Theoretical Approach to the Configuration Fluctuation in Sm-Chalcogenides

  • T. A. Kaplan
  • S. D. Mahanti
  • Mustansir Barma


We present a theoretical approach to the problem of configuration mixing in Sm chalcogenides. Properties in the collapsed phase of SmS are discussed in terms of an essentially localized or excitonic picture of most of the Sm 5d-electrons, with a small number (≃.1 electrons per Sm) occupying the free-electron-like states at the bottom of a broad 5d-band. We review how such an essentially localized model for the 5d-electrons can lead to the observed results on volume and degree of mixing vs. pressure, low-T specific heat, dc electrical conductivity, plasma frequency, XPS intensities and magnetism. We also present new calculations of phase boundaries in the T-p plane, obtaining for SmS both a first order and a second order boundary within a simplified model. The latter boundary, which occurs e.g. at high p and low T makes contact with a recent experimental result of Guntherodt et al. The unusual shape of the observed first-order boundary is shown to be in accord with our general model, A new evaluation of the low frequency dielectric constant (ħω⋝.lev) shows behavior very similar to the unusual experimental results obtained recently by Batlogg et al. and by Allen. We compare the above picture with the more common models where all the 5d-electrons (~.7 per Sm in SmS) occupy free-electron like states, and give a critique of the latter.


Conduction Electron Narrow Band Hubbard Model Mixed Phase Plasma Edge 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Valence Instabilities and Related Narrow-Band Phenomena, Edited by R.D. Parks, Plenum Press 1977.Google Scholar
  2. 2.
    T.A. Kaplan, S.D. Mahanti and Mustansir Barma, ref. 1, p. 153.Google Scholar
  3. 3.
    A. Jayaraman, V. Narayanamurti, E. Bucher, and R.G. Maines, Phys. Rev. B11, 2783 (1975); and ref. 1 p. 61.Google Scholar
  4. 4.
    J.M.D. Coey, S.K. Ghatak, and F. Holtzberg, AIP Conf. Proc. 24, 38 (1974); and ref. 1, p. 211.CrossRefGoogle Scholar
  5. 5.
    M.B. Maple and D. Wohlleben, Phys. Rev. Letters 27, 511 (1971).CrossRefGoogle Scholar
  6. 6.
    M. Campagna, E. Bucher, G.K. Wertheim, and L.D. Longinotti, Phys. Rev. Letters 33, 165 (1974). The X-ray photoemission studies were not made on SmS under pressure, but rather on Sm1-XRXS where R = Y or Gd. The concentration x acts similarly to pressure (see e.g. ref. 3).CrossRefGoogle Scholar
  7. 7.
    S.D. Bader, N.E. Phillips, and D.B. McWhan, Phys. Rev. B7, 4786 (1973).Google Scholar
  8. 8.
    E. Kaldis and P. Wachter, Sol. St. Comm. 11, 907 (1972);CrossRefGoogle Scholar
  9. 8a.
    B. Batlogg, A. Schlegel, and P. Wachter, Phys. Rev. B14, 5503 (1976).Google Scholar
  10. 9.
    B. Batlogg and P. Wachter, ref. 1, p. 537; J. W. Allen, ref. 1, p. 533 found a similar result in the mixed valence compound SmB6.Google Scholar
  11. 10.
    E. Yu. Tonkov and I.L. Aptekar, Sov. Phys. Solid State 16, 972 (1974).Google Scholar
  12. 11.
    C.M. Varma, Rev. Mod. Phys. 48, 219 (1976).CrossRefGoogle Scholar
  13. 12.
    Some examples in addition to those discussed in ref. 11 are: L.L. Hirst, ref. 1, p. 3; T. Penney, ref. 1, p. 86; S.K. Ghatak and M. Avignon, ref. 1, p. 229; B. Alascio, ref. 1, p. 247; B. Coqblin, A.K. Bhattacharjee, J.R. Iglesias-Sicardi, and R. Jullien, ref. 1, p. 365; J.H. Jefferson and K.W.H. Stevens, ref. 1, p. 419. For others consult ref. 1.Google Scholar
  14. 13.
    According to the titles of the papers by S. Doniach, ref. 1, p. 169 and in Proc. of the Oxford Conference on Itinerant Electron Magnetism, Sept. 1976, to appear in Physica, his work would appear to fall under this band-model category. However, the specific model Hamiltonian studied is rigorously a localized model; whereas the connection to the band model is not rigorous.Google Scholar
  15. 14.
    T. A. Kaplan and S.D. Mahanti, Phys. Lett. 51A, 265 (1975).Google Scholar
  16. 15.
    S.D. Mahanti, T.A. Kaplan and M. Barma, Phys. Lett. 58A, 43 (1976).Google Scholar
  17. 15a.
    J. Schweitzer, Phys. Rev. B13, 3506 (1976), studied stability properties of the localized model.Google Scholar
  18. 16.
    A localized picture for most of the d-electrons in connection with SmB6, was discussed earlier by R.L. Cohen, M. Eibschütz and K.W. West, Phys. Rev. Lett. 24, 382 (1970)CrossRefGoogle Scholar
  19. 16a.
    and in more detail by J.C. Nicfcerson, R.M. White, K.N. Lee, R. Bachmann, T.H. Geballe, and G.W. Hull, Jr., Phys. Rev. B3, 2030 (1971). A spatially homogeneous mixing was not considered, in contrast to ref.’s 2, 14, 15, and 15a.Google Scholar
  20. 17.
    F. Holtzberg and J. Torrance, AIP Conf. Proc. no. 5, 860 (1971).Google Scholar
  21. 18.
    The interpretation of ref. 17 followed the earlier work on the analogous situation in Eu compounds, by M.J. Freiser, S. Methfessel, and F. Holtzberg, J. Appl. Phys. 39, 900 (1968).CrossRefGoogle Scholar
  22. 19.
    J.W. McClure, J. Phys. Chem. Sd. 24, 871 (1963);CrossRefGoogle Scholar
  23. 19a.
    A. V. Golubkov, E.V. Goncharova, V.P. Zhuze and I.G. Manilova, Sov. Phys. Sol. St. 7, /1963/ (1966).Google Scholar
  24. 20.
    H.L. Davis, Proc. 9th Rare Earth Res. Conf., Va. Poly. Inst. and State Univ., Blacksburg, VA (1971).Google Scholar
  25. 21.
    K. Lendi, Phys. Cond. Matt. 17, 215 (1974).CrossRefGoogle Scholar
  26. 22.
    D.L. Greenaway and G. Harbeke, Optical Properties and Band Structure of Semiconductors, Pergamon Press (1968), p. 114.Google Scholar
  27. 23.
    A. Jayaraman, P.D. Dernier, and L.D. Longinotti, High Temperatures, High Pressures 7, 1 (1975).Google Scholar
  28. 24.
    The lowering of 0.6 ev. follows from T. Penney and F. Holtzberg, Phys. Rev. Lett. 33, 165 (1975), with ΔV/V — 12% across the transition.Google Scholar
  29. 25.
    Estimates were also made with a circular density of states, resulting in only minor changes.Google Scholar
  30. 26.
    C.M. Varma and V. Heine, Phys. Rev. B11, 4763 (1975).Google Scholar
  31. 27.
    One might worry about a possible discontinuous electron rearrangement as a function of V, such as that occurring in the Falicov-Kimball theory (Phys. Rev. Lett. 22, 997 (1969)), It is argued in Appendix 1 that this band theory is probably inconsistent with the stability of Sm2+ at atmospheric pressure.Google Scholar
  32. 28.
    L.L. Hirst, J. Phys. Chem. Solids 35, 1285 (1974).CrossRefGoogle Scholar
  33. 29.
    C.M. Varma and Y. Yafet, Phys. Rev. B13, 2950 (1976).Google Scholar
  34. 30.
    J. Hubbard, Proc. Roy. Soc. (London) A276, 238 (1963).Google Scholar
  35. 31.
    D. Penn, Physics Lett. 26A, 509 (1968).CrossRefGoogle Scholar
  36. 32.
    A marked difference between Ni and SmS is that the narrow band (3d) in Ni is much broader than that (4f) in SmS. This suggests that interactions neglected in the Varma-Yafet29 Hamiltonian (e.g. interband Coulomb interactions) which, if properly treated, give a tendency for exciton formation might be crucial in understanding the difference. Haldane (ref. 1, p. 191) seems to have taken a small step in the direction of including such interactions (still within a single-impurity model).Google Scholar
  37. 33.
    Like those considered by P.W. Anderson and A. Hasegawa, Phys. Rev. 100, 675 (1955).CrossRefGoogle Scholar
  38. 34.
    CM. Varma, ref. 1, p. 201.Google Scholar
  39. 35.
    J.B. Sokoloff, Phys. Rev. B2, 779 (1970);Google Scholar
  40. 35a.
    G. Beni, T. Holstein, and P. Pincus, Phys. Rev. B8, 312 (1973).Google Scholar
  41. 36.
    Lest one be misled, one should realize that the same (U → ∞) Hubbard model, for certain 3-dimensional lattices, is probably ferromagnetic (Y. Nagaoka, Phys. Rev. 147, 392 (1966)), at least for ρ ≅ 1.CrossRefGoogle Scholar
  42. 37.
    J.M. D. Coey, S.K. Ghatak, M. Avignon, and F. Holtzberg, Phys. Rev. B14, 3744 (1976); ref. 1, p. 211.Google Scholar
  43. 38.
    I. Nowik, ref. 1, p. 261.Google Scholar
  44. 39.
    I. Nowik, ref. 1, p. 167.Google Scholar
  45. 40.
    Nowik’s result is that the electron released in Sm2+ → Sm3+ is not felt at the Eul53 probe. His conclusion38 that this implies that the electron is in a d-band seems strained in light of the fact that the electron released in R2+ → R3+ where R = Y, Gd in Sm1-xRXS is felt at the Eu153 and presumably goes into the same (d) band as the electron in Sm2+ → Sm3+ + e. The remarkable result is the difference (observed at the Eu153 nucleus) between the electron in Sm2+ → Sm3+ + e and that in R2+ → R3++e. As noted by Nowik38,39, consistent with his result is localization of the extra Sm electron at that Sm site (postulated by us, and recently claimed to be found by Coey et al., ref. 37). Understanding the difference between Sm3+ and Gd3+ or Y3+ is a fundamental problem. A possible source is the charge asphericity of the Sm3+ core (4f5), not present in the other ions.Google Scholar
  46. 41.
    G. Güntherodt et al., ref. 1, p. 321.Google Scholar
  47. 42.
    Unfortunately, the low-p transition is found at 10 kbar in ref 41, in disagreement with the consensus of earlier results,3,410 introducing some quantitative uncertainty in the experimental situation. It is important to have this resolved.Google Scholar
  48. 43.
    We note here that anharmonicity, particularly in E(0,V), that might begin to become appreciable at these higher pressures goes in the direction of reducing p’O.Google Scholar
  49. 44.
    S.M. Shapiro, R.J. Birgeneau and E. Bucher, Phys. Rev. Lett. 34, 470 (1976).CrossRefGoogle Scholar
  50. 44a.
    Obtaining the qualitatively correct picture (fig. 5) within the simple theoretical model we’ve presented suggests the nonexistence of a new semiconducting phase (which had been suggested, and labelled B’3). We add the comment that the calculations of this phase boundary were also carried out for rectangular and elliptical densities of states — this resulted in no important changes.Google Scholar
  51. 45.
    R.L. Melcher, G. Guntherodt, T. Penney, F. Holtzberg, 1975 Ultrasonics Symposium Proc. IEEE Cat. #75 CHO 994–45A, pg. 16.Google Scholar
  52. 45a.
    P.D. Dernier, W. Weber, L.D. Longinotti, Phys. Rev. B14, 3635 (1976).Google Scholar
  53. 46.
    If one neglects the term -k(.3 1n.3-K7 In. 7) ≅.6k which arises from the different spatial configurations of 30%f6and 70%f5. The latter term is appreciable in the difference AS, and increases the difficulty for the band theories.Google Scholar
  54. 46a.
    D.H. Templeton and C.H. Dauben, J. Am. Chem. Soc. 76, 5237 (1954).CrossRefGoogle Scholar
  55. 47.
    N.P. Bogoroditskii, V.V. Pasynkov, Rifat Rizk Basili, and Yu. M. Volokobinskii, Sov. Phys. Doklady 10, 85 (1965).Google Scholar
  56. 48.
    J.W. Allen, private communication.Google Scholar
  57. 49.
    This was found by E. Carlson (private communication) in a Hartree-Fock Calculation for Sm2+ in the excited configuration 4f5d.Google Scholar
  58. 50.
    The numerical value of the broad-band width used in Sections II and III is rather larger than that (~2.5 eV) used in Sec. 1. The essential reason for the difference is as follows: The smaller value is based on the picture from the band calculation (Davis, ref. 20) where the conduction-band minimum is at the X-point, with a local curvature corresponding to a nearly-free-electron mass; due to the symmetry-reiated minima, the density-of-states is 3 times that of a Γ-minimum with the same mass. The difference is completely unimportant for the thermodynamic properties considered in Sec. II. Although it wasn’t necessary to choose the (less realistic) larger value for these considerations, we did so to obtain symmetry with the simplified model used in Sec. III to crudely estimate optical properties. It will also be noted that the width of the narrow band,.047 eV, is rather larger than that quoted in ref. 2 (.033 eV); this discrepancy is due essentially to an error made in the earlier work. The value.047 eV is correct and depends only on the low-T specific heat (γT) and the value of b; for parabolic density of states it is easy to see that D1 = (1-b/2)1/3 π2k2Na/γ where k=Boltzmannfs constant, Na = Avogadro’s no. and γ is per mole. None of our qualitative conclusions are affected by these changes.Google Scholar
  59. 50a.
    Comparison with experiment9 of ε1(ω)), as calculated above (i.e. from Eq.(6)), indeed suggests that there are large additional contributions in the range.1 to.8 eV (roughly), which might be due to these intraatomic excitations.Google Scholar
  60. 51.
    D. Sherrington and S. von Molnar, Sol. St. Comm. 16, 1347 (1975).CrossRefGoogle Scholar

Copyright information

© Plenum Press, New York 1978

Authors and Affiliations

  • T. A. Kaplan
    • 1
  • S. D. Mahanti
    • 1
  • Mustansir Barma
    • 2
  1. 1.Physics DepartmentMichigan State UniversityEast LansingUSA
  2. 2.Tata Institute of Fundamental ResearchColaba, BombayIndia

Personalised recommendations