The dark current of a CCD camera arises from thermal generation of electrons and consequently thermally excited current in the detector during data acquisition [1,2]. Therefore, such EELS or EFTEM intensity is an artifact and should be subtracted for data correction. In general, as long as the CCD camera conditions stay constantly, the dark current is proportional to the acquisition time [3,4]. However, the dark current changes rapidly at the beginning of an experiment and taking 1 or 2 hours to stabilize. On the other hand, the CCD temperature must also be taken into account  since any temporal fluctuation in the CCD temperature may affect the signal characteristic of EELS spectra even though the camera is always cooled.
In summary, the dark current is dependent on the following factors:
i) CCD temperature,
ii) acquisition time. Therefore, one needs to calibrate it from time to time during an experiment. On the other hand, the characteristics of the
dark-current can be parameterized at individual pixels as a function of the image acquisition time.
iii) dark current increases significantly at the beginning of EELS acquisition due to temporal heating of the CCD
camera after the intense zero loss illumination.
current signal increases for longer acquisitions. This affects the data acquisition of low-signal images at high energy-loss range since it require long acquisition times. The signals in high energy-loss range are more susceptible to CCD conditions, e.g. the dark current.
Due to the facts above, refreshing the dark current several times is a useful CCD initialization process to be performed just before every
Each EEL spectrum can be corrected for dark current and gain variations between the detector elements of the CCD and for multiple scattering applying Fourier–log deconvolution. For accurate analysis, for each core-loss edge recorded, dark current spectrum should be collected. A simple way to remove dark current is to subtract a constant value from all spectra. Such value for the dark current can be obtained from a spectrum taken far away from the specimen in the vacuum with the same acquisition time for analyzing spectra.
 W. J. de Ruijter, Imaging properties and applications of slow-scan charge-coupled device cameras suitable for electron microscopy, Micron 26 (1995) 247–275.
 J.M. Zuo, Electron detection characteristicsof a slow-scan CCD camera,
imaging plates and film, and electron image restoration, Microscopy
Research and Technique 49 (2000) 245–268.
 I. Daberkow, K.H. Herrmann, L. Liu, W.D. Rau, Performance of electron image
converters with YAG single-crystal screen and CCD sensor, Ultramicroscopy
38 (1991) 215–223.
 K. Ishizuka, Analysis of electron image detection efficiency of slow-scan CCD
cameras, Ultramicroscopy 52 (1993) 7–20.
 M.M. Blouke, J.R. Janesick, T. Elliott, J.E. Hall, M.W. Cowens, P.J. May, Current status of the 800 800 charge-coupled-device image sensor, Optical Engineering 26 (1987) 864–874.