In order to achieve knowledge about the emission sites of photoemitted electrons, raster scans are performed for various voltage ratios Γ.
5.3.1 Two-dimensional raster scans
By raster scanning the laser focus over the emitter, maps of the effective electron emis-sion sites are recorded. Fig.5.11 shows raster scans of two different tips for various voltage ratios Γ. In this experiment, the overall intensity on the detector is counted and the resulting maps are normalized in Fig.5.11Aand B, respectively. The images in Fig.5.11Aare obtained using a blunt tungsten tip with a diameter of about 400 nm. In contrast, Fig.5.11Bshows scans for a sharp tip with 40 nm apex diameter. Blue regions in Fig.5.11Acorrespond to laser positions which do not result in a signal on the detector.
For a low value of Γ, mainly electrons extracted by illuminating the tip shaft are able to reach the detector. This area shifts towards the apex region for increasing Γ, which corresponds to a shift of the cutoff-point, as discussed in Sec.3.2. The combination of maps for different voltage settings results in an image of the tip itself, convoluted with the laser focus size. The lower limit for the voltage ratio Γ is mainly given by the onset of static field emission. The increase of the maximum signal towards high Γ-values is a result of the focusing effect and the associated clipping of the electron pattern at the detector (cf. Fig.5.7A).
In Fig.5.11B, a second contribution to the emission is visible. This signal is independent of Γ and remains at the same position, and corresponds to electron emission from the tip apex. Hence, it is only visible for sharp tips, as substantial field enhancement required to obtain a signal of comparable intensity to the shaft contribution. The intensity of the apex signal is almost independent of Γ. This is expected, as an increase in Γ only leads to a slight focusing of the electron trajectories (cf. 5.1).
Figure 5.11: Intensity maps of the electron yield gathered by raster scanning the laser focus over the emitter for various values of Γ. The maps are measured using a A: blunt (dapex ≈ 400 nm) and B: sharp tip (dapex ≈40 nm). The first sub-figures show SEM
images of the corresponding tips.
Figure 5.12: Electron signal as a function of voltage ratio Γ and laser spot posi-tion along the tip axis z. A, B: experimental data of two different tungsten tips. C:
Simulation results.
The combination of voltage setting and laser focus position enables the selection of an emission region from the macroscopic tip. It can be used to effectively separate apex from shaft emission. The separation of the two is important in order to achieve optimal performance in a UTEM, since the source size should be as low as possible while maintaining high electron currents and a short pulse duration. Electrons emitted from the tip shaft are undesired, as they drastically corrupt the source size, and hence increase the emittance.
5.3.2 Laser focus position vs. voltage ratio Γ
A distinction between shaft electron yield and the apex signal is again visible in Fig.5.12 (A and B are measurements, C depicts simulated data). Here, the z-axis corresponds to the laser position along the tip axis. In the orthogonal direction (y-axis in Fig.5.11), the laser focus is centered with respect to the tip emitter using the emitter map gained in Fig.5.11. Again, the apex contribution is clearly visible at the tip axis position (z= 0µm) and is not shifting upon changes in Γ. The shaft contribution, on the other hand, is evident as a diagonal line. The signal from shaft electrons is decreasing in intensity for decreasing values of Γ. This effect is, as explained before, a result from the spatially increasing pattern of the electron signal on the detector and the subsequent clipping due to the limited detector size (cf. Fig.5.7A).
At the apex position, if Γ is decreased, as discussed in5.2, the ring pattern (arc pattern for blunt tips) disappears from the detector, and a homogeneous electron distribution is observed. However, an intermediate region is found in the measured data, in which the ring is not observed anymore but the uniform pattern is not yet intense. This
intermediate region can be seen in Fig.5.12A for values between Γ = 1.1 and Γ = 1.05 at z = 0µm, and can be explained considering the field enhancement at the tip apex.
The static field-enhancement is suppressed at the field reversal point and subsequently increases for decreasing values of Γ. As a consequence, between the Γ value where the ring shape disappears from the detector and the onset of the homogeneous pattern, the field enhancement is not yet large enough to allow for the same intensity in electron emission as for lower voltage ratios.
A large agreement is given when comparing simulated (Fig.5.12C) with measured (Fig.5.12A, B) data. In particular, the apex and shaft signals follow the same general behavior. Only
the enhanced emission near the field reversal point is weaker in the simulations.
5.3.3 Double image
The second diagonal line in Fig.5.12A stems from a ghosting effect of the incident light at the sapphire window of the vacuum chamber (Fig.5.13). Assuming that the laser light impinges on the glass with an angle α, the distance L between the original transmitted light and a doubly reflected beam can be calculated using Snell’s law to
L= cos(α)·2·tan
The window thickness is d = 3 mm, and the refractive index of sapphire for a laser wavelength of 400 nm isn2 = 1.7865 [154]. Assuming an incident angle ofα= 1◦ results in a lateral distance of about 58µm (n1 set to unity). This is approximately the distance found in Fig.5.12Abetween the two diagonal signals (the distance in x-direction between the two features). For a given tip axis position, the second signal is visible for a lower Γ. Hence, the ghosting signal is shifted towards the tip shaft.
5.3.4 Benefit for the UTEM
Considering the alignment of an ultrafast TEM, large raster scans or scans of the voltage ratio are not feasible. However, locating the position of the apex with a single electron image is an involving task. Utilizing the features shown above, one can easily locate the laser focus position on the tip. In case the laser focus is at the tip apex position, the electron beam should not move upon change of the suppressor voltage (thus changing
Figure 5.13: Sketch of beam paths through a glass window illustrating the ghosting effect.
Γ), but instead (de-)focus. Furthermore, the intensity of the signal should remain almost constant during this process.
Near the field reversal setting (Γ≈1.2), the detected electrons originate from the apex itself and from the apex-near region of the tip shaft (cf. Sec.5.2.4). In this setting, the source size is increased due to the shaft contribution and therefore corrupts the spatial coherence (cf. Sec.2.4.3) and the electron pulse duration. Latter stems mainly from different path lengths between electrons starting at the tip apex and such starting at the tip shaft, resulting in a time-of-flight difference. Although this “high-Γ”-mode is not the favorable setting for high magnifications, it might still be useful for lower magnifications.
It is clearly visible in Fig.5.11B or Fig.5.12A, B, that the overall electron yield for a setting close to the field reversal is two to three orders of magnitude larger compared to electron emission from the apex for lower Γ. Hence, a setting with Γ ≈ 1.2 can be used as a high current low, magnification mode which can be useful for, e.g., finding the relevant structure for an investigation.