MerlinEM

Within each pixel are two counters, which can be used simultaneously to give gapless readout – whilst one counter reads, the other reads out. This unique approach is of particular importance for beam sensitive samples – there is no gap in time for readout so every part of the exposure to the sample is being read and recorded. This feature and the speed of the detector can be used to record gapless multiframe films of dynamic effects and it can be also used to collect diffraction sample tilt series by continuous rotation of the goniometer, without loss of any electron data to the read-out. Due to the electron counting approach of the detection system as well as the fully digital readout, MerlinEM adds zero noise allowing a Signal to Noise Ratio (SNR) within a single frame as high as 16.7 million to zero.

The ROI Rows function allows a dramatic increase in speed of readout. The user can select how many Rows the software will read from the detector. The default value, 256, means that the whole chip will be read out. Selecting a different value will result in a rectangle image of N x 256 pixels where N can be N ∈ {4, 8, 16, 32, 64, 128}. The position of the Rows is fixed to the inner edge of the detector due to the architecture of the chip.  This mode can be used to significantly speed up the STEM acquisition. This feature is unique to MerlinEM detectors.

Each of MerlinEM’s pixels contains >1100 transistors within the 55 micrometer pitch, enabling intelligent processes to occur on chip. Each incoming electron is evaluated, and if it is above a user specified threshold energy level, the signal is counted. If it is not, then it reads as zero. This unique feature allows for noise-free data collection. This is unique to hybrid pixel technology and strongly differentiates it from analogue integrating detectors, such as indirect CCD technology or CMOS.

The high signal to noise is not the only advantage when hardware electron counting is employed. With continuous imaging mode, our users can take one hundred 1 ms acquisitions (as an example) which should be exactly the same as one 100 ms long acquisition. This will allow the user to analyse how damage sensitive samples react to the electron beam and calculate the correct dose / frame time for your sample in a single exposure. In practice, summing the first 10 images will show exactly the same image as a 10 ms frame time, the first 20 images the same as a 20 ms frame time, etc. This is a very useful feature of the camera for radiation sensitive samples.