Ari D. Palczewski, SRF Scientist Current state of LCLS-II recipe development and ongoing research to maximization flux expulsion for the already in production cavity material Ari D. Palczewski, SRF Scientist Jefferson Lab, USA MOBA07

100% recrystallized. Uniform size and equal-axed grains LCLS-II Material Parameter Unit E-XFEL / LCLS II Year - 2007&2010 / 2014 RRR > 300 CHEMICAL COMPOSITION Ta % or weight ppm ≤ 0.05 % W ≤ 0.007% Ti ≤ 0.005% Fe ≤ 0.003% Si Mo Ni H2 ≤ 2 w-ppm N2 ≤ 10 w-ppm O2 C MICROSTRUCTURE 100% recrystallized. Uniform size and equal-axed grains Grain size ASTM µm Predominantly ≥ 6 ≤ 45 Maximal grain size > 4 < 90 MECHANICAL PROPERT. Tensile strength, Rm * N/mm2 > 140 Yield strength, Rp0.2 * 50 < Rp0.2 < 100 Elongation, AL 30 * % ≥ 30 Hardness, HV (min. load 10 N) ≤ 60 26 Total heat lots from Tokyo Denkai (TD) material - 3155 sheets 16 heat lots from Ningxia Orient Tantalum industry co ltd (OTIC or Ningxia or NX) - 2745 sheets Compiled by Jacek Sekutowicz DESY/LCLS-II

How is material specification defined/tested Parameter Unit E-XFEL / LCLS II Year - 2007&2010 / 2014 RRR > 300 CHEMICAL COMPOSITION Ta % or weight ppm ≤ 0.05 % W ≤ 0.007% Ti ≤ 0.005% Fe ≤ 0.003% Si Mo Ni H2 ≤ 2 w-ppm N2 ≤ 10 w-ppm O2 C MICROSTRUCTURE 100% recrystallized. Uniform size and equal-axed grains Grain size ASTM µm Predominantly ≥ 6 ≤ 45 Maximal grain size > 4 < 90 MECHANICAL PROPERT. Tensile strength, Rm * N/mm2 > 140 Yield strength, Rp0.2 * 50 < Rp0.2 < 100 Elongation, AL 30 * % ≥ 30 Hardness, HV (min. load 10 N) ≤ 60 Each sheet is tested for hardness – must be less than HV 60, if harder they are sent back to Manufacture for annealing again. The hardest and softest sheet from each lot then goes to analysis (Done by DESY) for LCLS-II. Lots are defined by heat lots, not by ingot - TD uses more than one ingot per heat lot while Ningxia uses one ingot per heat lot.

LCLS-II Material Distribution Gigi Ciovati - JLab plots TD distribution is more expected “Predominately ASTM 6.0” OTIC/NX in predominately “ASTM 5.8” but strange distribution

OTIC/NX Material sorting based on ASTM crystal size – this was a guess Gigi Ciovati - JLab plots “Lot A” ~600-700 from softest/largest crystal sheets (ASTM 5 to <6) “Lot B” ~400-500 from the middle sheets (ASTM 6 to>7) “Lot C” ~150-200 from the largest/smallest crystal sheets (ASTM 7 to 8) TD material was not sorted or actively tracked

Current status LCLS-II recipe by cavity vendor, material vendor, heat treatment/EP recipe by cavity counts

Comments on testing conditions – Magnetic field Cavities JLab FNAL DESY First 20 from RI (TD) First 8 from Zanon (NX) 1-2mG (Dewar 7-8) 1-2mG (Dewar 1) 3-5mG (Dewar 2) Cavity after first articles Through April 15th 2017 5-7mG (Dewar 7-8) 6mG max 3-5mG (D2) 3-10mG (D3) – mostly D2 1-10mG large Dewar with 4 cavities. (8TD and 4 NX-A 900 tested for JLab) Cavities after April 15th 2017 (JLab change) 3-5mG (Dewar 5, 7, 8) 6mG max Magnetic field at JLab can be controlled on all Dewar's in the axial position, perpendicular position is ~ 0-1.5mG D1 at FNAL had composition coils around the cavities to minimize the fields D5 @ JLab and D2 & D3 @ FNAL hold 3 cavities.

TD 140µm/800°C Two tested in ~1-3mG at JLab and the other two and 3-5mG at FNAL. 1-3mG too low for modual 3-5 about modual spec This recipe will not be good enough – this was a verification from single cell results at FNAL and JLab 800°C degassing temperature is not enough Gonnella SRF2017 thxa09

TD 200µm/900°C FNAL data all pass Gonnella SRF2017 thxa09

NX 200µm/900°C FNAL data all pass Gonnella SRF2017 thxa09

TD 200µm/900°C vs NX 200µm/900°C Most data tested in 3-7mG field depending on time, and location. Gonnella SRF2017 thxa09

NX-A 900 vs TD @ DESY Tested in 2 different Dewar's and two different test stands.

NX-A 900°C retest from DESY @ JLab Dewar 7 When tested at DESY the cavity had a Q0 of 1.9e10. Q0 change appears to be both magnetic field and cooling rate related changes – See JLab D5 data later.

Re-testing at JLab D5 (position 1) vs. D7 standard cool down Sorry for scales, 2.2e10 in Dewar 5 to 3.1e10 in Dewar 7

Testing at JLab D5 (three on one test stand) – standard cool down Position 3 is on the inlet and outlet side. Positions 1 and 2 are on far side. Cooldown data corrupt, but looks like cavity 3 cooling like D7/D8 Cavities 1/2 top and bottom cool at the same time, trapping all the flux

Testing at JLab D5 (three on one test stand) – modified cooldown Standard cooldown Dewar warm at 300K, ramp up time 20 minutes 0 to 8-10 grams per second. Modified cooldown Dewar and cavity thermalize at 100K, , ramp up time 10minutes 0 to 8-10 grams per second. Cooldown data corrupt, but looks like cavity 3 cooling like D7/D8 Cavities 1/2 top and bottom cool at the same time, trapping all the flux

D5 modified cooldown from 100K

RF D5 standard vs. modified 100K

NX-B 900°C vs. 950°C

NX-A 950°C from Vendor A vs. NX-A 900°C Vendor B

NX-C 900°C from Vendor A vs. NX-C 975°C Vendor B

Back-up

Reminder - Flux expulsion 800°C vs 900°C SRF2017 Tutorial July 13-15th Reminder - Flux expulsion 800°C vs 900°C Ningxia Lot A Flux expulsion baseline 800°C (JLAB data) Flux expulsion post 900°C 3 hours (FNAL data) Ningxia Lot C Flux expulsion baseline 800°C (FNAL data) Tokyo Denkai Lot B-C Tokyo Denkai Lot A Flux expulsion post 900°C 3 hours (JLAB data)* ΔT 2K 10% 46% 2% 5% 23% 69% 14% 71% 6K 21% 70% 15% 35% 30% 95% 26% 90% 10K 80% 20% 42% 34% 100% NA Absolute value between different labs and different test setups have not been fully calibrated, look at relative changes. Both TD material lots show excellent expulsion after 900°C NX-A material shows “ok” expulsion after 900°C, 950°C would maximize it. Lot “C” NX material shows minimal expulsion after 900°C, and acceptable after 975°C, 9 cell prototypes results later. *Different test setup than 800°C data (800°C D7 and 900°C D3) Table initially compiled by Anna Grassellino FNAL, modified here