time per slice: 10-20 seconds to cut. The aim is to have about 1 slice per minute

9. Make a fiducial. Increase the depth of the cross from 500 nm to 750 nm.

10. Set the SEM imaging conditions in UI

* Choose OptiTilt

* Use a low kV

* Choose T1 detector. SEM beam current is not too important

– zoom in and put the ROI in the center

– focus, stigmate and link (!!), reset any beam shifts

– Update the FIB image (F9)

– in Slice and view, place the pattern correctly and find the fiducial

11. Green clean pattern

– Reason: to find the edge of the block

– Setting 1: 10 (makes 10 slices inside the yellow box)

– AOI overlap: (area of interest) 5: will make 5 outside and 5 inside the block

12. Setup imaging conditions in Slice and view

– Make sure Y-shift correction is ON

– assure the AOI is in the center

– autofocus: slows down the procedure tremendously. If you use it, then maybe every 50 slices or so.

13. Run

Slice and view advanced setting

1. Make sure you save all data on D

2. Multi SEM: add additional detector images (simulataneously)

– setup the images in the 4 quandrants of the UI

– get these 4 quadrant using MultiSEM in auto Slice and View

3. Selected area: will scan only one part of the image (=faster!). Zooming in using SEM is also reducing the unwanted info at the edges, but is not faster.

4. Alignment (under imaging). Can be anything (does not have to be the standard crossed fiducial). Aligned area is the image to the next slide.

5. Options that are enabled while running (not greyed out) can be changed – and will be applied – on the fly.

Nanopore drilling

Alignments

If needed:

– focus, stigmate: use 2D box on Si test sample

– beam shift alignment: 100 pA is the reference

– stage rotation alignment: do on the periphery of the stage. You can use continuous rotation, which is the most accurate, but is often struggles.

Make sure the vacuum is really good (1 * 10 -4). Bad vacuum makes larger, less round holes

Concepts

1. go to a 10 kV, 200 pA

2. Making smaller holes:

– less passes

– longer dwell time

3. Making rounder holes:

– less passes

– maximize dwell time

4. Check STEM to know if the hole is through (2 kV)

Protocol

1. Use the horizontal row bar

2. Rotate the stage 180°: the holder is on the right of the stage

3. Focus on the edge of the SiN window, Mag 25 000X

4. Link, zero the beam shift

5. Set the BCP using a landmark, and go back to 0°

6. Zoom out (600-1000X) and find the square. Save the position. Then reiterate 3-5 on a landmark very close to the window

7.  Tilt to 52°, set Mag to 15 000X

8. go to ion beam, set the mag 1000-2000X and find the window. Use 1pA. Take a snapshot

9. Draw a circle pattern with:

–  outer diameter: 7 nm

– 25 ms dwell time

– 1 pass

Si application

– time: 25 ms

10. Run pattern

(from here optional)

11. Don’t focus / image the result. Contamination may clog your result

12. make a relative stage movement of 180° = ready for STEM

13. insert the STEM detector, use horizontal holder option

14. load scanning preset C6, and check the image in the STEM.

AutoTEM (TEM lamellae preparation)

1. Load the sample and the Moon grids. The halfmoon grids go in the 38° tilted bar. Make sure the sample is not too high

2 Sharpen the tip. Use CTRL + J to get the overlays

3. Tip should be 1-2 µm diameter, about 5-10 µm long

4. If needed, do an esayLift alignment (guided)

5. Find ROI, focus, stigmate, link

6. Do a Pt e-beam deposition

7. Find the BCP

8. position the manipulator. position XY in SEM, Z in FIB.

Aborted due to a too low GIS position. Will be corrected by the technician.

MAPS

Make sure your data is saved on D

– Idea of a pyramind with layers on top of eachother, first is e.g. from the navcam

– HFW = horizontal field width,overlaps in tile stich should be 20%, not lower

– Open the data in HD view (installed on the SPC)

Non conductive samples / charging samples: What to do

1. coat with Au layer

2. Mount on ITO glass instead of other (non conductive support)

3. minimize the size / reduce the exposed area (zoom out)

4. Lower the Kv (< 2-3 kV)

5. Lower the Beam current (<50 pA)

6. Tilt (increases signal)

7. Use BSE detectors

8. Lower the WD

9. Use optiPlan/OptiTilt

10. Tweak the scan presets:

– Scan Fast

– Interlace

– Use drift correction

Plasma cleaning

Two types of plasma cleaning exist:

– chamber cleaning. Takes about 5 hours

– sample cleaning. More gentle, takes about 5 minutes

Lower the stage (25 mm WD)

Go to standard mode

Sleep vs beam off – how to switch off after a session

Ga ion beam

The following compromise must be taken into account:

– Ga ion beam too long idle: waste of Ga

– Ga ion beam wake cycle: uses a lot of Ga

Therefore:

– If the FIB is used the next day: don’t go to sleep. Just beam off. This stops the extraction voltage.

– If the FIB is not used the next day, goto sleep. This cools down the FIB gun

In any case, you:

– beams are off

– switch off / retract the detectors

– samples out

– chamber is pumped

– check the reheat time (see below)

Reheat

– every 80 hours, the Ga source need to reheat: shutting down, cooling down and reheating. Hence, you cannot do experiments longer than 80 hours.

Pt deposition

– Switch off if the user does not need it

– 10 min to heat

Automated ion beam alignments

– use the graphite sample

– zero stigator and beam shift

– CTRL + F to assure 19 mm WD

I column alignments, and run the selection

Chamber closing

Contact

Dr Dimitri Vanhecke | tel: 9509 | dimitri.vanhecke@unifr.ch