Albanova NanoLab FEI Nova 200, application course

Application course, teacher Dr. Chengge Jiao, FEI

2009-03-24 — 2009-03-26, lecture notes

FIB and dual beam, theory and applications PDF, restricted access.
Nova SEM construction PDF, restricted access.
Dual Beam application training PDF, restricted access.

Dual beam functionality and cross section

Eucentric height — the height of the sample where it stays fixed in image during tilt of stage.
Co-incident point — the point where the e^--beam and ion-beam crosses.
SEM-imaging is optimal at 5 kV ! Different from our experience where we mostly use 10 - 15 kV with good result.
Ion beam acceleration voltage can be varied between 30 down to 5 kV.
Electrons have an interaction volume diameter of 6 µm into aluminum.
Ions (Ga⁺) have an interaction volume diameter of 50 nm into aluminum.
Gas Injector System (GIS) used for depositioning, Pt in our system. But it can also be used to increase the etch rate, XeF₂, I₂, H₂O,...
Working Distance (WD) — distance from end of column to focus point. Ebeam has a nominal WD of 5 mm, eucentric distance should be 4.99 - 5.02 mm for FEI system. Ion beam has a nominal WD of 16 or 18 mm depending on model of column (Magnum ion column?). If adjusted properly F sets nominal WD and sample should be in focus. You should be able to switch current and stay in focus.
(Figure crossection, protective Pt-dep., staircase, bulkmill)
Bulkmill: 7 nA, Crossection 1 nA.
Deposition current I = X times Y times 6 pA per square µm
This can be used to decide what current to use for deposition. Electron depositioning, 15 µs per point (dwell time) typical value?

Why deposit Pt before crossection (CS):

To easy locate area to crossection
Protect against milling
Avoid source curtaining

(Figure redepositioning, curtaining, gaussian beam profile)
Curtain effect is from geometry, variations in material hardness. Pt depo. fills out these variations, makes milling effect uniform. Also inner cavities can cause this effect.
(Figure, round objects, stronger milling between)
Image CS with ion beam, small current. If sample is a mix of isolating and conducting grains, ion-beam image gives more info., cracks and corrosion show up better.
Two CS to make a corner must be made in parallell mode. Cleaning CS cannot be used, cannot be made in parallell mode. Instead make manual rectangles in parallell.

Curtaining effect, solutions:

Platinum deposition
Lower beam current
Change OverLap (OL) and dwell time (figure showing overlap)
Larger tilt angle in cleaning CS, 1-2 degrees more (53.5°) (Figure)
Angeled cut, rotate 90°, tilt 25°, use image scan rotation -90° to get back same geometry. (Figure)

Image with ion beam before SEM imaging of a CS can enhance SEM image.
Dig holes (mill spots) at back of CS to ground a floating (isolated) part.

SPI, Simultaneous Patterning and Imaging.
iSPI, intermittent SPI.
SPI mode: i_SEM >> i_FIB, fast SEM scanning.
Multiscan CS faster by three times than stair step pattern (regular CS). "Cleaning line three times".
ASV + 3D reconstruction.

Application demo at system

Things to check before imaging

Check SEM aperture correspond to hardware aperture, 7 is smallest and is 30 µm. There is no sensor telling the software which aperture is set by turning the handle. The software setting affect all current values given in the drop down menu for currents.
Check LUT = default (linear) for all channels.
Align feature to one main axis, horisiontal or vertical.
Here is was discovered that stage rotation was not calibrated for compucentric rotation. This was done with calibration 17 Quick stage rotation center.
Then Stage —< Align XT feature could be used to align edge of sample horisontally, just draw a line along the edge and it is rotated to horisontal position.

Ion beam default WD is 19.5 mm, set by <ctrl>F.
Electron beam and ion beam does not point at same spot within 5 µm, now it is 15 µm. Will be adjusted later.
Important: Decouple magnification, magnification of electron beam and ion beam are independent.
Ebeam deposition of platinum first:

Pt-nozzle should be 150 µm above sample and 150 - 200 µm away from center of ebeam.
Open Pt-nozzle and check for increased pressure in main chamber (10^-5 mBar?).
Form a sub-window over the same area where you want to deposit, start imaging at low speed, this will deposit.
Go on for 5minutes. This example area is 3.5 x 1 µm.
You have control over drift while imaging, compensate with beamshift.
Ebeam deposit possible with immersion mode.

--- Service alignment: rotation between e-beam and ion-beam.
30 pA ion beam and 1.6 nA ebeam SPI works well (SEM imaging while milling).
Regular CS —> Advanced —> Scan Method = Multiscan, 3 passes
This is about three times faster than normal mode Regular CS.
Shift + left mouse button —> beam shift.
Tilt 53°. Decrease ion-current to 0.1 nA. Shift away to do focus, shift "back" (right click on shift control).
<ctrl>B - beam on/off.
Arndts pillars: drill holes to ground area near intresting sample area.
Ebeam Pt. depo. 5 minutes. Ion beam 0.1 nA, 30 kV, Multiscan CS, then 50 pA.

Lecture 2009-03-25
FEI FEG SEM. Three parts: Source, Column, Detector.
Source: Thermoassisted field emission source, Schottky effect: lower workfunction for escaping electrons using ZrO, 4.5 -> 2.8 eV. Two Ion Getter Pumps (IGP) upper (2) vac < 2x10^-9 mBar, lower (1) vac < 2x10^-7 mBar.
Spot 1 - lowest current - smallest aperture
Spot 7 - highest current - largest aperture
Gun drift - source tilt/shift, changes during lifetime of source.
Mode I - search mode - HR (High Res) mode.
Mode II - UHR (immersion mode) 5-10 um image shift if well adjusted.
Light material (low Z) has deep penetration of electrons, better with low current.
Free WD = 5 mm = FWD in HR mode
Optical WD = 31 mm = OWD.
OWD = 3 mm in UHR mode.
EDX mode can be used to image magnetic material, only TLD detector used!
15 kV, WD = 5 mm, spotsize 0.5 nm
5 kV, WD = 3 mm, smallest, might give better image for mode II. For mode I a decrease of WD does not give better resolution.
Good backscatter electron collection -> should use small WD.
(Figure electrons collected by TLD and ETD)
Ebeam deposition gives more carbon into deposited Pt.
Ebeam depositioning:

Cover surface

Growth Mode I, lower kV (5 kV or less), higher current (> spot 4)

Image mode (F7) dwell time 15 us or slower, use beam shift to compensate for drift, area 20x2 um and 5 minutes gives 250 nm thickness
Pattern page:

Pt depo
Eectron beam
Dwell 25 us
Overlap (OL) 75%

Nanopatterning

Mode II, immersion mode
increase WD ( > 5 mm)
20 - 30 kV
to make thin line, play around with PASS parameter in Advanced Tab.
There is no stable value for volume per dose for Ebeam. For IBID it is 5 x 10^-2 µm³/nC

Papers, publications about this: Zarazago university, dept. of Electronic Engineering. Technical University of Eindhoven.

Alignment

User should do adjustment

No image

Check detector settings
Digital enhancment? Turn to Default
Switch on Crossover mode, check adjustment.
Auto contrast.

Before final image in mode II

In user interface: Lens alignment
Alignment 13: Stigmation alignment

Supervisor alignment

Mechanical aperture adjustment.
81. Crossover hole point calibration
10. Source tilt/shift (kV dependant)
44. UHR lens alignment
42. UHR stigmation alignment
43. UHR image shift correction
45. HR image shift correction

44, 42, 43 and 45 all WD dependent.

17. Stage Rotation Center.

Non-conducting samples, isolators: "Dust with Pt":

Ion beam at highest current, 20 nA
Insert nozzle
image at low mag (100-150 um)
drive stage around to cover appropriate areas

More added on work computer

2009-04-22

Ion alignment

Alignment 110

Compare current reading at drop-down menu with reading at lower right. Beam must be blanked ????
Wobbling: Adjust aperture position with knobs under hood on ion column. Note it is possible to move too far with the knobs so the next aperture comes into beam. (Figure: aperture strip).
Final check: Trick:

wobble on, zoom in very much to mill a small spot.
zoom back out, look at outer rim of circle for ellipticity.

20 kV, 10 kV OK. 5 kV all apertures black, no beam thru.

Alignment 210
(Figure: keep sample at top of holder when tilted, minimize risk of running into column)

ETD (Everhart-Thornley detector)
30 kV, use 100 pA as reference current.
Leave Shift adjustment to last, change L2 for best focus.
Adjust Stig sin, Stig cos and Stig.
Aperture position does not need adjustment, adjusted in 110
change aperture and repeat above for each
go to smaller apertures first.
smallest aperture need 110 adjustment
Please note there is a slow settling between changing apertures, about 10-20 seconds before image stabilizes.
Stig Sin and Stig Cos only depenent of acc. voltage, kV, not on current, i.e. no need to adjust for higher bigger apertures. Only Stig. adjustment.
Set lower magnification due to milling changing object, more milling for larger apertures.
Start with Bshift for all apertures.
Goto reference aperture (100 pA), make new hole by zooming in.
Position this hole in center of SEM image for a particular acc. voltage, here 5 kV.
Switch to all apertures and center hole with shift in ion-image. Mag should be about a 10 um scalebar.
When all is done all ion beam apertures should have same image position and focus as SEM image.

Adjustment 254 GIS

Nominal temp 45°C, min 40, max 45.
Check chamber pressure for increase when turning on GIS
if not check for clogged nozzle, close and open several times
if this does not help then increase heating.

Adjustment 253

Idle - Shutdown - Acquire cycle if no ion beam

Ion beam pattern parameters

outer to inner, inner to outer
bottom to top, top to bottom
left to right, right to left
Scan along the long axis of pattern is best (Figure: right to left)
Parallell / serial

Mill holes in circle, long dwell time, long step size (50 nm?) gives separate holes instead of a filled circle pattern for milling.
InA ion milling, 1.6 nA SEM image plus averaging gives very nice image quality during milling.

Square grating, pitch 1 um in X, 25 nm in Y, then overlay with pitch X = 25 nm, pitch Y = 1 um, this makes a matrix of squares.

Isolating sample

Taras G. sample, with copper tape on top. (Figure)

Check charging: goto 30 kV, zoom in
go back to 5 kV
if seriously charging you should see imager of the chamber, like in a mirror.
Move to copper tape
Platinum deposition with biggest aperture.
Deposit a connector between coppertape and sample surface.
Move stage during depo.
Thus make a band of Pt across whole sample
Thickness of Pt probably less than 50 nm

To reduce charging in general:

minimize sample size
Scan SEM concurrently is a way to neutralize charge, but uncontrolled
mill a via to a conducting layer below,

mill long time
switch Pt depo on and off during milling

This is the end of my lecture notes as taken down during the lectures and demos. Some of this is very terse and can only be understood when doing the respective adjustment.

Anders Liljeborg Nanostructure Physics, KTH.