Fundamental Studies on SRF Materials – Interfacial Oxidative Chemistry of Niobium Progress Report July 2007 to June 2008 Lance Cooley SRF Materials Group Technical Division Fermilab Steven J. Sibener The James Franck Institute and Department of Chemistry The University of Chicago

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 Nb 2 O 5 : many polymorphs exist; most have octahedral Nb-O coordination an well-known feature of metal-oxide magnets such as manganates Surface morphology Few-nm thick layers, so all layers can “communicate” with underlying Nb superconductor Mismatch at interfaces, defects, etc. (“selvedge”) – “Layers” is too simple blue = Nb red = O Image taken from

BACKDROP: IIT/ANL/FNAL Discovery! Signatures of magnetism in SRF niobium at ~2K We suspect defects in the Nb 2 O 5-x since pure samples (e.g. Nb 12 O 29 ) 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, (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

LHe Cooled UHV Scanning Tunneling Microscope with XPS, Auger, ISS Reactive Gas-Surface Oxidation with O( 1 D) and O( 3 P) Unique Surface Instrumentation in the Sibener Group for Scattering, Imaging, & Interfacial Spectroscopy Inelastic He Atom Scattering for Determining Surface Forces of Clean/Oxidized Surfaces Beam Scattering with In Situ Vibrational & Electronic Spec

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

5-micron view of Niobium Polished at Cabot Microelectronics Atomic Force Microscopy Image of As-Received Sample Average roughness is <0.5 nm Imaging: Nataliya Yufa, Graduate Student at UChicago

Initial XPS -- Survey UC Grad Student: Miki Nakayama Nb 3d C 1s Nb 3p O 1s F 1s Co 2p

Overview - Nb XPS Spectroscopy Nb 3d peak positions from prior literature oxidation state (3d 3/2, 3d 5/2 ) (units in eV) Nb +5 (210.0, 207.3)Nb +4 (208.8, 206.0) Nb +2 (206.8, 204.0)Nb 0 (205.0, 202.2) (Nb 2 O 5 ) (NbO 2 ) (NbO) Thin Solid Films, 192 (1990) 351

XPS of Pristine Sample -- Nb 3d Reveals Nb 2 O 5 -dominated layer on top of metallic Nb for both samples Nb Nb 2 O 5

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 Nb Nb 2 O 5

Sputtering Effects on Single Grain Sample Nb 3d peaks gradually convert to NbO, suggesting massive rearrangement of oxygen. But why NbO? Nb 2 O 5 NbO The single-grain sample was sputtered with 2 keV Ar + at a sample current of 2.5  A in 5 or 15 minute increments

Heating Effects + Sputtering = NbO After 790 K heating, Nb 2 O 5 is gone, NbO remains and is thick enough to almost mask Nb underneath. Nb 2 O 5 apparently gives up its oxygen easily, either by annealing or by sputtering. This is not as true for NbO… 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 Nb 3dO 1s

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

Q-Drop of Solid Nb Cavities (Detlef Reschke, DESY; SRF 2007) What is a bake??? Open Air, Humidity, Pure Nitrogen, Vacuum, UHV??? Temperature of bake??? C 1-2 days; C hours Cooldown rate??? It matters! What occurs at the atomic level wrt O, H, Nb, & Interfacial Structure???

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) NbO Nb 2 O 5 Heating in Air - “Air of the Day”

Carbon and NbC: Sputtering Induced Chemistry C 1s peak shows that almost all of the graphite converted to NbC after the lower energy sputters C, graphiteNbC Carbon’s Role in Nb Interfacial Chemistry?

Key Questions! What is the nature of the clean Nb interface? What is the oxidation mechanism? Kinetics? Stability? How do other species, e.g. H 2 O in “air of the day”, affect the oxides ? H 2 ? How do crystal faces (100, 110, 111) and polycrystallinity (which includes grain boundaries) affect oxides? What is the role of Carbon and NbC? How does the oxide “communicate” w/ the bulk? What is the effect of baking/cooldown on the interface? How do different polishing or cleaning procedures affect the interfacial chemistry? Can we design an optimal chemical / mechanical polishing procedure? How does the above affect superconductivity?

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 Year 2: Emphasis on Precision Studies of Oxidation Mechanisms on Single Crystal and Technical Substrates Vibrational and STM studies will complement XPS Experiments to assess oxidation kinetics (including O 2, H 2, C, H 2 O), structure & defects, and correlations with physical properties Stability of the Interface? Passivation vs. Conditioning?

Backup slides

In Situ Vibrational Spectroscopy (HREELS) Gives Chemical View of Interface in Real-time During Oxidation Experiments: Example Ni(111) Synergistic Effects of Electron Irradiation During Ni Oxidation At Low Temperatures Surface Passivates with Chemisorbed Oxygen Overlayer Synergistic Influence of Electrons Produce Surface and Bulk Oxide

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, (2001) J. Chem. Phys. 115, (2001) Surf. Sci. Lett. 496, L29-L34 (2002) 160 sec200 sec240 sec 80 sec120 sec t=40 sec 840 Å 230 Å Zippering Event During Oxidation T xtal =465, 0.15 L O 2 STM Image of Zippering Event

Time-Lapse STM Movie of O-Induced Step Doubling Summary: View Surfaces as Dynamic Systems During Oxidative Chemistry