Gallium focused ion beam (Ga+-FIB) scanning electron microscopy (SEM), one of several types of volume electron microscopy (vEM), enables biologists to visualize the 3D ultrastructure of cells and tissues at a nearly isotropic spatial resolution of a few nanometers in specimens prepared by heavy atom staining and plastic embedding using electron beam energies of 1.5 keV and fluences greater than 104 e-/nm2. It has not previously been understood how such high fluences can be used without radiation-induced collapse of the specimen block. In addition, the z-resolution (perpendicular to block-face) in FIB-SEM is substantially better than predicted by Monte Carlo modeling [1]. Here, we show directly that Ga+ ion-milling in the FIB-SEM implants a thin layer of Ga atoms at the block surface, thereby improving the z-resolution by reducing the depth from which backscattered electrons are detected, creating a hardened metallic Ga/C glassy phase and avoiding collapse of the specimen block, while also preventing specimen charging. By imaging the sectioned FIB-milled block-face with annular dark-field (ADF) scanning transmission electron microscopy (STEM) at a primary beam energy of 300 keV, we have shown previously that there is a substantial increase in mass within a ∼20-nm thick layer under the block-face, which have suggested in attributable to a glassy Ga-rich phase containing between 30 and 50 atomic percent Ga. In the present study, we have used electron energy loss spectroscopy (EELS) in the 300 keV STEM, to confirm directly the presence of elemental Ga at the sectioned block-face (Figs. 1a and 1b). This analysis required an electron fluence of ∼108 e-/nm2 due to the relatively small Ga L2,3 edge scattering cross section. Despite loss of Ga from the analyzed region caused by the high electron fluence, and presumably resulting from the labile movement of Ga, which liquifies at 30 °C and lacks any stable carbides, the EELS image and spectrum nevertheless confirm a concentrated layer of gallium extending 20 or 30 nm into the block as shown in Fig. 1b and 1c. Whereas the maximum fluence in serial block face (SBF)-SEM is limited to ∼20 e-/nm2 due collapse of the specimen block, our EELS data shows that Ga+ ion implantation accounts for the radiation resistance of epoxy embedded specimen blocks in the Ga+ FIB-SEM with 1.5 keV electrons and fluences above 104 e-/nm2 due to the formation of radiation-resistant, and conducting glassy metallic phase. The higher density of the Ga-rich surface layer also restricts the depth from which backscattered electrons are collected and provides higher z-resolution [1,2]. Annular dark-field (ADF) STEM and STEM-EELS spectrum imaging of a section through a Ga+ FIB-milled block of epoxy resin cut perpendicular to the block-face. The block-face was prepared in a Zeiss Crossbeam 550 FIB-SEM, operating with an electron beam energy of 1.5 keV and with a Ga+ ion beam energy of 30 keV, and used for volume electron microscopy of biological specimens. a) Diagram indicating direction of the Ga+ beam, which generates the block-face, and the ∼200 nm thickness of the microtomed section. b) STEM-ADF image of a sectioned Ga+-milled block exhibits a sharp increase in signal within ∼20 nm of the block-face, which is attributed to a high level of Ga implantation [1]; STEM-EELS images from a section through another Ga+-milled block confirm that the surface layer contains Ga as well as carbon from the epoxy resin. c) Integrated background-subtracted energy loss spectrum in the region within 20 nm of the block-face shows a clear Ga L2,3 edge at an energy loss of 1115-1150 eV, as well as the much stronger C K edge signal. Implantation of Ga+ ions perpendicular to the block-face explains the radiation resistance of epoxy resin blocks in the FIB-SEM, as well as the nearly isotropic spatial resolution achieved in volume EM of biological specimens [1].
No takes yet. Share an insight, caveat, or question.
Leapman et al. (2026) studied this question.
Synapse has enriched 5 related papers. Consider them for comparative context: