SEM Contrast Problem

• System working normally, then SEM went dark on return from spot mode. 
• Contrast control has no effect. 
• In lens detector no signal
• SE2 detector turns on and off when selected, but SEM software does not control voltage. Good image with manual control.
• Image from Auger detector is good, beam hits sample.
• Exit SmartSEM, restart windows, restart SmartSEM, still broken.

Suspect SmartSEM programing voltage for CEM failed. Fuse??

Check voltages.

Brightness is +2.6 V for 100%, 0 for 50%, -2.6 for 0%, both SED and InLens. This seems right.

(pin to right of center in audio plug)

Contrast (center pin) is stuck at 1.92 V for SED and between 0.51 and 0.74 V (non linear) for InLens

Both should be 0 to +5 V for for 0 to 100%.

HV Monitor on Supra is +1.227 V for 100%, 0.314 for 25%.

On InLens is +0.018 V. 

work around

still no news from Omicron 11/8/12

Run SED in local mode, works, but must remember to shut down if not required, for sputter, etc. Also no brightness control.

Modify InLens, open housing, disconnect voltage monitor wire and use BNC for external brightness control. Write external LabView app to control voltage.

Ion Gun Operation

FIG-5 Setup

1. Start Diff Pump Turbo (pump control 3)
1. run for at least 10 minutes after reaching 100%
2. PV Turboviewer gives details, pump power should be <~ 10 W steady state
2. Start 20-066 ion gun control power supply (brown box in rack)
3. Start the Pfeiffer RVC050 controller (grey box above 20-066) with the white toggle switch in Set Point mode.
4. Start FIG5 control program
5. Open FIG5 differential pumping
1. system pressure should stay constant or drop somewhat
6. In software set mode to Standby then click the ON button
1. Standby means the filament will heat but no the ion acceleration voltage is OFF
2. Parameters in Orange are not at desired values, parameters in green are good. In standby mode it is normal for Energy, Condenser and Objective to be orange, since the ion accelerating voltage is off, as shown below. They go green when the gun is on (Etch/Neutralize)
7. Wait for the Extractor Pressure to drop below 1 mPa. This shows the gas pressure in the ion gun. Since the Ar valve is still closed, any pressure reading is due to outgassing. This should drop with time.
8. Once the ion gun has degassed, open the leak valve. Click the magnifying glass to get a big pressure display, put it where you can see it on the computer screen from the vacuum system.
1. Slowly (1/10 turn adjustments) open the leak valve. Wait for the pressure to start to rise, ie to several mPa. Also watch the system vacuum. This will rise from the 1e-10 or 1e-11 range into the 1e-9 range.
2. It should take 2-3 turns to open the leak valve. If you open the leak valve too far, too fast, you could cause the ion gauge, ion pump, SEM field emission source to shut down.
1. This is bad.
2. Do not do this. Open the value slowly
3. Open the valve until the pressure exceeds the setpoint, typically 20 mPa. Now the Pfeiffer should stat to regulate the pressure (there will be a non-zero current showing on the Pfeiffer meter). The pressure regulates at ~ 25% above the setpoint when the gun is off, since the pressure sensor reading changes with the gun comes on.
4. Now you are ready for ion sputtering.
5. The system pressure should be in the low 1e-9 range when not sputtering. It may rise to ~ 1e-8 while sputtering.

Operating Conditions

Max current at condenser 66, Obj 61 for 2 kV.

Recorded current, clean copper sample no bias:

66   3.5 uA

70   2.2

80   0.27

85   0.11

90  0.05

Not much beam shift with Cond, Obj

At 1 kV, center shifts to (-0.5,-0.5) from (-0.9, 0.2), FWHM increases to ~ 1.5

Cond 60 387 nA

62  692

64   950

66  1070

68  860

70  590

75  193

80  65

85  27

Update profile Obj 61 Cond 66

Ion Gun tests and calibration

Check ion gun alignment, try to optimize focus etc.

Operating Conditions

Max current at condenser 66, Obj 61 for 2 kV.
Similar at 4 kV.

Recorded current, clean copper sample no bias:

66   3.5 uA

70   2.2

80   0.27

85   0.11

90  0.05

Not much beam shift with Cond, Obj

At 1 kV, center shifts to (-0.5,-0.5) from (-0.9, 0.2), FWHM increases to ~ 1.5

Cond 60 387 nA

62  692

64   950

66  1070

68  860

70  590

75  193

80  65

85  27

Update profile Obj 61 Cond 66

Sputter test

Change to oxide on Si sample, (old and dirty, lots of particles).

2 kV no raster, offset -.9, -.5, cond 66 obj 61, 25 mA emission, 20 mPa gas

After 60 sec carbon reduced, oxygen little change except peak shift.

150 s Si peak LVV starts to evolve, no SEM change yet. Ar peak starts to appear

Approx 300 um diameter spot, ~ 430 um below center. (Aligned by max count rate with apertures open) 

Clear contrast in auger image 93 pass 10, weak contrast SE2, more with inlens

Additional sputter, spot size does not change much. Now try line raster. Also record ion excited Auger. Clear Si, Ar, S

Move to lower end of test area. Center on sputter signal with 2 mm aperture. 

Record Ar excited Auger during sputter. SiO peak and O peaks clear.

So-called ion excited spectra may actually be ions+electrons...

Do not see expected sputter crater. May be shadowed at low angle or problems due to tip transfer lever.

Remove sample. Si sputtered clean except shadow area.

New Test

Second SiO piece same wafer. Put in round beveled 6 mm holder. This one with Faraday Cup, shows 17.8 nA for 3 kV max. 

Had to slightly pull out plate (away from back stop) to get current measurement.

30 deg tilt, 5 kV sputter no raster. Current exceeds 10 uA during sputter. 

Sputter 250 sec total, not thorough yet. Notice noise on AES. Look at beam current stability. Occasional glitches. Effect of sputter? SiO sputter test 3.aes spec 9, 10. 

After several spectra checking stability, spot mode, big C peak. 

Gets better. spec 11 same spot, spec 12 elsewhere, lots of C???? 

Sputter into contamination layer? Was gone before. Contaminated sputter gas?

EIS crashes, most data was saved.

Turn on mass spec, close gun valve, start turbo, close ion pump. 

Turbo only (was not run for a while, may not be that clean. 

Pressure to 3.6e-9, 

MS 28 - 20%,

2 17%

40  12.5

18 8.3

15 7.9

16 7.6

44 5.8

27 4.5

26 4.0

MS 200 uA emission, baseline 3.7 nA H peak 8.8, 28 peak 4.8

Open leak 2e-8, baseline 3.3, Ar+ 8.5, Ar 4.7, H 4.9, 28 3.6

Trend graph Ar+:++ is 70%, 14% or ~ 5:1

slight relative increase in 28.

Open leak 1/8 turn, pressure to 1.3e7

20 rises to 36%, 40 drops to 62%. H2 is 2%. 

Baseline 3.0 nA, 20 is 4.55, 40 is 5.0, H 3.2, 28 3.08

Strange, Ar pressure above ~ 1.5e-8, mass 40 drops wrt 20

Flash TSP close Ar pressure to 1.9e-9 dropping.

Open Ion Pump, 1.5e-9

Much later, +50 s sputter no C but still oxide.

Total 360 sec, now get spot, poor quality, 620 um shifted.

93 eV auger 10 eV pass

Fresh spot

2 kV ions offset -0.9, -0.5 (should have been, first shot -5.0)

First had no effect, sample current was 50 nA at -5, now 5 uA at -0.5.

Now 300 s sputter removes C, oxide still present, peak increased no contamination

1050 s burnthrough, ~ 250 um diameter, 460 um low

93 eV pass 10

Spectrometer apertures and transmission

Sample was flat clean Cu110, tilt 30 degrees, 3 kV max current.

Count rates vs aperture, images saved. 500 eV, pass 50 eV

• 1/1 900 kHz
• 2/1 830
• 3/1 380
• 3/2 285
• 3/3 125
• 4/1 225
• 5/1 350

• Conclusions: exit slit widths, 5:3:1, count rate 2.8:1.8:1, "hot" part of channeltron in center
• Exit slit no effect on image area

• Entrance 5 (1x12) aligned with 4 (1 mm) with 50% more counts
• Entrance 1 (6x12) only 8% more counts than 2 (6 mm)
• Round apertures 6:2:1, count rates 3.7:1.3:1, ideally if lens images sample on slit should be no difference. Angles should be fixed by magnification of 3. This may suggest tweaking lens factors to try to optimize transmission. These results may depend on pass energy....

Au low voltage ref spectra

Try to take spectra from Au/foil Cu110 sample. Go to 3 kV, after previous work at 10 kV. Max current mode, tweak settings. At first poor counts/shape below ~ 150 eV, also 'strange' octopole settings for 30 degree tilt. Expected ~ 49,50. Instead ~ 46 60.

Cu Graphene

Ultrapure Cu foil, anneal/recrystallize in H2 30 min 1020 C then 15 min growth CH4+H2 1010 C. Changhyun, Wu group

• 121016-1-L-b, 1/6 "large", SEM
• 121016-1-L-a, 1/6 "large"
• 121021-1-R-a, 1/6 "small"
• 121021-1-R-b, 1/6 "small"

large and small refer to Cu grain size. Even "small" is ~ 100 um

Turbo Pump Problem

• Prep chamber turbo did not pump down. Stalled at a few thousand RPM, ~ 40 Watts, bearings started to heat. Weird noise. Possibly scroll pump did not start. Foreline pressure gauge showed ATM. Airlock pressure ~ 25 Torr.
• Cycled power on pump. Now system goes to full speed eventually, but power stays at 66 W instead of dropping to 12-18 W. Foreline still showed ATM.
• Swapped scroll pumps for nano and prep. Start pumps together. Now both start normally, foreline pressures 0.13 mBarr prep (with nano scroll), 0.19  nano (with prep scroll). Both turbos ~ 15+-3 W, normal. Ultimate foreline pressure was ~ 0.055 mBarr, or 0.075 during bake.
• Possibly scroll pumps have trouble, or roughing lines leaking due to sharp bend on hoses. Exchanging scroll pumps seemed to fix problem. need to invert pumps again.

Let turbos run. Both prep and nano running about 10 W next day.

On Friday 10/26 switch scroll pumps back. Now prep turbo is slow to start, eventually reaches 100% with 88 W power. Prep pirani reads "atm". Shut down, disconnect scroll, pull out from rack, clean and reconnect quickflange with no stress on hose. Try again. Better, quick pump down, standby 18 W, foreline 0.10 mbar by pirani

problem was bad foreline pressure, possibly due to leak from strained hose. 

More trouble Nov 8, 2012

• Try to start turbo. Stalls at 63 W 27%. Pirani reads ATM. 
• Vent and retry. Similar results. Airlock shows ~ 10-15 torr.
• Stop again, examine turbo. Remove hose from scroll pump. Near pump the transparent plastic hose looks "misty". Remove hose, moisture condensed on inside ???.
• Reconnect hose, try again. This time system pumps down, foreline pirani indicates rapidly (< 60 s), goes below 0.5 mbar, pump reaches 100%.
• Then power and foreline pressure increase, ~ 40W at 100%, foreline 1.1 to 1.2 to 1.5 to 1.7 mBar over several minutes 
• This is very strange.
• Stop turbo, disconnect foreline, try again. This time 19W at 100%, foreline 0.16 mB and dropping slowly (normal).
• Possibly some water condensed in scroll pump when foreline was leaking, now being slowly pumped away. 
• after one hour foreline 0.12 mB, 17 W

Au foil testing

Move to Au foil reference on Cu sample
Good Auger Au 74, 164, 243 peaks, very weak at 2022 peak, probably due to weak excitation at 3 kV

Try 10 kV, large Octo offset to maximize at 2000 V, octo (50,66) instead of (50,50), now 2 kV peak/background is 11.5 % for 280 kHz at 3 nA aperture, was << 1% at 3 kV

Also 3 nA gun align way off at 10 kV, optimized near (-62,-68)
Max 10 kV gun (-48,-58), current ~ 9x greater

Try full scan this fixed Octo, weird below 1 kV

Optimize 2 kV (50.66)
Optimize 1 kV (50,56) repeat
optimize 500 eV (46,54) repeat, still OK up to 2.2 kV
optimize 250 eV (44.7.51.8) repeat,  OK from ~ 155
optimize 120 eV (43.2.51.0) repeat,  OK from ~ 155
optimize 50 eV (50.2.48.8) repeat,  worse, poor results below 250 eV, 120 opt better, fine above 1 kV
optimize 90 eV (47.6,47.7) strange, worse LV than opt at 250...OK about 700...

repeat 90 V opt with CRR instead of CAE, different results....still below

FPGA and LV 2012 update

Trouble getting FPGA to PC data streaming working after upgrade

ISM testing Cu 110

Testing ISM multipoint, line and SAM image modes

File CuAu test1
30 deg tilt apertures open, 3 kV max current, spot mode
sample cleaned ~ 4 weeks ago
Spectra in CRR mode 5, 10, 20 (max) trace C less O
Spectra in CAE 50, 40, 30, 20, 15, 10, 5

Peak/background and shape very different 5 eV pass, Cu peak/background different
internal scattering?

SED Detector How To

1. Start SmartSEM software
2. Turn on SED Control Unit
3. Push Remote button to enable control by SEM
4. In SmartSEM select signal SE2 on main or Dual Channel window
5. Adjust contrast and brightness
1. "contrast" is Channeltron voltage displayed on controller, value depends on beam current. For large currents may be ~ 1500 V, for small currents, may max out at 2500 V.
2. "brightness" is an offset voltage on the SEM signal, for the SED larger brightness makes the image darker. This is the opposite of the In Lens detector.