Scanning Transmission Electron Microscopy Facility
Frequently Asked Questions
Will PCMass run on Mac or Linux?
No. As the name implies, it needs a Windows PC. However the hardware requirements are modest (1024x780 or greater display)
so it should run on most word processor or laptop PC’s.
Is there another way to view STEM images?
Yes. The STEM image format is a 4096 byte header followed by two 8-bit channels interleaved. PhotoShop has an
“Open As” option “Raw” that will read this. Gatan’s Digital Micrograph. However, these programs lack many of the
features of PCMass for mass measurement. Images can be compressed for display by up to a factor of 100 (.jpg) for
visual use in publications, but are no longer useful for quantitative measurement.
What causes the TMV calibration to be other than 13.1 kDa/A?
The STEM calibration varies slightly over time and with hardware modifications to the data acquisition system.
We frequently include TMV controls in unused slots of the freeze dryer and plot the long-term trend of all measurable
TMV segments. We will include our best estimate of the current TMV calibration when we notify you that your images are
available on ftp.stem.bnl.gov. If the TMV appears intact but deviates from the specified value, it is usually for one
of four causes:
Residual salt or small protein fragments have accumulated in the central hole of the TMV or in the crevice where the cylinder of
the TMV contacts the plane of the substrate. This increases the TMV M/L and distorts the radial density profile. Frequently the
accumulation is quite variable in different segments. Residual salt crystals (if present) tend to show up with diffraction contrast
on the small angle detector (brighter or darker than the same area on the LA image (insensitive to diffraction contrast,
proportional only to mass). Since this depends on geometry and local affinity, specimen mass may or may not track the TMV
value and caution is necessary. Please consult STEM staff if you suspect this.
Residual salt or small protein fragments are coating the background between particles but not the region between particles
and film, causing the background measurement to be artificially high. Since the substrate mass is frequently as much as half
the measured mass of particle plus background, too much will be subtracted and the TMV value will be low. One indication of
this will be an aberrant plot of Mass or M/L vs. radius of integration. This will increase with radius, reach a peak, then
decrease as the radius of integration increases further.
The dose was high, leading to mass loss. We aim for a dose of 10 el/Å2 which gives 2.5% mass loss at –160C.
The dose for each image is computed from the counting statistics measured by the background program and printed in
the PCMass display. If the image was recorded with a 0.256u scan, the dose will be 4x higher than with a 0.512u scan
and probably not useful for mass measurement. Dose response curves for TMV (96% protein) and control protein track, so
the TMV normalization effectively corrects for small amounts of mass loss. Nucleic acid does not lose mass, so care
should be used in TMV normalization.
The specimen is overcrowded. If crowded with small particles, these may not be masked completely by the background program
and the masking threshold needs to be set lower. If the masking excludes too many pixels in any of the 16 sub-areas the program
will exclude that area (marked with red in the background display) and try to interpolate from adjacent valid areas. We
recommend skipping such images in favor of these with 1-10% of the area occupied by specimen and 1-5% occupied by TMV.
STEM operators search for such areas and only record overcrowded or aberrant areas for documentation of problems with a
What is the expected accuracy of STEM mass measurements?
In an ideal biological specimen inaccuracy results from two factors: counting statistics and background subtraction.
Typical STEM beam current gives 1,000 electrons incident on each pixel (10A pixels with a dose of 10 el/A2). Of those,
5-10% are scattered. For a TMV segment 360A long and a measuring width of 220A, this would give ~80,000 electrons
scattered with a standard deviation (SD) of 280. However, roughly half the scattered electrons come from the TMV,
giving an uncertainty of 280/40,000=0.7%. Increasing the dose improves statistics but damages the specimen (2.5%
mass loss per 10 el/A2 of dose) and blurs the edge of the particle, requiring a wider measuring area. Smaller and/or
thinner objects give larger SD with a practical cutoff of ~20% for compact 30kDa proteins at 10el/A2.
The carbon substrate is not perfectly flat, as assumed when its contribution is subtracted from the total measured mass.
This is mainly important for single atom imaging at very high dose. The background program attempts to mask all particles and
measure background in the remaining “clean” areas. As discussed in Q3, any non-uniform residue between or underneath particles
can generate serious errors.
A practical approach to determine the overall accuracy of a series of mass measurements is discussed in Tutorial 8.
What is the expected spatial resolution with freeze-dried or negatively stained specimens?
Generally, freeze-dried biological specimens at -160ºC lose roughly 20A of resolution for every 10el/A2 of dose,
becoming “stable” at a width equal to their initial height. This is seen most clearly on the ends of TMV particles, one of
which is concave with very sharp corners. In a dose-response series, the TMV end gradually becomes completely rounded.
Negative stain gives higher resolution but very limited possibility for mass mapping. The stain forms a shell and penetrates
into crevices, preserving some of the shape even when the protein is vaporized. Nanovan (methylamine vanadate) gives the best
performance in STEM, but may disrupt some complexes. A layer of vanadate thicker than the particles of interest permits
determination of baseline between particles and rudimentary mass mapping. It is particularly useful for heavy atom cluster
labeling, since both the clusters and the fine structure of the protein are visible at
<5A resolution. Uranyl acetate gives higher contrast and seems gentler to specimens (even with the required pH<3),
but granulates rapidly with dose.
What causes "dirty" background? How does it affect data quality?
The problem usually comes from buffers or impure proteins. Bacterial contamination of any solution will give all manner
of secreted material. Sometimes a complex that is “stable” at high concentration (for crystallization) will release subunits
on dilution, giving a heterogeneous mass histogram and a poor background. In that case, brief fixation with glutaraldehyde
or other cross linker may solve the problem. We perform frequent TMV and water controls to assure the purity of substrates
and reagents used in our portion of specimen preparation.
Can STEM examine frozen-hydrated specimens?
This is possible but the thick background and lack of phase contrast are unfavorable for viewing biological specimens.
In a combined TEM/STEM, the STEM mode may be useful for finding heavy atom clusters. STEM tends to be sensitive to charging,
which may also be present with frozen-hydrated specimens.
Can other STEM instruments perform similar studies?
Links are provided on the STEM Home Page to facilities at Basel, Switzerland and other locations. In principle,
it should be possible to use TEM/STEM instruments and we are collaborating with Grigoriev’s group at Brandeis to test this.
Last Modified: June 12, 2009
Please forward all questions about this site to: