Angular Distribution Measurements of Sputtered Particles at UCSD Presented by Russ Doerner for Jonathan Yu, Edier Oyarzabal and Daisuke Nishijima QMS measurements in unmagnetized plasma Moly Carbon clusters Optical measurements in PISCES-B
Angular distribution of sputtered Mo and C Jonathan Yu and Eider Oyarzabal April 26, 25 Sputtering of grids is a critical issue for ion thruster lifetimes. Modeling requires measurement of angular distribution of sputtered atoms. Previous sputtering measurements: profilometry, weight loss, visible spectroscopy. Here, we use an energy-selecting mass spectrometer to detect sputtered ions. 2
Plasma Chamber and Quadrupole Mass Spectrometer (QMS) n = 2 cm -3 T e = 3 ev P = 5 mtorr Γ ~ 8 ions s - cm -2 Target Plasma source θ -V bias QMS Energy filter Mass filter Electron multiplier detector Ionizing stage used for neutral detection 3
Mo sputtering with Xe plasma; Ion detection Count Rate (c/s) 5 4 3 2 T ixe =.8 ev f M = E /2 e -(E-e φ s)/t Sheath potential drop ~ 5T e Xe ions elastic scattering Sputtered energy ~/2 E b =3.5 ev.5 Xe peak Angle 6 Angle 35 No signal when sample is not facing QMS Sputtered Mo ionized in plasma 5 5 2 25 Energy (ev) 4
Mo ion sputtered signal increases with plasma ion bombardment energy Count Rate (c/s) 3 25 2 5 5 Mo ions P = 5 mtorr θ = 6 V bias = 5 V V 5 V 2 V 5 5 2 25 Energy (ev) 5
Angular distributions of low and high energy Mo ions 3 Count Rate (c/s) 25 2 5 5 5 5 2 25 Energy (ev) Target QMS Low energy peak High energy peak 26 24 22 2 8 6 4 2 8 6 4 2 2 4 6 8 2 4 6 8 2 22 24 26 27 3 24 33 2 8 3 5 6 2 9 bias25 bias5 bias75 bias bias25 bias5 bias75 bias2 bias225 9 8 7 6 5 4 3 2 2 3 4 5 6 7 8 9 27 3 24 33 2 8 3 5 6 2 9 bias25 bias5 bias75 bias bias25 bias5 bias75 bias2 bias225 6
Xe on Mo comparison with other work. Sputter Yield (atoms/ion). E-3 E-4 JPL Weijsenfeld Rosenberg Bhattacherj Blandino Tartz Doerner Current study E-5 2 4 6 8 2 4 6 8 2 22 24 26 Ion energy (ev) 7
Carbon weight loss is not explained by atom sputtering and chemical processes [R.P. Doerner, D.G. Whyte and D.M. Goebel, J. Appl. Phys. 93(23)586] Sputtering Yield (atoms/ion).... -5 Weight Loss (Mo) Spectroscopy (Mo atoms) Weight Loss (C) Spectroscopy (C atoms) Spectroscopy (C 2 molecules) Chemical erosion of CO is estimated to be within the weight loss error bars at 5 ev Chemical erosion of CO is ~independent of incident energy Carbon dimer erosion appeared more dominant than the carbon atom sputtering term -6-7 Mo Spectroscopic Detection Limit 5 5 2 Ion Energy (ev) Carbon trimers, etc. were not investigated at the time 8
Neutral particle detection with carbon target /24/5 kw Ar plasma on Carbon, P=4.mTorr, QMS neutrals Count Rate (c/s) 2 5 5 Mass 2 Ion energy = 2 ev Ion energy = 5 ev Ion energy =.*Mass 28 signal 2 4 6 8 Intensity (bias on) / Intensity (bias off).5.4.3.2..9 Mass 2 Mass 44 Mass 28.8 5 5 2 25 Neutral particle energy (ev) -V bias (V) 9
Appearance mass spectroscopy shows that measured mass 2 signal is cracked CO and CO 2 4 8 Mass 2 8 Mass 28 Count Rate (c/s) 6 4 2 IE(C)=.3 ev DI(CO)= 22.4 ev DI(CO )= 2 24.6 ev Count Rate (c/s) 6 4 2 IE(CO)= 4. ev 2 4 6 8 Electron Energy (ev) 2 4 6 8 Electron Energy (ev) Carbon Target Plasma source V bias QMS Ionizer Future plans for neutral sputtered particle detection: Set electron energy in ionizer to 5-2 ev. Use long-time accumulation and/or lock-in detection to increase signal to noise. Upgrade ionizer stage to increase ionization efficiency.
Ion detection -- Carbon cluster sputtering by Xe High energy tail represents sputtered particle distribution. Low energy component is nearly thermal, indicating collisional scattering. Count Rate (c/s) 8 7 6 C 3 ion, θ = 6 o C 3 ion, θ = 9 o 5 4 3 f M f 2 T f + f M T 5 5 2 25 3 Energy (ev) Maxwellian: f M = E /2 e -(E-e φ s )/T Thompson: f T = E/(E+E b ) 3 Fits yield φ s =. V T =.5 ev E b =. ev
Carbon cluster low- and high-energy angular distributions Count Rate (c/s) 3 25 2 5 5 5 5 2 25 3 Energy (ev) Mass C 2 24 36 32 28 24 2 6 2 8 4 27 3 Mass 24 (high energy population) 33 3 6 9 V 5 V V 5 V 225 V 25 2 5 5 27 3 Mass 24 (high energy population) 33 3 6 9 V 5 V V 5 V 225 V C 3 36 ounts/s 2 8 6 4 2 2 Mass 36 (low energy population) 33 3 27 3 6 9 V 5 V V 5 V 225 V ounts/s 35 3 25 2 5 5 5 27 3 Mass 36 (high energy population) 33 3 6 9 V 5 V V 5 V 225 V 2
Mo & C Sputtering Summary High energy ion signal represents sputtered atoms that are ionized in the plasma. This angular distribution is peaked at θ ~ 6 o. Low energy ion signal represents sputtered atoms that have been ionized and elastically scattered. Lower-pressure plasmas are needed to reduce the complication of ion-neutral elastic scattering. Direct detection of sputtered neutrals is a challenge due oxygen contamination and low ionization efficiency in the QMS. We are in the process of upgrading the ionizing stage. 3
QuickTimeý Dz TIFFÅiLZWÅj êlí ÉvÉçÉOÉâÉÄ Ç Ç±ÇÃÉsÉNÉ`ÉÉǾå ÇÈÇžÇ½Ç ÇÕïKóvÇ-Ç Behavior of eroded beryllium atoms in PISCES-B D. Nishijima, R.P. Doerner, M.J. Baldwin, R. Seraydarian and G.R. Tynan Center for Energy Research, University of California at San Diego, USA J.N. Brooks Argonne National Laboratory, USA 4
Experimental setup PISCES-B Spectroscopy (absolutely calibrated system) The light is guided with a mirror and focused with a lens to the entrance slit of spectrometer equipped with a 2-D CCD camera. Vertical (y) profiles of Be I (2s2p P-2s3d D: 457.3 nm) line intensity emitted from sputtered Be neutrals were measured at several z positions (every 2 mm near the target). Spatial resolution: z ~ mm y ~.64 mm y or r Be z = 52 mm Double probe n e ~ -3x 8 m -3 5-6 cm θ z T e ~ 8 ev B Γ i ~.5-3x 22 m -2 s - Flat radial profile Be target (~ 2. cm dia.) 5
Abel inversion can be applied to derive local emissivity (as a function of radius) from line integrated intensity (as a function of y). Abel inversion: ε(r) = π r a di(y) dy dy y 2 r 2 PISCES-B Be I intensity [ 6 ph/(s m 2 sr)] 2.5.5 E i = 5 ev @ z = 2 mm 25425, 4436_4 measured fit Be I emissivity [ 6 ph/(s m 3 sr)] 7 6 5 4 3 2 25425, 4436_4 E i = 5 ev @ z = 2 mm -3-2 - 2 3 y [mm] 5 5 2 25 3 35 r [mm] 6
Derivation of local ground state Be density 4πε σv 457.3nm n e n Be = 4πε n Be = σv 457.3nm n e PISCES-B σv 457.3nm : Photon emission coefficient [ph m 3 /s] of Be I line (2s2p P-2s3d D: 457.3 nm) from ADAS data base Note that excitation only from the ground state (2s 2 S) is taken into account. There is no information available about a metastable state (2s2p 3 P) in the experiment. n Be [ 6 m -3 ] E i = 5 ev 7 25425
Angular distribution does not makedly change from E i = 5 to 5 ev. Distribution at E i = 5 ev with lower P n is slightly different from others at θ ~ 3-6 deg. Further investigation on effects of P n is needed. n Be /n Be (θ=) n Be /n Be (θ=).2.8.6.4.2.2.8.6.4.2 E i = 5 ev 25425 d = mm d = 6 mm d = 22 mm 9 9 d = 22 mm P n = 6.-6.3 mtorr E i = 5 ev E i = 5 ev E i = 8 ev E i = 5 ev cosine cosine 6 6 PISCES-B 3 P n = 4.7 mtorr 3 8
Differences between modeling and experiment are seen at large angle (θ > 6 deg). WBC monte carlo code simulation (by J.N. Brooks) uses TRIM-predicted cosine angular distribution. At larger d, the distribution becomes closer to cosine distribution for point source. In experiments, n Be does not become small at large angle (θ > 6 deg.). -- Deviation from cosine distribution? n Be /n Be (θ=) n Be /n Be (θ=).2.8.6.4.2.2.8.6.4.2 WBC simulation Experiment d = mm d = 6 mm d = 22 mm cosine 25422 d = mm d = 6 mm d = 2 mm cosine 9 9 6 6 PISCES-B E i = 5 ev E i = 5 ev 9 3 3
Summary and conclusion PISCES-B Measurement of 2-D profiles of Be I emission and density eroded from Be target has become possible using spectroscopic techniques with the absolutely calibrated system. The incident ion energy dependence of angular distribution of eroded Be atoms is small. At large angle (θ > 6 deg), the measured angular distribution deviates from the WBC modeling result, where the cosine distribution is employed. The measured axial e-folding distance of eroded Be atom density is in good agreement with the WBC modeling. 2