Ice Contamination Rate Test

Theory

• All intensity measurements are proportional to number of electrons passing through sample. Measure them either directly from counts on CCD camera, or from measured OD from scanned film. Can also use Tecnai exposure time readings, but these are less precise.
• Let Io represent counts over a hole.
• Let Ic represent counts over a carbon area with no ice
• Let It represent counts over a carbon area with ice contamination
  
• Thickness of carbon: Tc=Kc log (Io/Ic), where Kc = proportionality constant representing scattering of carbon
  • John Berriman says that Kc=3750 A on EM 420 microscope, 120 kV, 50 micron objective with 3.6A cutoff
  • If using a Quantafoil grid, can assume that carbon is 200A thick, so can calculate Kc=200/(log(Io/Ic)) for a particular microscope
• Thickness of carbon + ice: Tt=Kc log (Io/It)
• Thickness of water: Tw=Kw log (Io/Iw), where Kw= scattering constant of water (ice)
  • Can assume that Kw=2 * Kc
• Thickness of contaminating ice: Tt = Kw/Kc * (Tt - Tc) = 2 * (Tt - Tc)

Caveats

• As with most tests for this, there are several assumptions which will affect the end result. Perhaps the most important of them is the assumed value for the thickness of the carbon. The carbon of Quantifoil grids can vary from ~20 to ~40nm in thickness. Furthermore, the scattering of electrons is strongly dependent on the accelerating voltage, and at 300kV, there is considerably less scattering than at 100kV, thus at 300kV, this test will yield an underestimation of the contamination rate.

Practical

• Use a Quantafoil grid to determine Kc, if not already known. Assume that thickness of grid = 200 A. Use an objective aperture so that carbon scattering = 10% (#2 on JEOL3200F, 50 micron on Tecnai)
• Since Quantafoil grids are often dirty and are quite thick, may be better to use home-made thin carbon grid for contamination measurements. Assume thickness = 70A, or calculate from Quantafoil data
• Set up low-dose on microscope such that in photo mode, you are imaging a fairly large area, including part of a hole, and focus area is inside photo area, preferably near hole
• Photo mag of 12K and Focus mag of 200K seems to work. Use about 50 micron underfocus for nice contrast
• Find a clean area of carbon, then pre-expose in photo mode for 10 minutes, 10 pA/cm2. Just want to be sure anything volatile is gone.
• Set shuttering to CCD control, so that lifting screen blanks beam.
• Take a CCD image of photo area
• Go to Focus, and lower screen to expose the small area for 1 minute, 10 pA/cm2 (JEOL) or 1sec (Tecnai)
• Raise screen, go back to photo, and take a second image. You should not see any difference -- this proves volatiles are gone and you are down to bare carbon.
  • With Quantafoil grids, this area may actually be darker because of plastic in the grid. You do not need to do this step
• Close gun valve, and wait at leat 2 hours. Longer if contamination rate is low. We usually do 4 hours
• Open gun valve, go to Focus mode. Lower screen, expose for 10 seconds
• Raise screen, go to photo mode and take CCD image
  • Alternatively, just take a photo picture, without re-burning the focus
• If significant contamination has built up, you should see a "burned" area where Focus beam has evaportated ice.
• Continue experiment until there is a significant difference between carbon and carbon + ice.
• Use the first image (after 10 min exposire at 12K) to determine the scattering coefficient
• Use the scattering coeffcient to calculate the carbon+ice thickness of the second image
• Image Analysis: Use your favorite program to measure counts over hole, carbon, and Focus area to get Io, Ic, and It respectively.
  • In EM Menu3, open "linescan", open "Linescan Average" to create box, and use "width (real)" arrow to increase box size.
  • In ImageJ^?, import he mrc file as "raw" and select area for density determination

Scripts to do this in SerialEM

pre-exposure macro

   R  # take record image before burn
   ReportSpotSize
   oldspot=$reportedValue
   screenDown
   SetSpotSize = 1
   WaitForDose 5000
   ScreenUp
   SetSpotSize = $oldspot
   R  # take a new one, compare buffers to verify that area is clean.

dose macro

   # need to have file opened for saving (File --> New)
   # need to have beam and record mode set up
   # need to have area pre-cleaned with above macro
   R  # take record image
   S  # save it
   ResetClock
   loop 180  # minutes -- increase to 240 for 4 h
      delay 60  # seconds. 
      ReportClock  # print a timer
   endloop
  R  
  S
  SetColumnOrGunValve 0  # close column valve
 

SerialEM setup

F20 setup

• Align scope
  
• Start Serial EM, make sure Tecnai low dose is off
  
• Prepare gain reference with dose calibration
  

Setup modes:

• Record: binning 2, 10,000x, 1sec exp, spot 3-4
  
• Focus: binning 4, 0.4 sec exp, 200,000x, same spot size as record
  remember to click Continuous update of mag&beam to save settings
  

Prepare sample for contamination test:

• Microscope: click Dose button. 2000 e/sqA ~ 10 min. Spot 1, minimize beam to area of interest
  

Check to make sure no contamination:

• Take a record
  
• Screen down, show focus, make sure over carbon
  
• Take another record and confirm no more contamination over area (burn spot from focus mode)
  

Begin contamination test:

• Open new file for writing
  
• MRC stack, file name *.st (save under E:\NYSBC\cuny_tests\contam_***)
  

• Run macro, make sure Area to show when screen down = Focus

# a 3 h contamination test 
# should have beam set up and area pre-exposed.
# Should have file open for saving.
# Should have record mode set up for proper time and binning
# be in low-dose, with focus same as record, but condensed beam
# be sure "area to show when screen down" is Focus

#Ttime = 3  #hours  -- time for contam test.

R # take record image
S  # save it

SetColumnOrGunValve 0  # close gun valve
echo "Waiting for contamination test to finish. Do not touch!"

ResetClock
Loop 3600 
   delay 3
EndLoop

SetColumnOrGunValve 1  # open gun valve
R
AlignTo B # make sure we are still in same area
R
S
ReportClock
# now burn:
ScreenDown
delay 15   # 15-second burn in Focus mode
ScreenUp
GoToLowDoseArea R
R
S
SetColumnOrGunValve 0  # close gun valve
ScreenDown   
EndMacro

Sample Calculation

• From John Berriman's notes:
• F20 with 3rd objective aperture
• Exposure time reading for 250 Angstrom carbon: 1.1 sec (remember inverse relationship between time and number of electrons)
• Exposure time for hole: 0.96 sec
• K = 250 A / (log(1.1/0.96)) = 4230 A

• Set ALLOWTOPICVIEW =

-- BillRice - 12 Apr 2007 * tf20contam_081312.st: tf20contam_081312.st