3.2 Estimations of the AO Flux-Density generated by the AO Generator
In the present work, the well-known Kapton mass loss method for measuring AO flux densities is adopted. In addition, a self-assembled monolayer (SAM) of ethyl trimethyl siloxane (ETS) formed on an oxidized silicon wafer is exposed to the AO beam of the AO generator for testing the AO oxidative behavior. Both the changes in hydrophilicity by water contact angle measurements (WCA) and surface composition by X-ray photoelectron spectroscopy (XPS) are characterized as a function of AO exposures. The experimental details are included in SI.
Results and Discussion
The experimental results from the Kapton mass loss measurements and the WCA measurements are included in Figure S5 . In brief, the Kapton mass loss measurements confirm that for the injection of 300eV O+ at a flux density of 1×1015 ions cm-2 s-1 into the tube chamber of the AO generator, a broad AO beam with a flux density of 1.5 ×1016 atoms cm-2s-1 is detected at a distance of 6cm from the exit of the tube chamber. The Kapton mass loss measurements also confirm that the AO fluxes at the specimen-placement plane of 6cm from the exit plane are uniform with a change of less than 10% within the central area of 30cm in diameter. Since the tube chamber has a diameter of 30cm, the diameter of the broad AO beam at 6cm outside the tube exit is expected to be about 36cm, due to random scattering of AOs. Thus, the AO flux density right at the tube exit is expected to be 2.2 ×1016 atoms cm-2s-1 and the AO flux is 1.53×1019atoms s-1. Since the ion flux density of 1×1015 ions cm-2s-1 is injected into the tube, with an injection diameter of 20cm, the ion flux is 3.14 ×1017 ions s-1. Therefore, on the average, 48.7 AOs are generated by one 300eV O+ and roughly about 24 collision-induced dissociations of O2 are driven by each ion injection.
As for the exploitation of AOs, a drop in WCA of 20ois evidence with one second of the AO beam-pulse onto the C2H5-SAM. The surface oxidation is also revealed with the C1s XPS spectral analysis shown in Figure 5(a). In this set of data, the sample prior to any AO exposure should have a surface composition of SiO2 terminated by -C2H5 and should thus show a C 1s spectrum of those carbon atoms in -C2H5. The trace amounts of spectral signals of C-O, C=O and O-C-O chemical species are attributed to the practically inevitable adsorption of residual carbonaceous contaminants such as lubricating oils in the non-perfect ultrahigh vacuum environment during XPS analyses. A comparison of the spectral data with and without an AO dosage of 1.5×1016 atoms cm-2 clearly shows a significant increase of the amounts of C-O, C=O and O-C-O chemical species on the AO-exposed sample. Thus, the carbonaceous contamination merely compromises but does not mask out the detection of AO-oxidation of the -C2H5 chemical group. Nevertheless, a quantitative analysis of the degree of oxidation of the -C2H5 chemical group is impossible due to this carbonaceous contamination. The drop in WCA of 20o, however, firms up the effectiveness of AO surface oxidation of a hydrocarbon surface/film.
Atomic layer deposition (ALD) of aluminum oxide is widely used as a protective encapsulant. Typically, one cycle of such an ALD process comprises a half cycle of TMA adsorption and a subsequent half cycle of TMA oxidation. Commonly, a flux of water molecules is used to complete the half cycle of TMA oxidation within a fraction of a second, but the complete purge of the residual water molecules is slow and typically requires a duration of 10 seconds or more due to the high surface sticking efficiency of water. In the present work, the flux density of 1.5×1016 AOs cm-2s-1 is expected to be sufficient for the completion of TMA oxidation in less than one second and the purge of AOs is virtually instantaneous. A trial run of ALD, with one second of TMA exposure followed by one second pulse of AOs, for 50 cycles indeed yields the deposition of about 7nm of aluminum oxide on a copper disc. The XPS data summarized in Figure 5(b) confirms the fast deposition of aluminum oxide at room temperature with AO-assisted ALD. The effect on the passivation of a copper disc against thermal oxidation is demonstrated by the experimental evidence summarized in Figure S5 .
Conclusion
In the present work, AIMD simulations offer clear atomistic snapshots of collision cascades induced by hiring O+ with a kinetic energy of 0-300eV to O2. Dissociative collisions of O2 and generation of AOs are evident, and the dissociation threshold requirement of the O+ kinetic energy can be as low as 14eV which critically depends on the exact collision configuration. Through the examination of some 200 collision configurations, a comprehensive picture of collision-induced O2 by hyperthermal O+ is obtained. The database is then used to facilitate DSMC simulations of AO production in the practical experimental design of a high-flux AO generator equipped with an ECR oxygen ion source powered to 800W for the extraction of an ion flux density of 1×1015 ions cm-2s-1 with an extraction area of about 300cm2, and with the ions fired into a drift tube (30cm in diameter) containing O2 at 0.14Pa. In essence, DSMC simulations confirm each 300eV O+ bullet initiates cascades of collisions with some of them having collision energy higher than the dissociation thresholds estimated by AIMD simulations. As such, a microscopic mechanistic portrait of AO generation is yielded. Subsequent experimental AO flux density measurements by the well-known Kapton erosion method gives a quantitative estimate of 49 AOs generated by one 300eV O+ injection. Since the dissociation energy of O2 is about 5.1eV, on the average, about 122eV of the kinetic energy carried by each O+ bullet is used to drive collision-induced dissociations of 24 O2, with the rest of input energy raising the kinetic energy and vibrational/rotational energy of each collision partners to an average of roughly 0.18eV which is significantly higher than their average energy in thermal equilibrium at room temperature. As such, the experimental AO broad-beam with a beam-size of 1000cm2and a record-high flux density of 1.5×1016 atoms cm-2 s-1 bears an average energy of 0.18eV for each AO or an AO temperature of slightly higher than 1100oC. When a self-assembled molecular layer of ethyl trimethyl siloxane on a silicon wafer is exposed to this AO beam for one second, the water contact angle drops by 20o and XPS surface analysis also confirms the surface oxidation of the hydrocarbon. In addition, the same AO-beam has also been demonstrated to facilitate the atomic layer deposition of Al2O3with a record-fast growth rate of 7nm per 100 seconds. In comparison, the typical growth with water vapor to replace AO as the oxidizing agent has a typical growth rate of 7nm per 500-1000 seconds. The present work thus articulates the science and technology of practically generating a high-flux broad-beam of AOs and demonstrates some applications of such a technology innovation.