In order to study the influence of a deposition parameter such as the electron beam current [Ie] in the microstructure and composition of the Co-based FEBID nanodeposits at the nanometer scale, two FEBID magnetic nanodeposits were fabricated at room temperature using a field emission gun scanning electron microscope electron column. The deposits were grown on an oxidized silicon wafer SiO2//Si substrate using a working voltage of 30 kV. In order to compare the effect of the working current Ie on the final metallic content, one of the deposits was grown at a low Ie (in picoampere range) and another one at a high Ie (in nanoampere range). In both cases, the Co2(CO)8 precursor gas was brought onto the substrate surface by means of a gas injection system and decomposed under the focused electron beam. Common parameters for this rectangular shape Co-based deposition process were the following: Co nanostructures with dimensions (width × length × thickness) = 0.5 × 1.0 × 0.2 μm3; Vol/dose = 5 × 10-4 μm3/nC; dwell time = 1 μs; beam overlap = 50%; refresh time = 0 s; base chamber pressure = 1 × 10-6 mbar; process chamber pressure = 4.3 × 10-6 mbar; scan strategy = bottom to top in serpentine mode; vertical distance between gas injection system needle and substrate = 135 μm; horizontal distance = 50 μm; and pitch = 2.21 nm (deposit 1, 0.044 nA), 13.16 nm (deposit 2, 2.4 nA).
Following the nanodeposit growth, in situ EDS analysis has been performed on them (deposit 1, Co:C:O 64:17:19; deposit 2, Co:C:O 93:5:2). Prior to the lamella preparation, the Co deposits were covered with a layer of FEBID-grown platinum and a second layer of focused ion beam induced deposition [FIBID]-grown platinum. This standard procedure was carried out to protect the deposit from the ion beam damage during lamellae preparation. The in situ lift-out and cross section TEM lamellas of the Co deposits have been fabricated using the focused ion beam present in the same equipment. The final thinning and polishing have been done at an ion beam acceleration voltage of 5 kV to decrease the amorphization layer. The final lamella thickness was lower than 50 nm.
The microstructure of the nanodeposits has been investigated by HRTEM, whose results were obtained using an image Cs-aberration-corrected FEI Titan Cubed at 300 kV (FEI Company, Hillsboro, OR, USA). The correction of the spherical aberration of the objective lens leads to a spatial resolution of at least 0.1 nm.
The composition of the nanodeposits at the nanometer scale has been investigated by means of spatially resolved chemical analysis, carried out in a STEM VG HB 501 with a field emission gun operated at 100 kV and fitted with a Gatan 666 spectrometer (Gatan Inc., Pleasanton, CA, USA), optically coupled to a CCD camera. Spatially resolved EELS analysis was used to investigate the metallic cobalt content and the oxidation state in each deposit. Thus, the electron beam is scanned on the sample, and a series of spectra is collected for each point; thus, the obtained spectra can be compared as a function of the point of collection in the sample. This technique is known as spectrum-line or line scan acquisition [13
]. For each line scan, spectra were acquired at steps of 1 nm, and then summed every five spectra for the calculation of intensity ratios of the Co L2,3
) and I
), respectively). I
) and I
) were calculated as the intensity maximum for each edge. For the analysis of chemical composition as a function of growth direction, 200 spectra were acquired for each point, realigned, and summed. Principal components analysis [PCA] was applied to each series of spectra to decrease experimental noise and so as to obtain a better signal to noise ratio [14
]. After applying PCA to each spectrum for a single point, five resulting consecutive spectra of a line scan were summed, and the intensities of the white lines were calculated after a power-law removal of the background and a linear fit below the lines. Therefore, the chemical state of Co has been first estimated by means of the intensity ratio of the L2
peaks. The reference values of I
), 0.31 for metallic cobalt and 0.27 for cobalt oxide [CoO] [15
], were calculated using the same technique.
On the other hand, the relative O/Co concentrations were also calculated, integrating their respective signal intensities from a series of 200 summed EELS spectra at a single point inside the deposit and dividing by their respective cross sections. An energy dispersion of 0.2 eV/channel was used for the analysis of the fine structure for each element, whereas an energy dispersion of 0.5 eV/channel was used for the quantification of the relative amounts of each element, with a collection angle of 24 mrad and a convergence angle of 7.5 mrad. Both types of experiments had an acquisition time of 0.8 s/spectrum.