Extreme uv photon and hydrogen radical interaction with graphene and ruthenium surfaces
Baibhav Mund, a former PhD student in the Industrial Focus Group XUV Optics defended his thesis Thursday November 22nd 2018. His supervisor was prof.dr. F. Bijkerk from the Faculty of Science and Technology.
Graphene, a covalently bound, hexagonal 2D network of carbon atoms, is generally reported to be a chemically inert material with a high mechanical strength. Due to its high optical transmission, graphene is a great candidate as an optical material in the EUV to soft X-ray wavelength range. In this thesis, the chemical and physical interactions of both single-layer and multi-layer graphene is experimentally investigated, in order to verify and corroborate grapheneâ€™s viability for optical uses.
Firstly, oxidation studies on single-layer graphene (SLG) are done. SLG is transferred from the Cu surface on which the layer was grown, to an amorphous Si surface. SLG is shown to oxidize when dosed with H2O and exposed to EUV radiation in the presence of H2. This oxidation signifies degradation of the aromatic layer of graphene in the form of keto-enol groups, such as enol of forms of 1,3 di-ketone, most likely occurring at point defects and/or grain boundaries. Experimentally, this is noted by in-situ reflection-absorption infrared spectroscopy (RAIRS) which indicates that ring structure degradation occurs through graphene oxidation by reactive oxygen species. These are generated through interaction of EUV with adsorbed H2O. Also, XPS shows an increase in the amount of C bound to oxygen and also an increase in formation of sp3 bonds, indeed confirming that the surface is oxidized, leading to graphene being more defective. Higher pressures of hydrogen enhance the oxidation of graphene, most likely by creation of defects by EUV generated H radicals, which subsequently can oxidize under exposure to EUV and H2O. This study shows that a low defect density of SLG is paramount for optical use, as a balance between oxidation and reduction reactions on a defective graphene surface cannot be achieved.
Next, the oxidation of transferred and as-grown multi-layer graphene was investigated in the presence of EUV radiation and a H2 background. In this case, as-grown MLG is shown to be resistant to oxidation when dosed with H2O and exposed to EUV radiation in the presence of even up to 1 Ã— 10-3 mbar of H2. Most likely, this is due to the morphology of MLG, which conforms to the grain boundaries of the surface it is grown on, which in this case is Mo. Additionally, secondary electrons generated in the underlying Mo substrate (which has a higher secondary electron yield than carbon) have a lower mean free path than the thickness of MLG, and therefore cannot reach the adsorbed water layer on top the MLG surface. Therefore, the only way for EUV-induced chemistry to take place would be through either EUV photons, or photo-electrons generated from graphene.
In contrast, transferred MLG (on an aSi surface) is shown to oxidize and form keto-enol groups (as observed for transferred SLG). However, XPS shows that this formation does not occur in the top-most layers of the graphene stack, and most likely takes place at the interface between the graphene stack and the surface onto which the graphene was transferred (aSi). SEM images show that graphene does not follow the morphology of the underlying aSi surface, but retains the morphology of the original Mo film. Consequently, during dosing water gets trapped between the aSi and the MLG, which most likely leads to graphene being oxidized when irradiated with EUV photons. Furthermore, SEM images show that MLG contains holes of <200 nm diameter, and image analysis indicates that small holes (< 50 nm diameter) are shown to increase in number after exposure to EUV. As such, transferred MLG is shown to be more susceptible to oxidation by formation of keto-enol groups when compared to as-grown MLG.
Multi-layer graphene is further investigated as a diffusion barrier for atomic hydrogen and compared to sputter-deposited amorphous carbon of similar thickness. In this case, MLG was transferred onto a Ru capped Y film, and spectroscopic ellipsometry was used to probe hydrogenation of the Y film to YH2 and YH3. By comparing the in-situ ellipsometry measurements acquired during atomic H exposure of a reference Ru/Y sample with carbon and graphene covered Ru/Y samples, it was demonstrated that carbon and graphene both act as barrier against diffusion of hydrogen to underlying layers. However, since the exposure of graphene and carbon to atomic hydrogen leads to etching of carbon, eventually the carbon/graphene is removed, after which the Y film does hydrogenate. The experiments demonstrated that etching of graphene by atomic hydrogen is considerably slower than etching of an amorphous carbon film of similar thickness. This investigation shows that when the surface is protected against direct exposure to atomic hydrogen, MLG can be used as barrier against diffusion of atomic hydrogen.
As a final sub-study, Ru was investigated as a protective material with respect to photochemistry of carbon and oxygen containing molecules, notably CO2. A Ru(0001) surface was dosed with CO2 and exposed to EUV in the presence of H2. At an H2 background pressure of at least 1 Ã— 10-5 mbar, a combination of Ru-catalyzed and EUV-induced reactions of CO2 is shown to lead to the formation of H2O and amorphous carbon on the Ru(0001) surface. The reaction pathway is shown to proceed via partial dissociation of CO2 into CO and O, due to the catalytic activity of the Ru surface, instead of by photo-chemical processes. Due to photochemical reactions and EUV-generated hydrogen radicals, CO is further dissociated to amorphous carbon and H2O is produced. The formation of amorphous carbon deactivates Ru surface sites and therefore reduces the partial dissociation of CO2. This was investigated by repetitive cycles of CO2 dosing on the Ru surface without cleaning the formed amorphous carbon. This CO2 dose cycle leads to a decrease in water formation, indicating that EUV-induced CO2 dissociation is either negligible or does not lead to any detectable reaction end-products. Furthermore, it was demonstrated that reduction of ruthenium oxide by EUV generated hydrogen radicals proceeds at relatively low pressure, where carbon cleaning is still not effective.