SLAC-PUB-15116 SURFA ACE CHAR RACTER RIZATION N OF THE E LCLS RF R GUN CA ATHODE E* # A. Brachmann B , F.-J. Deckeer, Y. Ding, D. D Dowell, P. P Emma, J. Frisch, F S. Giilevich, G. Hays, H Phh. Hering, Z. Z Huang, R. Iverson, H. Loos, A. Miiahnahri, D. Nordlund, H.-D. H Nuhn, J Turner, J. Welch, W W. White, W J. Wu,, D. Xiang P. Pianetta, J. SLAC Naational Acceelerator Labooratory, Mennlo Park, CA A 94025, USA A, U.S.A. A Abstract The first coppper cathode installed in thee LCLS RF gunn was used duriing LCLS com w mmissioning foor more than a y year. Howeverr, after high charge c operatioon (> 500 pC)), t the cathode showed s a deccline of quantum efficiencyy w within the areea of drive lasser illuminatioon (fig. 1). We r report results of SEM, XPS S and XAS stuudies that werre c carried out on this cathode after a it was rem moved from the g gun. X-ray absorption and X-ray photoelectronn spectroscopy reveal surfacee contaminatioon by variouus h hydrocarbon c compounds. In n addition wee report on the p performance of the second installed cathode withh e emphasis on thhe spatial distriibution of electtron emission. MOTIV VATION We assumpttion the reason n for the decay of the cathodees q quantum efficiiency is the co ontamination of its surface byy h hydrocarbon s species contain ned in the vacuuum system of o g and adjaccent beam linee. The UV drrive laser beam gun m a and/or the elecctron beam intteract with the residual gas of o t vacuum syystem or alread the dy physisorbedd species, whichh a then ionized by drive laaser beam and electron beam are m a and depositedd on the caathodes surfacce forming a c chemisorbed layer resultin ng in an inncreased workk f function. Uponn the removal of the cathodee a dark ring of o ~ 3 cm diameeter on its surfface, macrosco fopically visiblee, w discovered (fig. 2). It iss unclear if the deposition of was o t ring occurrred during RF processing of the gun or if it the i h formed durring normal op has peration. This study was w conducted d to confirm our assumptionns f surface conntamination an for nd study the chhemistry of the c cathode surfacce layer, in partticular the com mposition of the d dark ring. In addition, a a charracterization off the surface byy SEM techniquues was desirrable after ~ 18 months of o o operation. Figure 1: Electron beam m image, low quantum efficciency where catthode is exposed to the drivee laser beam (L LCLS cathode 2) c cathode center ~3 cm dia. ring r X-RAY PHOTO-ELEC CTRON (XP PS) AND X-RAY ABSO ORPTION SP PECTROSC COPY (XAS)) X-Ray Photto-electron speectroscopy wass carried out at a SSRL’s SPEA AR 3 beam lines l 8-2 and 10-1 [1,2] too i investigate thee surface chemiistry on a qualiitative level. Figure 2: circularly shaaped ‘dark ringg’ contaminatiion of cathode suurface ___________________________________________ *Work supported by DOE Contract No. DE-AC02-76SF00515. # brachman@slac.stanford.edu Presented at the 1st International Particle Accelerator Conference (IPAC 2010) Kyoto, Japan, May 23 - 28, 2010 spectrum show the absorption spectrum of oxygen and copper species. Both are the result of absorption of higher order photons. A detailed spectrum of the carbon edge is shown in fig. 7. 20 15 C-O K1s Cu L2 2p1/2 2.0x10 6 1.5x10 6 C K1s 25 I [A.U.] The chamber’s mounting system allowed a translation of the sample in x, y and z by a few inches in either direction. Photoemission spectra were recorded by a curved mirror analyzer system. X-ray Absorption Spectroscopy (XAS) spectra were recorded using the ratio of incident beam intensity and sample current. The XPS spectrum indicates a surface layer composition containing oxygen, nitrogen and carbon. A variety of peaks can be attributed to the bulk material (Cu, Pb, Bi and S). An overview spectrum is given in figure 3. 10 Cu L3 2p3/2 C K 1s O K1s CTS 1.0x10 6 5 Bulk Cu Contaminants (ppm level) N K1s 0 292 C(A) 290 288 S 2p3 ? Pb/Bi 4f 7/2 5.0x10 286 284 282 280 Binding energy [eV] Cu M13s 5 Figure 5: Deconvolution of the C K1s XPS peak Cu M23p1/2; Cu M3 3p1/2 Cu 3d 0.0 Fermi edge (Eb=0) eV 1000 800 600 400 200 2.6 Carbon species 0 Binding Energy [eV] Figure 3: Overview XPS spectrum 2.2 CTS To determine the location of the ‘dark’ ring (see fig. 2), measurements of the C-K1s and Cu-3d peak were performed in a grid pattern and their ratio was used to obtain information on the relative C distribution (fig. 4). Cu 3rd order photons (Cu2O) 2.4 ? 2.0 1.8 Oxygen 2nd order photons 1.6 1.4 1.2 260 280 300 320 340 360 Photon Energy [eV] Figure 6: XAS Overview spectrum 2.4 C-H 288.4 Absorption [A.U.] 2.2 2.0 1.8 * 292 C-C σ * 293 C-O σ shoulder ~ 287.5 C-CH3 and C-CH2 292 C-H 286.0 * Figure 4: Ratio of the C-K1s/Cu-3d XPS peak (normalized) Higher resolution measurements of the carbon peak and peak deconvolution were carried out to resolve the speciation of the surface carbon layer (fig. 5). X-Ray absorption spectroscopy was used to further determine the nature of the carbon layer. An overview spectrum is given in fig 6. The spectrum shows the absorption of photons 260 – 370 eV. The main feature is the carbon absorption edge. Adjacent areas of the C 1s - π 285.2 1.6 1.4 280 285 290 295 300 Photon Energy [eV] Figure 7: XAS details of Carbon (C K1s) absorption edge ANALYSIS OF XPS AND XAS XPS spectra were aligned to the Fermi edge of the spectrum, thereby converting the spectrum to units of binding energy in electron volt. XAS spectra were calibrated using the exciton peak of a highly oriented pyrolytic graphite (HOPG) reference spectrum (291.65 eV). For peak assignment of carbon, nitrogen and oxygen species, the data in [3] were used as a reference. Main emphasis was the analysis of the surface layer Carbon distribution and speciation. For other species, the tables published by the Center for X-ray Optics and Advanced Light Source of the Lawrence Berkeley National Laboratory were used [4]. The measurement of the spatial distribution of the carbon to copper ratio indicates a continuous layer of carbon species across the entire cathode. The carbon XPS peak (C K1s) was observed at any location on the cathode. A larger fraction of carbon is evidence of a larger thickness of the carbon layer. The XPS measurements of the distribution of the C/Cu ratio confirm the presence and carbon containing composition of the macroscopically visible dark coloured ring. Higher resolution measurements of the C K1s peak at various locations reveal the complex structure of the observed carbon species. Slight intensity shifts of the peaks contributing to the spectrum were found at different locations. Qualitatively, spectra are very similar across all regions of the cathodes surface. The peak at ~ 284 eV indicates the presence of elemental carbon. Deconvolution of the XPS spectra show a main peak of elemental carbon but also indicate a significant fraction of C-O bonds ( ~ 288 eV, fig. 5). A more detailed carbon speciation is possible using XRay absorption spectroscopic measurements of the C K1s absorption edge (see fig. 7). The main identified features are the π* antibonding state of graphitic carbon located at ~ 285 eV as well as states in the region of 287-289 eV that are associated with C-CHx and C-CH bonds. The broad peak at ~ 292 eV can be attributed to the σ* orbitals of C-C and C-O bonds. A C-H bond can also be expected here. In general, EDX analysis revealed the foreign nature of particulates (mainly composed of C, O, Mg, Al, Si Ca and Ti). Surface carbon was detected; however a conclusive study was impossible due to the rapid alteration of the surface layer by the impinging electron beam. Figure 8: SEM detail of cathode surface: center and lower left – RF breakdown events, upper middle – artefacts of laser cleaning CONCLUSIONS As expected, the XPS and XAS measurements confirmed the presence of complex hydrocarbon chemistry on the cathodes surface. Similar phenomena currently occur on the second installed LCLS Cathode. In general it is apparent that this process cannot be entirely avoided. A study of ion migration within the LCLS RF gun and a discussion of possible surface contamination mechanisms is given in [5]. It is therefore desirable to have the cleanest UHV system as possible to ensure a slow speed of surface contamination. Further experiments are planned to investigate this process in more detail, and to develop techniques for surface cleaning and passivation. SEM ANALYSIS ACKNOWLEDGEMENT SEM imaging and EDX analysis were carried out on an Aspex PSM-76 LS scanning electron microscope. Various areas of the cathode surface were analyzed. In particular, a clearly macroscopically visible dark ring of surface deposits (~ 3 cm in diameter), the center of the cathode and other surrounding areas were studied. A variety of features were observed: We would like to thank SLAC’s Klystron group for assistance in sample preparation and SEM measurements. We are also very grateful for the tremendous experimental support we received from the operations staff at SSRL’s SPEAR 3 facility. 1. [1] [2] [3] [4] [5] 2. 3. Surface damage due to RF breakdowns that occurred during RF processing of the gun and cathode (fig. 8). Surface perturbations that are results of ‘laser cleaning’ of the surface (fig 8). Contamination by particles REFERENCES http://www-ssrl.slac.stanford.edu/beamlines/bl8-2/ http://www-ssrl.slac.stanford.edu/beamlines/bl10-1/ J. Stohr, NEXAFS spectroscopy, Springer, 1992 http://xdb.lbl.gov/ A. Brachmann et al., Simulations of ion migration in the LCLS RF gun and injector., this conference (TUPE064).