aLandolt B

1
Micro-Raman studies of zincblende MgS
Th-P-26
D Wolverson, L. C. Smith, University of Bath, UK
d.wolverson_at_bath.ac.uk C. Bradford, B. C.
Cavenett, K. A. Prior Heriot-Watt University, UK

• Zincblende MgS is a wide bandgap semiconductor
  (5eV).
• MgS can now be grown by MBE (in the zincblende
  structure) close to lattice-matched on GaAs.
• Main conclusions of this work
• Resonant micro-Raman scattering has been carried
  out on ZB MgS with improved spectral range
  (improved dielectric filters)
• Good agreement with calculated phonon dispersion
  and DOS.

• Experimental observations
• Overtones of the LO phonon seen at 849, 1261 and
  1674 cm-1 (n 24 respectively) they are strong
  because MgS is highly polar and the Raman
  cross-section depends quadratically on the
  polaron coupling constant
• Fitting the overtones shows that their widths are
  proportional to n whilst their peak positions lie
  below nth multiples of the LO phonon by
  approximately (15n) cm-1
• This shift with n is evidence of the increasing
  participation in the multiple scattering process
  of phonons with finite wavevectors (though
  phonons close to the G point still dominate). The
  shifts of the overtone peaks are significant
  because of the substantial curvature of the LO
  phonon dispersion around the G point in ZB MgS
  (see calculated dispersion above).
• The other features that can be identified are
  those at 504 cm-1 and 759 cm-1 which are
  numerically close to the second and third
  multiples of a strong feature in the phonon
  density of states at around 250 cm-1 (see panels
  (c,d) of the comparison to experiment, above
  left).
• This feature arises from the LA phonon branch in
  the energy region of all of the X, L and W
  critical points and, since all two-phonon
  overtone scattering is Raman-active in zincblende
  crystals, 2LA and 3LA processes offer a possible
  explanation of these bands.
• However, scattering by a sum of G point LO and TO
  phonons (predicted Raman shift 773 cm-1) gives
  another possible interpretation of the band at
  759 cm-1.
• Finally, a weak band (marked ) can be see in in
  the experimental spectrum (panel (a), above
  left) with Raman shift 692 cm-1, assigned to the
  2nd overtone of the G-point TO phonon (predicted
  frequency 688 cm-1).

• Calculations
• PWSCF code (www.pwscf.org)
• Local density approximation
• BHS pseudopotentials for Mg and S of
  norm-conserving type
• Exchange-correlation effects accounted for via
  the Perdew-Zunger parametrization
• Non-linear core correction was included for Mg
  (the Mg 2p states were here not included in the
  valence states)
• (6,6,6) Monkhorst-Pack grid
• Convergence was checked with respect to this and
  to the kinetic energy cutoff (30 Ry)
• Lattice parameter found from the minimization of
  the total energy per unit cell 10.65 Bohr
  (5.633ºA) (figure, above left)
• experimental values range from 5.66ºA to 5.622ºA
• Phonon dispersion and density of states
  calculated via PWSCF.

Calculated phonon frequency Measured value
TO 344 cm-1 344 cm-1 (1 cm-1)
LO 430 cm-1 429 cm-1 (1 cm-1)
a Landolt Börnstein b RWG Wyckoff, Crystal
Structure Wiley, New York 1963) c G Kalpana, B
Palanivel, RM Thomas, M Rajagopalan, Physica B
222 (1996) 223 d L Konczewicz, P Biegenwald, T
Cloitre, M Chibane, R Ricou, P Testud, O Briot,
RL Aulombard, J. Crystal Growth 159 (1996)
117 e C Verié, J. Electronic Materials 27 (1998)
782 f C Bradford, CB O'Donnell, B Urbszek, A
Balocchi, C Morhain, KA Prior, BC Cavenett, Appl.
Phys. Letts. 76 (2000) 3929 g For the NaCl
structure, d a/2 h Y Morinaga, H Okuyama, K
Akimoto Japan J. Appl. Phys. 32 (1993) 678 i U
Lunz, C Schumacher, J Nürnberger, K Schüll, A
Gerhard, U Schüssler, B Jobst, W Faschinger, G
Landwehr, Semicond. Sci. and Technol. 12 (1997)
970 j I Suemune, H Uesugi, H Suzuki, H Nashiki,
M Arita, Phys. Status Solidi B 202 (1997) 845

• Conclusions
• Best Raman data yet obtained on ZB MgS, and
• good agreement with detailed theoretical
  calculations.

Agreement good - better than expected from the
LDA and linear response approach - fortuitous!