The detection system of a scanning electron microscope (SEM) is just as important for image quality as the design of its electron optics [1]. Recent studies [2, 3] emphasize the importance of angle- and energy-filtering of backscattered electrons signal (BSE) to reveal details not usually observable in the integral BSE signal.

Here we introduce a novel detection system tested on a recently developed ultra-high resolution electron optical column [4, 5] that comprises three BSE detectors capable of angle- and energy-filtering. The system can be switched between a “high-signal” mode where no bias is applied to the filtering grid and an “energy-selective” mode where the filtering voltage can be tuned to achieve various degrees of signal separation.

High-pass energy filtering enables detection of BSEs which have undergone minimal energy loss. Their escape depth is limited and the resulting micrograph becomes surface sensitive. This is beneficial for samples with complex surface and sub-surface structure. Clear visualization of the topmost layer of the sample is critical e.g. for failure analysis of 3D semiconductor devices, given the continuously decreasing dimensions of the local structures in such devices.

Figure 1 shows a SEM micrograph of a semiconductor memory device cross-section acquired at 3 keV primary beam energy and 265 pA probe current. Contrast of the images decreases continuously with increasing bias of the filtering grid. With no bias applied the material contrast is highest, yet it contains information about both surface and subsurface area of the specimen. With increasing filtering bias only the electrons backscattered from the very surface of the sample are detected. This is in agreement with theoretical simulations for 3 keV primary beam in silicon that show an escape depth of 80 nm for the integral BSE signal and only 2 nm for the low-loss BSE signal.

References:

[1] B J Griffin et al, Proceedings of  the 18^th International Microscopy Congress (2014).

[2] I Müllerová et al, Material Transactions 48 (2007), p. 940.

[3] H Jaksch, Proceedings of  EMAS 2011 (2011), p. 255.

[4] J Jiruše et al, Ultramicroscopy 146 (2014), p. 27-32.

[5] J Jiruše et al, Proceedings of Microscopy and Mircoanalysis 2014, (2014).

Figures:

Figure 1. SEM micrographs of a FIB cross-section of a semiconductor memory device. The image shows metallic layers with through silicon vias (TSVs) acquired at 3 keV primary electron beam energy. Different values of filtering bias were applied to demonstrate increasing surface sensitivity of the imaging system. With no bias applied, the TSVs in the subsurface layers are clearly visible while at the highest bias only surface structures are resolved.

« Back to The 16th European Microscopy Congress 2016

EMC Abstracts - https://emc-proceedings.com/abstract/surface-sensitive-investigation-of-semiconductor-devices-with-a-signal-selective-sem-detection-system/