In this paper we have introduced a simple model which is believed to describe the relevant influence of a boundary on the polariton waves. Our treatment leads to the condition that an appropriate linear combination of P and 8P / 8x vanishes at the surface. We have also extended the theory of bulk waves and boundary conditions to include two or more dispersive resonances.
Original Description:
Original Title
Focusing of Phonons in Crystalline Solids due to Elastic Anisotropy
In this paper we have introduced a simple model which is believed to describe the relevant influence of a boundary on the polariton waves. Our treatment leads to the condition that an appropriate linear combination of P and 8P / 8x vanishes at the surface. We have also extended the theory of bulk waves and boundary conditions to include two or more dispersive resonances.
In this paper we have introduced a simple model which is believed to describe the relevant influence of a boundary on the polariton waves. Our treatment leads to the condition that an appropriate linear combination of P and 8P / 8x vanishes at the surface. We have also extended the theory of bulk waves and boundary conditions to include two or more dispersive resonances.
s4s2 T.
plained by incorporating nonlocal loss mecha~
nisms.
Vil. SUMMARY
In this paper we have introduced a simple model
which is believed to describe the relevant influence
of a boundary on the polariton waves, In contrast
to the considerations in Ref. 2, we do not neglect
the coupling to light in the boundary-condition prob
lem. Generally, our treatment leads to the condl-
tion that an appropriate linear combination of P
and aP/@r vanishes at the surface [see Eq. (16)].
However, when inserting actual parameters of erys-
‘8. 1, Pekar, Zh. Eksporim. 1 Teor. Fiz, 33, 1022
0957) (soy. Phys. JETP 6, 785 (1958)),
83, J, Hoplield and D. G, Thomas, Phys. Rev. 192,
563 496s),
81, J, Hopi, J. Phys. Soc. Japan, Suppl, 21, 777
0966),
v. M, Agranovich and V. 1. Ginsburg, Spatial Dis-
persion in Crystal Optics and the Theory of Excitons
(intorscience, Joba Wiley and Sone, London, 1960),
‘c. W, Deutsche and C. A, Mead, Phys. Rev. 158,
Ass (1965),
"J, J, Hopleld, Phys. Rev, 112, 1955 (1958).
1D, G, Thomas, J. Phys. Chem. Solids 15, 86 (1960),
SKETTRUP AND 1.
BALSLEV a
talline resonances we find that P0 at the surface
ian adequate condition throughout the spectrum.
‘The influence of the thickness a of the surface layer
has been treated strictly in the long-wave limit.
‘Thus, if a polariton mode carrying a considerable
part of the energy has a wave vector comparable
to 1/a, then a more elaborate calculationis neces-
sary. We have algo extended the theory of bulk
‘waves and boundary conditions to include two or
more dispersive resonances, Finally, we have
demonstrated the agreement with experiment of the
boundary condition in case of the exciton Lines in
200.
‘Hyg, Cardona, J. Phys. Chom. Sols 24, 1549 (1963)
(0-090).
“p, B, Dietz, J. J. Hopfield, and D. G, Thomas, J.
Appl. Phys, $2 Suppl. 2282 (961), (Wf=1,5m, If
‘corrected for the new value of the electron ma
25m).
“Sp,” sogall, Phys. Rev. 163, 769 (1967) (AT=2m)-
‘Tho time-saving approximations introduced in pro~
lows computations of exciton spectra in Zn0 (see Ref. 13)
have not been applied here. Consequently, we obtain
somewhat different values for oscillator strengths and
resonance frequencies,
itp. Skettrup, Solid State Commun, 1, 869 (1965).
'Y, 8, Pankand J. R, Schneider, J. Appl. Phys. 29, ‘ir, skettrup and I. Balatev, Phys, Sia, Solid 40,
049 (2966), 93 4970).
45, Pitinskt and 7, skettrup fenpublished).
PHYSICAL REVIEW B VOLUME 3, NUMBER 4 15 FEBRUARY 1971
Focusing of Phonons in Crystalline Solids due to Elastic Anisotropy*
B. Taylor, H. J. Maris, and C. Elbaum
Department of Physics, Brown University, Providence, Rode Island 02012
(Received 14 September 1970)
Large differences (ap toa factor of 100) have been observed in the intensity of phonons of
ilferent polarizations propagating bellistially in LiF, KCI, and ALO. These observations
‘were made by means of heat-pulse experimenta carried out on single crystals, in the tempera
{ure range 1.5-8.5°K. The results axe explained in terms of phonon focusing evo to the {Bet
tht in elastically anisotropie erystals the phonon phase and group velocities are, in general, not
collinear. Caleulations ofthe focusing effect, to detormine the relative intensities of phonons
‘of each polarization in many erystal dizections, have been curried out for a number of solid
The results ofthese calewations are in good agreoment with the experimental results obtained
obtained by other investigators on Si, Ce, and NaF,
{im this study, as well as with tho
1 INTRODUCTION
1m heat-pulse experiments, * phonons are gener
ated in a thin metallic film on one face of a crystal
by means of short electrical, microwave, or laser
pulses. The phonons are detected, alter propagat-
ing through the erystal, by means of a suitable bo-
lometer. Metallic tllms which undergo a super-
cconducting-to-normal transition when the phonon
energy arrives are very effective detectors, For
heat-pulse experiments, a crystal with phonon mean
free path comparable to its linear dimensions is,
required so that the phonons will propagate ballisti-
cally rather than diffusively. Pulses due to phonons
of different polarizations then arrive at the detector
at different times and therefore ean be studied in-
dependently, These experiments have considera
ble advantages over thermal-conductivity studies3 FOCUSING OF PHONONS IN
a5 a means of investigating phonon-phonon and pho-
non-