The line shapes of multiplet components of the complex x-ray photoemission spectra of the Ta 4f core levels in 1T-and 4Hb-TaS2 are shown to have different asymmetries associated with the charge density wave induced differences in the densities of states at distinct Ta sites which produce site specific local screening of the photohole. This underlines an important general effect in the interpretation of line shapes in x-ray photoemission. PACS numbers: 79.60.Jv, 79.60.Bm The screening of the photohole produced in x-ray core level photoemission (XPS) from a metal produces an inelastic tail to the low kinetic energy (KE) side of the XPS line. The resulting asymmetry depends on the excitations available to the screening carriers, i.e. , on the joint density of states (JDOS) near the Fermi energy (EF) [1]. A charge density wave (CDW) splits XPS lines because of the distinct chemical shifts for atomic sites with different local electron densities [2]: We show that the line shapes of the resulting multiplet components also differ because the local densities of states (LDOSs) are also modified by the CDW, and that detailed line shape fitting for several polytypes of TaS2 extracts parameters related to the electronic structure near specific atomic sites in the CDW supercell.TaS2 has hexagonal sheets of Ta atoms sandwiched between similar S sheets to form layers; the Ta coordination by the S atoms is either trigonal prismatic or octahedral.The 2H and 1T polytypes have unit cells with two trigonal prismatic and one octahedral layer, respectively, while 4Hb-TaS 2 has four, alternating between "1T"-and "2H"-like. Their electronic structures are similar; above the S 3p valence bands lies the Ta d band holding one electron per formula suit, and resulting in metallic properties. But in detail these d bands, and therefore their CDW behavior, differ; a quasicommensurate inplane~13 X~13 CDW exists for 1T-TaS~even at room temperature [3], at which 2H-TaS2 has no CDW. Below 180 K, 1T-TaSz adopts the 1T3 phase studied here, with a commensurate~13 &&~13 CDW with three distinct Ta sites -a, with one atom, and b and c with six atoms each -with different local charge densities. 4Hq-TaS2 shows behavior characteristic of both layer types; below 315 K it exhibits a~13 X~13 CDW associated with its 1T layers and below 22 K a 3 X 3 CDW characteristic of the 2H polytype [3,4]. Empirical arguments [2], and scanning tunneling microscopy (STM) [5], suggested a CDW amplitude in 1T3-TaS2 of a considerable fraction of an electron, though another [6] based on chemical shifts and a self-consistent calculation suggested 0.05 electron. A simple linear combination of atomic orbitals (LCAO) calculation for 1T3-TaS2, explicitly including the effects of A A2 P",(E) =, B(E -e") + 1 -,B(E),where A is the appropriate matrix element; convolving these for multiple excitations, and assuming A is small, gives the overall probability of energy loss E, P(E)~A~( e" ' -1) e ' 'exp g dt.-jL V P, V the CDW on the Ta d bands, shows that the LDOS is different for each...

Lineshapes in x-ray core-level photoemission (XPS) reflect the local electronic environment of the atomic species from which they originate, and the excitations of the material system which reduce the emerging photoelectron's kinetic energy. For metallic systems, we show that detailed analysis of such lineshapes provides important information on the local conduction band electronic structure at distinct sites of the same atomic species. A form is derived for the core-level lineshape which is based on the spectrum of excitations in the conduction band, predominantly the formation of electron-hole pairs which screen the core-level hole produced in XPS; this is more general (but less rigorous) than the Doniach-Sunjic formulation. The lineshape is used as the basis for a numerical data analysis package-SHAPER-which extracts lineshape information from experimental XPS data by least-squares fitting. Various lineshapes derived from hypothetical conduction band profiles, and the reliability of the fitting process, are examined.

SHAPER, a computer package for x-ray photoemission lineshape analysis, is applied to experimental data on 2H-TaS 2 and some of its transition metal intercalates, revealing that charge transfer into the conduction band resulting from intercalation affects the core-level lineshape by modifying the density of states at the Fermi energy. The observations are discussed in terms of the rigid-band model of intercalation. In order to emphasize the interactive character of the fitting process, and its strengths and limitations, a range of fits of increasing complexity is presented.

High-resolution x-ray photoemission (XPS) studies of the lineshapes of the Ta 4f core levels in 1T-and 4H b -TaS 2 are reported. These materials show charge density wave (CDW) behaviour at room temperature, and the local electron densities at different atomic sites are shown to produce different local screening of the core-level photohole, so the components of the doublet or triplet 4f emission line resulting from the CDW-induced shifts in the local potential at different atomic sites have different core line asymmetries. The complex lineshapes are analysed using the iterative fixing procedure, SHAPER, allowing connections to be made between the observed spectra and the site-dependent density of states.

The photoemission lineshapes of the Ta 4f levels in 2H-TaS2 and its intercalates MnIi4TaS2 and CoIi,TaS2 have been studied using synchrotron radiation. The peaks show marked asymmetries because of the many-electron screening response to the core-hole potential. The different lineshapes are discussed in terms of the rigid band model of the electronic structure of the intercalates, and of the effects of charge transfer into the narrow conduction band on the many-electron response. Calculations of the lineshapes using a simple model for the conduction band density of states allow the experimental lineshapes to be understood in terms of intrinsic losses through multiple excitations within the narrow conduction band, together with some extrinsic plasma losses.