Why Do Subtle Changes in SRF Cavity Treatments Produce Profound Changes in Performance? Niobium cavities (melting point 2000 °C) – a bake at 150°C for a few hours can transform a good cavity into an ILC-qualifier or even a record-setter – why?? - Any changes must be subtle
- Any changes must happen at the surface
- Oxygen is the most likely actor
The Sibener Group has great expertise in surface and interfacial chemistry, including a detailed understanding of the atomic-level aspects of metallic oxidation
Review of Structure Crystal structures of Nb and some of its oxides - Nb: body centered cubic
- NbO: rock salt structure
- Nb2O5: many polymorphs exist; most have octahedral Nb-O coordination
- an well-known feature of metal-oxide magnets such as manganates
BACKDROP: IIT/ANL/FNAL Discovery! Signatures of magnetism in SRF niobium at ~2K We suspect defects in the Nb2O5-x since pure samples (e.g. Nb12O29) are known to have various magnetic phases at 2-5 K (Cava 1991) Magnetism is anti-superconductivity, hence this is a (the first?) plausible reason why subtle changes in the oxide would have profound effects on performance - Thomas Proslier et al., IIT, Appl. Phys. Lett. 92, 212505 (2008)
That work cross-fertilized to ANL work led by Pellin to coat cavities with aluminum oxide. Pellin’s work has been successful! We now have a recipe for ameliorating the oxide layer entirely! - Proslier et al, APL submitted
- Pellin, Norem et al., report on coated-cavity tests to FNAL 28 May 2008
Year 1 UC/FNAL result: An explanation of the baking effect? FNAL asked UC to look at “real” niobium first, instead of crystals as planned, in view of these developments Cabot Microelectronics – niobium polished to atomic flatness The pentoxide DEGRADES EASILY when attacked by a mild ion beam. Baking the native oxide TRANSFORMS it IRREVERSIBLY into a TOUGH layer with different bonding configuration (NbO). Tentative conclusion: The oxide that forms at room temperature (e.g. after etching) is not desirable because it contains defects that spawn magnetism. Fortunately baking heals these defects (but for how long?). Details are not known; now (year 2) we need to explore more ideal systems (e.g. crystals) to understand
Growth Possibilities Enabled by This Seed Project submitted to University-Based Research for the ILC (George Gollin, UIUC – lead) but proposal to NSF cancelled by Omnibus - Nonetheless, it alerted us to possible synergisms
Regional center opportunity (FNAL/ANL/UC/IIT/NWU): - Themes revolve around niobium oxidation science
- Full-scale non-acid polishing of cavities in new Fermilab tumbling machine using Cabot slurries
- Capping of post-tumbled surfaces using ALD to prevent magnetic oxide
- Understand mechanism of polishing (which involves embrittlement by oxygen injection) to optimize slurry
- Understand structure-property relationships of surfaces
- Long term: consider new accelerator possibilities of coatings, other materials topics (carbides…), mitigation of field emission
- Possible new members: FNAL-APC, NIU-CADD, UIC
Year 1 Accomplishments Focus: Polished polycrystalline Nb from SRF cavity stock; also an unpolished piece cut from a single Nb grain Surface and interface chemistry using new X-ray Photoemission Spectroscopy (XPS) tools Ion sputtering & annealing in UHV to simulate cavity etching and baking under controlled conditions Heating in the “air of the day” for studies of the “baking effect” Instrumentation development for infrared and STM studies
Initial XPS -- Survey UC Grad Student: Miki Nakayama
Overview - Nb XPS Spectroscopy - oxidation state (3d3/2, 3d5/2) (units in eV)
- Nb+5 (210.0, 207.3) Nb+4 (208.8, 206.0)
- Nb+2 (206.8, 204.0) Nb0 (205.0, 202.2)
XPS of Pristine Sample -- Nb 3d Reveals Nb2O5-dominated layer on top of metallic Nb for both samples
Heating Effects Mild Changes at Modest Temperatures Oxygen diffusion into Selvedge and Interstitials Broadening of the Nb 3d peaks with initial heating (522K) clearly indicates that oxygen is mobile, even at this temperature
Sputtering Effects on Single Grain Sample Nb 3d peaks gradually convert to NbO, suggesting massive rearrangement of oxygen. But why NbO?
Heating Effects + Sputtering = NbO After 790 K heating, Nb2O5 is gone, NbO remains and is thick enough to almost mask Nb underneath. Nb2O5 apparently gives up its oxygen easily, either by annealing or by sputtering. This is not as true for NbO…
Order of Sputtering & Heating: Notable Stability of NbO For Nb 3d and O 1s, we see that the end result is the same both ways, but heating after a sputter has no effect That is, once the transformation to NbO happens, it remains stable
Exposure to Air: Return of Other Oxides Nb 3d peaks show that NbO is still the dominant oxide species after air exposure, but higher oxides are forming The third peak does not match Nb+4 nor Nb+5, indicating a mixture of the higher oxides
1 hr 140 C Nb 3d peaks show that more of the Nb are converting from NbO to higher oxides; O being added to sample (in UHV, oxygen tends to be subtracted from the surface and move into niobium metal)
Carbon and NbC: Sputtering Induced Chemistry C 1s peak shows that almost all of the graphite converted to NbC after the lower energy sputters
Key Questions! What is the nature of the clean Nb interface? What is the oxidation mechanism? Kinetics? Stability? How do other species, e.g. H2O in “air of the day”, affect the oxides ? H2 ? How do crystal faces (100, 110, 111) and polycrystallinity (which includes grain boundaries) affect oxides? What is the role of Carbon and NbC?
Year 1 Summary and Year 2 Plans Year 1: A Good Start! Strong Interactions Between UChicago and FNAL Emphasize the Strength of Joint Efforts on this Project Real effects observed on SRF cavity niobium, and correlation of effects with changes in gap (through IIT/ANL work) were found FNAL Supplied and Cabot Polished Samples used at UC for Oxide Studies We have begun to explore the efficacy that different heating, sputtering, and polishing preparations have on the quality of the interface
Backup slides
Interfacial Dynamics & Oxygen Driven Reconstruction of a Stepped Metallic Surface Via Time-Lapse Scanning Tunneling Microscopy T.P. Pearl and S.J. Sibener J. Phys. Chem. B105, 6300-6306 (2001) J. Chem. Phys. 115, 1916-1927 (2001) Surf. Sci. Lett. 496, L29-L34 (2002)
Dostları ilə paylaş: |