Delicacy in experimental study of NEG coated vacuum chamber

1. Delicacy in experimental study of NEG coated vacuum chamber Dr. Oleg B. Malyshev Accelerator Science and Technology Centre CCLRC Daresbury Laboratory 45th IUVSTA Workshop on NEG coatings for Particle Accelerators and Vacuum Systems

2. Introduction Two concepts of the ideal vacuum chamber: Traditional: • surface which outgasses as little as possible (‘nil’ ideally) • surface which does not pump otherwise that surface is contaminated over time Result in • Surface cleaning, conditioning, coatings • Separate pumps • Baking in situ New (NEG coated surface) • surface which outgasses as little as possible (‘nil’ ideally) • a surface which does pump, however, will not be contaminated due to a very low outgassing rate Result in • NEG coated surface • There should be no un-coated parts • Activating (baking) in situ • Pump for CxHy is required 45th IUVSTA Workshop on NEG coatings Oleg Malyshev for Particle Accelerators and Vacuum Systems

3. Introduction Studying the vacuum properties: Non-pumping surface outgassing: • surface thermal, electron, photon, ion, positron, etc. stimulated desorption • surface does not pump, therefore there are no this uncertainty Vacuum pump: • ISO-standard pumping speed dome • Large pumping capacities (larger than gauge outgassing/pumping) NEG coated surface outgassing: • surface thermal, electron, photon, ion, positron, etc. stimulated desorption • surface does pump: • this affect the sensitivity • pumping speed may change NEG coating pumping properties: • No standard procedure as it is not a pump, it is surface with pumping properties • Low pumping capacity 45th IUVSTA Workshop on NEG coatings Oleg Malyshev for Particle Accelerators and Vacuum Systems

4. Introduction • The NEG coated vacuum chamber is a new subject in vacuum science and all traditional technique of study must be reviewed • Main question are • What and where exactly was measured • How the used tools affect the studies subject • How to interpret the measured data 45th IUVSTA Workshop on NEG coatings Oleg Malyshev for Particle Accelerators and Vacuum Systems

5. What we can measure • Pressure, P • Pressure gauges are the basic tool (or even the only tool) • It is necessary to consider that an accuracy may change with pressure level and time and depends on the gauge history • Pressure is measured inside the pumping port or in the near environment of the gauge, Pgauge, not in the location where you need to know pressure Ptest. • All UHV/XHV gauges pump and/or desorb – i.e. they change in the location: pressure without a gauge may significantly differ from the pressure in the same place if a gauge was installed. Pgauge Ptest 45th IUVSTA Workshop on NEG coatings Oleg Malyshev for Particle Accelerators and Vacuum Systems

6. What we can measure Gas Injection flux • continues expansion method – primary method: • by pressure change dP/dt in known (measured) volume V P V Q 45th IUVSTA Workshop on NEG coatings Oleg Malyshev for Particle Accelerators and Vacuum Systems

7. What we can measure Conductance, c, – secondary method: • pressure difference between two vessels • known (measured) gas flow P1 P2 V c Q 45th IUVSTA Workshop on NEG coatings Oleg Malyshev for Particle Accelerators and Vacuum Systems

8. What we can measure Gas flux between two vessels – secondary method: • conductance method (conductance must be measured earlier) P1 P2 V c Q 45th IUVSTA Workshop on NEG coatings Oleg Malyshev for Particle Accelerators and Vacuum Systems

9. Traditional outgassing measurements P1 P2 c Test sample • All desorbed molecules move to the pump • Outgassing flux can be easily measured by the conductance method • Errors:  P1, P2, c • ougassing of the ‘green’ vessel and gauge P1 • Pumping of gauge P1 Pump 45th IUVSTA Workshop on NEG coatings Oleg Malyshev for Particle Accelerators and Vacuum Systems

10. Open geometry traditional PSD, ESD, ISD outgassing measurements Test sample P1 P2 • The same methods used for photon, electron and ion stimulated desorption measurements •  is a desorption yield •  is a photon, electron or ion flux c  Pump 45th IUVSTA Workshop on NEG coatings Oleg Malyshev for Particle Accelerators and Vacuum Systems

11. Close geometry traditional PSD, ESD, ISD outgassing measurements For testing the prototypes of the accelerator vacuum chamber P1 P2 Test Chamber c Pump • Errors: • P1, P2, c, molecular beaming effect • Outgassing of the ‘green’ vessel and gauge P1 • Pumping of gauge P1 45th IUVSTA Workshop on NEG coatings Oleg Malyshev for Particle Accelerators and Vacuum Systems

12. NEG coated sample • What kind of sample • How to measure 45th IUVSTA Workshop on NEG coatings Oleg Malyshev for Particle Accelerators and Vacuum Systems

13. Problems in the NEG studies/measurements • Activation procedure – poisoning during activation • Low pumping capacity: • The coating might be poisoned by pumped gases before the measurements started • Sources of gas: • Gas injection • Induced gas desorption from the NEG coated surface • Induced and thermal desorption from uncoated parts • Gauges desorption • Gas molecule cracking in gauges • Molecular recombination on the NEG surface • Low pumping speed: gauge pumping speed might affect all measurements • Low outgassing/desorption: difficult to measure 45th IUVSTA Workshop on NEG coatings Oleg Malyshev for Particle Accelerators and Vacuum Systems

14. NEG coating in the open geometry set-up: pumping speed test • The main question is how to activated the NEG film: • The installation bake-out keeping the NEG film below the activation temperature • Baking the NEG film to activation temperature without temperature increase of uncoated parts • Pressure during activation (attention to H2O, CO and CO2) • Outgassing of gate valves • Limited pumping speed through the conductance c and right-angled valves P1 P2 gas c Test sample with NEG coating Pump 45th IUVSTA Workshop on NEG coatings Oleg Malyshev for Particle Accelerators and Vacuum Systems

15. NEG coating in the open geometry set-up: pumping speed test If pressure during activation is 10-9 mbar, then the amount of molecules hitting the wall is an equivalent of If pressure of NEG-sorbing gases (CO, CO2, H2O) during activation P > ~10-10 mbar => • the NEG film is continuously poisoning by these gases => • the activation is not full 45th IUVSTA Workshop on NEG coatings Oleg Malyshev for Particle Accelerators and Vacuum Systems

16. Reducing of CO, CO2 and H2O pressure in the open geometry set-up: • Increasing the effective pumping speed of the pump -> limited • Outgassing of gate valves • Limited pumping speed through the conductance c and right-angled valves • Long bake-out of the set-up (uncoated parts) before the activation • Back side of the test sample -> P1 P2 gas c Test sample with NEG coating Pump 45th IUVSTA Workshop on NEG coatings Oleg Malyshev for Particle Accelerators and Vacuum Systems

17. Reducing of CO, CO2 and H2O pressure in the open geometry set-up: gas • Increasing the effective pumping speed of the pump -> limited • Outgassing of gate valves • Limited pumping speed through the conductance c and right-angled valves • Long bake-out of the set-up (uncoated parts) before the activation • Back side of the test sample -> • NEG coated P1 P2 c Test sample with NEG coating Pump 45th IUVSTA Workshop on NEG coatings Oleg Malyshev for Particle Accelerators and Vacuum Systems

18. Reducing of CO, CO2 and H2O pressure in the open geometry set-up: gas • Increasing the effective pumping speed of the pump -> limited • Outgassing of gate valves • Limited pumping speed through the conductance c and right-angled valves • Long bake-out of the set-up (uncoated parts) before the activation • Back side of the test sample -> • Either NEG coated • Or outside the vacuum chamber P1 P2 c Test sample with NEG coating Pump 45th IUVSTA Workshop on NEG coatings Oleg Malyshev for Particle Accelerators and Vacuum Systems

19. Reducing of CO, CO2 and H2O pressure in the open geometry set-up: Cooling channel gas • Increasing the effective pumping speed of the pump -> limited • Outgassing of gate valves • Limited pumping speed through the conductance c and right-angled valves • Long bake-out of the set-up (uncoated parts) before the activation • Back side of the test sample -> • Either NEG coated • Or outside the vacuum chamber • Cooling channel P1 P2 c Test sample with NEG coating Pump 45th IUVSTA Workshop on NEG coatings Oleg Malyshev for Particle Accelerators and Vacuum Systems

20. Reducing of CO, CO2 and H2O pressure in the open geometry set-up: • There is an area where temperature changes from the temperature of the NEG coated sample TNEG to the temperature of the rest of vacuum chamber TVC: • During the set-up bake-out this are is under-baked • During the NEG activation this area temperature is higher than TVC and outgases. It might be the main source of gas. Area with transitional temperature Cooling channel TNEG TVC Test sample with NEG coating 45th IUVSTA Workshop on NEG coatings Oleg Malyshev for Particle Accelerators and Vacuum Systems

21. Reducing of CO, CO2 and H2O pressure in the open geometry set-up: gas • Increasing the effective pumping speed of the pump -> limited • Outgassing of gate valves • Limited pumping speed through the conductance c and right-angled valves • Long bake-out of the set-up (uncoated parts) before the activation • Back side of the test sample -> • Either NEG coated • Or outside the vacuum chamber • Cooling channel • Increase the coated area and decrease uncoated area P1 P2 c Pump => NEG coated chamber 45th IUVSTA Workshop on NEG coatings Oleg Malyshev for Particle Accelerators and Vacuum Systems

22. Reducing of CO, CO2 and H2O pressure in the open geometry set-up: gas • Increasing the effective pumping speed of the pump -> limited • Outgassing of gate valves • Limited pumping speed through the conductance c and right-angled valves • Long bake-out of the set-up (uncoated parts) before the activation • Back side of the test sample -> • Either NEG coated • Or outside the vacuum chamber • Cooling channel • Increase the coated area and decrease uncoated area P1 P2 c Pump => NEG coated chamber ≡close geometry 45th IUVSTA Workshop on NEG coatings Oleg Malyshev for Particle Accelerators and Vacuum Systems

23. NEG coating in the close geometry set-up: pumping speed test gas • A tube is a natural choice for close geometry sample: • tube is the cheapest close geometry sample • easiest to make a uniform NEG coating • easiest to model • it is quite easy to make separate bake-out of coated and uncoated parts • less likely probability of contamination/poisoning from uncoated parts • full activation P2 NEG coated tube c P1 Pump 45th IUVSTA Workshop on NEG coatings Oleg Malyshev for Particle Accelerators and Vacuum Systems

24. NEG coating in the close geometry set-up: pumping speed test gas P2 NEG coated tube c P1 • Tube is the cheapest close geometry sample: • easiest to make a uniform NEG coating • easiest to model for initial condition • it is quite easy to make separate bake-out of coated and uncoated parts • less likely probability of contamination/poisoning from uncoated parts • almostfull activation (except extremities) Pump 45th IUVSTA Workshop on NEG coatings Oleg Malyshev for Particle Accelerators and Vacuum Systems

25. NEG coating in the close geometry set-up: pumping speed test gas P2 NEG coated tube c P1 But there still some problems: Gauge P1 ougassing during activation and experiments may be significant to poison the NEG film => a low outgassing gauge Gauge P1 crack the molecules and change the molecular mass spectrum => variable emission current The main gases at gauge P1 might be H2, CH4 or noble gases => RGA is essential Pump 45th IUVSTA Workshop on NEG coatings Oleg Malyshev for Particle Accelerators and Vacuum Systems

26. Modelling of the injection into the NEG coated tube • Calculated pressures P1 (far end) and P2 (at the injection) as a function of the wall sticking probability • Ratio of two pressures P1 and P2 as a function of the wall sticking probability • More details in the following presentation by Keith Middleman 45th IUVSTA Workshop on NEG coatings Oleg Malyshev for Particle Accelerators and Vacuum Systems

27. NEG coating PSD, ESD, ISD experiments in the open geometry set-up • It is not as the simple way to measure samples as it look like: • All the same problem of activation of the sample in the open geometry • Reflected and secondary particles • Hit uncoated parts of the set-up vacuum chamber • Desorbs some amount of gas, (QVC), which might be larger than desorption from the NEG coating, QNEG. • Significant but unknown (if sticking probability is unknown) amount of desorbed gas will be sorbed by a sample P1 P2 c  Test sample with NEG coating Pump 45th IUVSTA Workshop on NEG coatings Oleg Malyshev for Particle Accelerators and Vacuum Systems

28. NEG coating PSD, ESD, ISD experiments in the open geometry set-up • For the activated NEG coating it might be that • That means that in formula • Then uncertainty in the NEG pumping speed SNEG will led to large error bar for  P1 P2 c  Test sample with NEG coating Pump 45th IUVSTA Workshop on NEG coatings Oleg Malyshev for Particle Accelerators and Vacuum Systems

29. NEG coating PSD, ESD, ISD experiments in the open geometry set-up • NEG coating could not be fully activated => • Pumping sped is not well defined • Desorption from the uncoated parts of vacuum chamber due to reflected and secondary particles is not defined and can be larger when studied • Desorption yield from the NEG coating could not be calculated • NEG coating desorption and pumping could not be accurately studied in open geometry set-up! P1 P2 c  Test sample with NEG coating Pump 45th IUVSTA Workshop on NEG coatings Oleg Malyshev for Particle Accelerators and Vacuum Systems

30. NEG coating PSD, ESD, ISD experiments in the closed geometry set-up P2 P3 P1 c NEG coated test Chamber • Initial conditions are much better: • NEG coating can be fully activated • Well reproducible NEG pumping speed measurements => initial NEG pumping (at least!) is known. Pump 45th IUVSTA Workshop on NEG coatings Oleg Malyshev for Particle Accelerators and Vacuum Systems

31. Monte-Carlo model for close geometry set-up P1 P2 P3 ↓ ↓ ↓ If gas desorption happens from the NEG coatings only then the pressure readings from all gauges have the same ratio P1:P2:P3 for different sticking probabilities 45th IUVSTA Workshop on NEG coatings Oleg Malyshev for Particle Accelerators and Vacuum Systems

32. Monte-Carlo model for close geometry set-up P1 P2 P3 ↓ ↓ ↓ If gas desorption happens from the NEG coatings only then the pressure readings from all gauges have the same ratio P1:P2:P3 for different sticking probabilities This modelling was performed for the set-up for Heavy ion induced desorption studies at CERN. Results will be reported by Edgar Mahner 45th IUVSTA Workshop on NEG coatings Oleg Malyshev for Particle Accelerators and Vacuum Systems

33. Monte-Carlo model for close geometry set-up P1 P2 P3 ↓ ↓ ↓ If gas desorption happens from the NEG coatings only then the pressure readings from all gauges have the same ratio P1:P2:P3 for different sticking probabilities 45th IUVSTA Workshop on NEG coatings Oleg Malyshev for Particle Accelerators and Vacuum Systems

34. Closed geometry set-up: reflected and secondaries P2 P3 P1 c NEG coated test chamber • But • Some part of reflected and secondary particles will escape from the NEG coated chamber; for example, let’s take as little as 1% • Desorption from baked stainless steel is much higher than from NEG coating, let’s take a factor 10 • => QVC=0.1QNEG • Pumping from ‘green’ vessel to NEG coated test chamber increase with sticking probability Pump 45th IUVSTA Workshop on NEG coatings Oleg Malyshev for Particle Accelerators and Vacuum Systems

35. Closed geometry set-up: reflected and secondaries P2 P3 P1 c Pump I.e. gauge P1 is the only gauge we can use to study what is happening inside the NEG coated vacuum chamber. 45th IUVSTA Workshop on NEG coatings Oleg Malyshev for Particle Accelerators and Vacuum Systems

36. Closed geometry set-up: what gauge P1 measures? • The problem with reflected and secondary particles exists for the gauge P1 port also. • Desorption inside the gauge port causes: • Pressure inside the port might be much higher than inside the test chamber • Desorbed gas may saturate the NEG film close to the port and local pressure inside the test chamber will be higher near the gauge. • Overestimated pressure and desorption yields for the NEG coating. Gauge port NEG coated test Chamber 45th IUVSTA Workshop on NEG coatings Oleg Malyshev for Particle Accelerators and Vacuum Systems

37. Closed geometry set-up: what gauge P1 measures? Gauge port • A solution was to make • a shadow for the gauge port: • Reduce the entrance hall to diam.10 mm or even less NEG coated test Chamber 45th IUVSTA Workshop on NEG coatings Oleg Malyshev for Particle Accelerators and Vacuum Systems

38. Closed geometry set-up: what gauge P1 measures? Gauge port • A solution was to make • a shadow for the gauge port: • Reduce the entrance hall to diam.10 mm or even less • Baffle above the hall NEG coated test Chamber 45th IUVSTA Workshop on NEG coatings Oleg Malyshev for Particle Accelerators and Vacuum Systems

39. Closed geometry set-up: what gauge P1 measures? Gauge port • A solution was to make • a shadow for the gauge port: • Reduce the entrance hall to diam.10 mm or even less • Baffle above the hall • Long bake-out of the gauge port before the NEG activation. NEG coated test Chamber 45th IUVSTA Workshop on NEG coatings Oleg Malyshev for Particle Accelerators and Vacuum Systems

40. Closed geometry set-up: what gauge P1 measures? Gauge port • A solution was to make • a shadow for the gauge port: • Reduce the entrance hall to diam.10 mm or even less • Baffle above the hall • Long bake-out of the gauge port before the NEG activation. • Outgassing the baffle by 300 eV electrons . NEG coated test Chamber 45th IUVSTA Workshop on NEG coatings Oleg Malyshev for Particle Accelerators and Vacuum Systems

41. Closed geometry set-up: when P1<<P2 L P2 P3 P1 d c NEG coated test chamber • How sensitive pressure P1 to pressure P2 • Ideally, pressure P1 should be insensitive to pressure P2 • In practice, it depends on • sticking probability, α • ratio d/L • Gauge port design Pump 45th IUVSTA Workshop on NEG coatings Oleg Malyshev for Particle Accelerators and Vacuum Systems

42. Sensitivity of P1 to P2 • Calculated for one of configuration with L/d =31 • P1 is almost insensitive to P2 for α > 0.02 • Sticking probability α can be measures with gas injection (in the volume with the gauge P2) in the range: 3·10-4 < α < 0.02 ↑ ↑ Not accurate α Accurate High sensitivity P1(P2) Low sensitivity 45th IUVSTA Workshop on NEG coatings Oleg Malyshev for Particle Accelerators and Vacuum Systems

43. Closed geometry set-up: 3 gauge method • Configuration is most representing the accelerator vacuum chamber • No collusions at normal incident angle • Full or reflected flux can be measured with a collimator • Dynamic measurements of sticking probability All analysis details are in J. Vac. Sci. Technol. A23(3), May/June 2005, p.570 and the experimental results are in the following presentation of Alexander Krasnov. 45th IUVSTA Workshop on NEG coatings Oleg Malyshev for Particle Accelerators and Vacuum Systems

44. What we need to know for Design of Vacuum System • Dynamic behaviour as a function of bombarding particles energy or c • SR • Electrons • Ions • Heavy ions • … • As a function of large doses • As a function of NEG coated wall temperature • After accident: • Air vent during activation • Water leak • Radioactive gases • Self activation process 45th IUVSTA Workshop on NEG coatings Oleg Malyshev for Particle Accelerators and Vacuum Systems

45. Conclusions Studying of the NEG coating requires: • Review the technique traditionally used for the outgassing studies in the laboratory • Serious modelling and analysis prior the experiment allowed to design good set-up to study NEG coating. • Open geometry study of the NEG coating takes a lot of effort with very little success! • Close geometry experiments is the best choice • Outgassing/pumping and molecular cracking effects of the gauges are very important to count • Use of the RGAs is essential for the NEG coating study • Pressure inside the gauge may differ dramatically the pressure inside the vacuum chamber • Honest and accurate sensitivity and possible errors estimations help to understand results of the experiments 45th IUVSTA Workshop on NEG coatings Oleg Malyshev for Particle Accelerators and Vacuum Systems