Biology Department Biology Department  

Scanning Transmission Electron Microscopy Facility


Accuracy of STEM Mass Measurements

Expected Accuracy of STEM Mass Measurements
   Overall precision of STEM mass measurements is determined by counting statistics (random), specimen quality and analysis strategy. PCMass software and internal TMV controls are intended to provide a clear understanding of the strengths and weaknesses of data quality and efficient means to extract reliable numbers.

Random Factors
   Repeated measurements of the same particle or filament (neglecting damage) will give different answers due to the finite number of electrons detected in the measurement. If N electrons are detected, the standard deviation (SD) will be
SD=sqrt(N). Furthermore, the process of mass measurement involves integration (sum) over pixels within a boundary containing the object of interest minus an estimate of what the sum would be for an equivalent number of pixels containing only background.

  Choice of radius of integration is important because too small a radius may cut off part of the mass while too large a radius adds pixels with only background, increasing the total number of counts without adding signal. If the sample is crowded, a large radius may include parts of adjacent objects. A plot of mass vs. radius of integration for an isolated particle with a well-defined edge on a clean background should increase rapidly, plateau, then oscillate with increasing amplitude. PCMass makes it easy to study this in several ways (see Tutorial 1). The on-screen measuring circle or rectangle shows what pixels are being included in the measurement. The zoom display shows the integrated mass and a trace of average signal as a function of radius of integration (changed by T/Y keys). The numerical values can be viewed by striking the 'TAB' key.

   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.

Carbon Film Noise
   The carbon substrate is not perfectly flat, as assumed when its contribution is subtracted from the total measured mass. Any single particle may be situated on an area thicker or thinner than average. Moving the identical particle to another area would result in a different random offset in mass.

   This is mainly important for single atom imaging at very high dose. At normal dose the carbon film noise is negligible compared to counting statistics. 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.

 

Top of Page

Last Modified: June 12, 2009
Please forward all questions about this site to: Kathy Folkers


Visit our Anniversary Website One of ten national laboratories overseen and primarily funded by the Office of Science of the U.S. Department of Energy (DOE), Brookhaven National Laboratory conducts research in the physical, biomedical, and environmental sciences, as well as in energy technologies and national security. Brookhaven Lab also builds and operates major scientific facilities available to university, industry and government researchers. Brookhaven is operated and managed for DOE's Office of Science by Brookhaven Science Associates, a limited-liability company founded by the Research Foundation for the State University of New York on behalf of Stony Brook University, the largest academic user of Laboratory facilities, and Battelle, a nonprofit, applied science and technology organization.

Privacy and Security Notice  | Contact Web Services for help