Overcoming the Physical Limitations of Electron Detection

In early microscopes employing visible light, the human eye was the detector – translating the observation into an image – and the human being was the data acquisition system – recording the image in a drawing.  With the advent of photographic film, the vagaries of the human as detector were supplanted by electronic detectors. Film has remarkable properties – grains can be sub-micron in size, and the sheet of film can be arbitrarily large – but the time from snapshot to picture is rather long, thus limiting the utility of film. As all of us with “digital” cameras, based at first on CCDs (Charge-Coupled Devices) and now on CMOS APS (Active Pixel Sensors) know, the speed, convenience (and now performance) of these electronic detectors have all but obviated film. Electron microscopes are capable of extraordinary spatial resolution, and tremendous sensitivity because of the large interaction cross-section of electrons with matter. For a perfect microscope, the spatial resolution, , which can be achieved is  where SNR is the Signal-to-Noise ratio of the detector (how perfectly does the detector detect an incident probe particle), C is the contrast of the specimen (how much is the probe modulated by the sample) and D is the “dose” (number of probe particles per unit area).  The contrast is a property of the specimen, so improving resolution (smaller d) requires either improving SNR or increasing the dose.  Many specimens, particularly soft (biological) matter, become damaged (lose contrast) after a certain dose.  In addition, since  , a factor of two improvement (in one spatial dimension) requires a factor of 4 increase in dose. A transformational change in electron microscopy became possible with new detectors, based on CMOS APS, which dramatically improve SNR.  These detectors, are much more sensitive than previous electronic detectors, and can operate at speeds 3 orders of magnitude faster than previous detectors.  This means that one now takes movies rather than pictures: enabling us to see materials undergoing changes in real time, at atomic resolution, and taking out the electron beam-induced blur associated with biological samples. This presentation will discuss how these new detectors are designed, together with the physical limitations on electron detection (and what might be done to get close to physical limits).