Femtosecond Electron Imaging and Spectroscopy (FEIS-2) 2015 Lansing, Michigan, May 6-9, 2015. Oriented single crystal photocathodes: A route to high-quality electron pulses W. Andreas Schroeder Benjamin L. Rickman and Tuo Li Physics Department, University of Illinois at Chicago Department of Energy, NNSA DE-FG52-09NA29451
UED: Resolution Limit Non-relativistic regime Short-pulse Child s Law: ε 0E cath Nq > 2π( x 0 ) 2 Transverse emittance: 1 εt = x0. p T 0 mc conserved Spot size x s Divergence p Ts /p 0 E cath Laser Photocathode Anode Electron gun p 0 Laser Sample (lattice constant a) θ m mh 2ap 0 Detector δa ξ a 2Nq m. pt 0 a mh x 0 Observable ; with ( hω φ) p 0 =?? s πε0e T cath 3 D.H. Dowell & J.F. Schmerge, Phys. Rev. ST Acc. & Beams 12 (2009) 074201 K.L. Jensen et al., J. Appl. Phys. 107 (2010) 014903
Photoemission Theory I The semi-classical three-step Spicer model 2 3 e - 1. Photoexcitation 2. Transport to surface 3. Emission from surface E F 1 ħω φ Transport Real electronic band Photoexcitation into upper state near vacuum level Emission from upper excited state High quantum efficiency (η PE ) AND Response time Lifetime (ps-ns) Photocathode Vacuum NOT suitable for UED Examples: NEA GaAs, KCsSb, GaSb, diamond, Cu(111)? C.N. Berglund & W.E. Spicer, Phys. Rev. 136, A1030-A1044 (1964) P.J. Feibelman & D.E. Eastman, Phys. Rev. B 10, 4932-4947 (1974)
Photoemission Theory II The quantum mechanical one-step model e - Photoexcitation into a virtual state (excited copy of filled band) emitting into the vacuum in one step Low η PE ~ 10-5 to 10-7 ħω φ Instantaneous emission process Suitable for UED E F Examples: Most metals Photocathode Vacuum G.D. Mahan, Phys. Rev. B 2, 4334-4350 (1970)
Experiment: Solenoid Scan 2W, 250fs, 63MHz, diodepumped Yb:KGW laser ~4ps at 261nm (ħω = 4.75eV) YAG scintillator optically coupled to CCD camera Beam size vs. magnetic coil (lens) current measured Analytical Gaussian (AG) pulse propagation model to extract p T0 J.A. Berger & W.A. Schroeder, J. Appl. Phys. 108 (2010) 124905
Results: Polycrystalline Cr Solenoid scan measurement Range for p T from p T 0 m0( hω φ) = 3 φ = 4.50(±0.05)eV and ħω = 4.75eV AG model simulation of experiment gives p T0 = 0.155(±0.01) (m 0.eV) 1/2
Results: Metals Ten polycrystalline metal photocathodes Only Ag and Cu (noble metals) consistent with p T,expt = p T 0 = m 0 ( hω φ) 3 for others p T,expt < m 0 ( h ω φ) 3 Band structure effects? e.g. m* < m 0 B.L. Rickman et al., Phys. Rev. Lett. 111 (2013) 237401
Band Structure Effects Transverse momentum p T conserved in photoemission Classical metal E 2 p T 2m 0 p T, max. 2 Metal with m* < m 0 = m0 E E > p = 2m * F E F 2 p T 2m 0 Vacuum level ħω φ ħω φ E F E E F E 2 p T 2m 0 2 p T 2m p F p T,max. p T p F p T,max. p T
Fermi Surfaces K www.phys.ufl.edu/fermisurface/
Fermi Surfaces K Ag www.phys.ufl.edu/fermisurface/
Fermi Surfaces K Ag Mo www.phys.ufl.edu/fermisurface/
Fermi Surfaces K Ag Mo Nb Full (p,e) band structure required for photoemission analysis www.phys.ufl.edu/fermisurface/
Work Function Anisotropy Example: φ (ijk) for Mo by electron emission microscopy Polycrystalline metal photocathodes generate inhomogeneous electron beams Any photoemission analysis must include φ (ijk) D. Jacobson & A. Campbell, Metall. Trans. 2, 3063-3066 (1971)
Thin-slab Evaluation of φ (ijk) Example: φ (001) = 4.53(±0.05)eV for Mo φ (001) E vac. E F NOTE Schottky effect not included: ee DC e 2 πε 0 ±50meV uncertainty in φ (ijk) C. J. Fall et al., Journal of Physics: Condensed Matter 11, 2689 (1999)
Photoemission Simulation: Ag fcc crystal lattice ħω = 4.75eV Lowest index face with lowest φ (ijk) E = ħω - φ (110) = 0.23eV ±50meV error in φ (ijk)
Photoemission Simulation: Ag fcc crystal lattice p T, max. 2 = m0 E E F Photoemitting electron-like states for ħω = 4.75eV (p T and E conserved)
Photoemission Simulation: Ag fcc crystal lattice E, p T conservation PLUS Barrier transmission, T(p z,p z0 ) Spatially-averaged p T0 = 0.267 (m 0.eV) 1/2
Photoemission Simulation: Mo bcc crystal lattice Lowest index face with lowest φ (ijk) E = ħω - φ (001) = 0.22eV ±50meV error in φ (ijk)
Photoemission Simulation: Mo bcc crystal lattice Both electron- and hole-like states contribute to photoemission
Photoemission Simulation: Mo bcc crystal lattice E, p T conservation PLUS Barrier transmission, T(p z,p z0 ) Spatially-averaged p T0 = 0.219 (m 0.eV) 1/2
Photoemission Simulation: Nb bcc crystal lattice φ (ev) Lowest index face with lowest φ (ijk) E = ħω - φ (001) = 0.73eV ±50meV error in φ (ijk)
Photoemission Simulation: Nb bcc crystal lattice φ (ev) Only hole-like states contribute to photoemission
Photoemission Simulation: Nb bcc crystal lattice E, p T conservation PLUS Barrier transmission, T(p z,p z0 ) φ (ev) Spatially-averaged p T0 = 0.196 (m 0.eV) 1/2
Experiment vs. DFT Analysis NOTE: Polycrystalline vs. singlecrystal comparison Other crystal faces with smaller E = ħω φ (ijk) contribute lower p T0 DFT analysis at T e 0K Experiment at 300K
Temperature dependence: T e p T (ħω,t e =300K,m*=m 0 ) p T = m 0 ( hω φ (001 ) 3 ) p T,DFT (T e =300K) pt = m0 k B T e ħω < φ (001) p T,DFT (T e 0K) p T,expt (Nb poly ) ħω = 4.75eV T. Vecchione et al., TUPSO83, Proc. of FEL 2013, pp. 424-426.
Summary Photocathodes for UED Single-crystals for homogeneous electron beam generation (ħω > φ (ijk) ) Higher η PE (?) and higher conductivity (σ and κ) Virtual excited state emission Instantaneous response Emission from low m* states: p F > p T,max Lower p T0 Hole-like emission states are preferred: Even lower p T0 and less sensitive to T e (e.g., laser heating) Robust (e.g., high m.p.) and chemically inert Future work Direct comparison with theory: Single-crystal photocathodes Crystalline compounds: Mo x Nb 1-x, semiconductors, A 3 B (e.g., Nb 3 Sb), Search for ultra-low p T0 solid-state photocathodes: p T0 approaching cold atom electron sources TID issues?
Finally Calculating p T (ijk) for all elemental metals Results available on-line at http://people.uic.edu/~tli27/database.html. - E.g., hcp Mg(1010) face emission: (ħω = 4.67eV) - p T (0110) Mg(poly) φ - (1010) = 3.66eV p T (0001) X.J. Wang et al., Proceedings of LINAC2002, Gyeongju, Korea.