Volume 7, Number 12 APPLIED PHYSICS LETTERS
have been repeated several times and examination
of other observations have been made to confirm
the findings described. From this exploration it has
been determined that the amplitude of the return
from scatterers in the atmosphere is strongly de-
pendent upon the angle between the incident laser
radiation and incident solar radiation.
As yet no physical explanation for these variations
has been determined. Close examination of sur-
rounding geography and terrain has apparently
eliminated a geographical explanation.
It has been suggested that a physical orientation
of the individual scatterers is produced by the pres-
sure of the sunlight and/or alignment of scatterers
having dipole moments with the electric vector
of the solar radiation. Such orientation would
seemingly however be difficult to maintain against
the random influences of thermal and mechanical
turbulence.
This variation of backscatter intensity may be in-
fluenced by the Raman effect or a similar effect.
The vertically polarized laser radiation may be at-
tenuated or shifted in wavelength when transmitted
DIRECT PIEZOELECTRIC COUPLING TO SURFACE ELASTIC WAVES
(CdS, fused silica, and quartz; transducers; E)
The methods used previously to generate and de-
tect surface elastic waves piezoelectrically! have in-
volved the mechanical coupling of a compressional
or shear wave transducer to the body on which the
surface waves are to propagate. We report here di-
rect piezoelectric surface wave transduction by a
spatially periodic electrode on the plane surface of a
piezoelectric plate. A periodic electric field is pro-
duced when an rf source is connected to the elec-
trode, thus permitting piezoelectric coupling to a
traveling surface wave. Direct coupling has a number
of advantages described below. Also described here
are measurements of surface wave velocity in quartz,
and the surface wave attenuation produced by thin
cadmium sulfide films.
Direct coupling to a surface elastic wave is pos-
sible at the boundary of a piezoelectric solid if any of
the components of strain at the surface is piezoelec-
trically active. For example, consider a wave prop-
314
Downloaded 26 Sep 2010 to 143.248.72.176. Redistribution subject to AIP license or copyright; see http://apl.aip.org/about/rights_and_permissions
15 December 1965
parallel to the plane of the electric field from the
solar radiation. A shift of 4 or 5 A could be sufficient
to put the return radiation outside of the narrow
bandpass of the receiver filter. The opposite effect
observed when transmitting at right angles to the
plane of the electric field from solar radiation might
produce the high level of back scattered energy re-
ceived.
These anomalies are being investigated to deter-
mine, if possible, the exact nature of the effect of
solar radiation on atmospheric laser returns. If
other investigators have observed this effect or if
they can provide a physical explanation for these
observations, the authors would appreciate infor-
mation on such observations or explanations.
1 G. G. Goyer and R. Watson, Bulletin of the American Meteoro-
logical Society, 44, No.9 (1963).
2 D. Atlas, Bulletin of the American Meteorological Society, 44, No.
12, (1963).
3 M. Ligda, "Meteorological Observations with Lidar", Stanford
Research Institute report at Proceedings of First Conference on Laser
Technology, Vol. II, Office of Naval Research, Boston, Mass.
R. M. White and F. W. Voltmer
Department of Electrical Engineering, University of California
Berkeley, California
(Received 14 October 1965)
agating along the top of the crystalline quartz plate
of Fig. I a. There will be components of particle dis-
placement in the x and y directions; both compo-
nents vary with x, becoming negligibly small at a
Pulsed
Oscillator Receiver
Quartz
Fig. lao Arrangement for surface wave transduction by
electrodes on crystalline quartz bar.
 Volume 7, Number 12 APPLIED PHYSICS LETTERS
distance of about one wavelength beneath the sur-
face. The xy strain component (56) is coupled piezo-
electrically to the y component of electric held, and
the xx(5 l ) strain is coupled to the x component of
electric field. 3 Thus an oscillator or receiver con-
nected to an electrode on the quartz plate having the
configuration shown in Fig. 1b will be coupled piezo-
electrically to a propagating surface wave. The
period, p, of the pattern is made equal to the sur-
face wavelength at the frequency of operation. The
electrode pattern will also operate at integral mul-
tiples of its fundamental frequency.
This method of surface wave transduction some-
what resembles the use of the comb transducer,1.2
for both involve the creation or detection of spa-
tially periodic surface displacements. (The comb
transducer consists of a block, typically of metal,
upon whose plane upper surface a compressional
wave transducer is mounted and whose serrated
lower surface provides spatially periodic contact to
the body on which surface wave propagation oc-
curs.)
We have observed the direct piezoelectric cou-
pling described above at 15 and 45 mc with the ar-
rangements illustrated in Fig. 1a. The electrode
pattern used at both frequencies (see Fig. 1b) was
produced byo photoetching in a dilute NaOH solu-
tion a 1000-A thick aluminum film which had been
vacuum-deposited on the quartz bar. Standard
photo-resist processes were used to mask the sur-
face prior to etching. The electrode pattern was
For connection
to oscillator
or receiver
w most surface wave transducers) were measured in
Fig. lb. Electrode pattern used in (a) at 15 and 45 mc. Ap-
proximate dimensions in inches: p = 0.008; W = 0.095; L =
0.079.
Downloaded 26 Sep 2010 to 143.248.72.176. Redistribution subject to AIP license or copyright; see http://apl.aip.org/about/rights_and_permissions
15 December 1965
used either as a source or as a receiving transducer,
while a comb transducer mounted elsewhere on the
bar served as the companion transducer for trans-
mission measurements. Pulses from the receiving
transducer were identified as being surface waves
by their phase velocity as determined from the trans-
ducer separation and the pulse transit time, and bv
the large attenuation of the pulse produced when a
drop of acetone was placed on the surface between
transducers. It is also possible with rf input pulses of
long duration to observe the characteristic trape-
zoidally shaped received pulse resulting from the
finite time of transit of the wave across an extended
transducer. l
The pulse amplitude obtained using one electrode
pattern and one comb transducer was about the same
as that obtained using two identical comb tram-
ducers for generation and detection. The relative
responses of the two types of transducer depend
upon many factors including the number of periods
of each transducer, the type of piezoelectric ma-
terial used with each, and the means used to couple
the comb transducer to the surface (e.g. pressure
alone or pressure together with an oil film).
The advantages of direct piezoelectric coupling
include the following: it does not require contact
pressure and so produces no surface damage or
distortion; the transducer may be operated at ex-
treme temperatures or in a vacuum; the transducer
may be easily handled and precisely fabricated: and
the electrode pattern may be extended in the di-
rection of wave propagation so as to increase the
active transducer area and the energy density of the
surface wave without resulting in the substantial
coupling of energy back away from the surface which
results when comb transducers are made very long l
The most apparent disadvantage is the limitation
to suitably oriented piezoelectric materials. How-
ever, both the electrode pattern and a piezoelectric
him (e.g. cadmium sulfide) could be deposited on
a non piezoelectric substrate.
Surface wave velocities (necessary for design of
fused silica and in crystalline quartz. Velocities were
computed from the measured change in phase delay
of a pulse produced by one comb transducer as a
second comb transducer was moved a known dis-
tance along the surface. The values obtained were:
fused silica, velocity 3.37 ± 0.08 X lOs cm/sec:; quartz,
propagation along surface parallel to x axis, normal
to y axis, velocity 3.20 ± 0.06 X 105 cm/sec; quartz,
propagation along surface parallel to y axis, normal
to x axis, velocity 3.25 ± 0.06 X 105 cm/sec.
3Fi
 Volume 7, Number 12 APPLIED PHYSICS LETTERS
Surface wave propagation and piezoelectric inter-
action with electrons in piezoelectric semiconductors
has been considered and preliminary measurements
have been made of the attenuation of a surface wave
propagating along a fused silica plate on which a
very thin cadmium sulfide film has been deposited.
Substantial attenuation was observed. For films
ranging in thickness from 2 to 5 p. and having elec-
trical resistivities between 200 and 2000 O-cm, at-
tenuation as high as 80 dB/cm was measured at 15
mc even though the ratio of film thickness to wave-
length (in the fused silica) was 2.2 x lO-2 or less. In
contrast, attenuation less than I dB/cm was pro-
duced by deposited aluminum films of comparable
thickness. The origin of the high loss in the CdS films
is not yet clear. The loss is not determined entirely
by film resistivity as a change in illumination pro-
duced virtually no change in loss even though the
films were moderately photoconductive.
Stoneley4.5 has shown that true surface wave prop-
agation is not always possible in an anisotropic solid.
It is interesting to note that the elastic constants of
CdS satisfy both the criterion developed by Stoneley4
for surface wave propagation along a plane normal
to the hexagonal axis, and the analogous criterion
for propagation along a plane parallel to that axis.
THE MILLING OF SOFT METALS TO PRODUCE
VERY LARGE LATTICE STRAINS
(AI and Pb; cold working; AI,03 used; E)
During recent years many studies have been made
of the effects of cold work on the crystallographic
properties of metals, utilizing the line broadening
of x-ray diffraction patterns to analyze the results.
In the course of these studies the behavior of differ-
ent metals has been compared and the influence of
such factors as temperature and impurities has been
observed both in the process of cold working and
during the subsequent removal of the line broaden-
ing by annealing.
The mechanism which has been used for cold
working in these investigations has been filing,
partially because of the ease with which it can be
carried out in different environments and also since
316
Downloaded 26 Sep 2010 to 143.248.72.176. Redistribution subject to AIP license or copyright; see http://apl.aip.org/about/rights_and_permissions
15 December 1965
Thus in CdS, the piezoelectric coupling of electrons
to surface waves should be possible with either of
these orientations. The velocity for the first orienta-
tion, as computed from Stoneley's determinantal
equation upon substitution of elastic constants for
CdS,6 is 1. 70 x 1Q5 cm/sec. Experiments with depos-
ited electrode transducers on CdS single crystals
are in progress. Electrically nonconducting glass
comb transducers (photoetched in HF after appli-
cation of KMER Photo-Resist with Additive D) are
also being used in this study.
This work was supported in part by the U.S. Army
Research Office-Durham under Grant DA-ARO-
D-31-124-G678. The authors are grateful also to
K. Dransfeld and R. M. Arzt for substantial assist-
ance, and to Mrs. D. McDaniel who photoetched the
transducers.
'I. A. Viktorov, Sov.-Phys. Acoustics 7,236 (1962).
2 R. M. Arzt and K. Dransfeld, Appl. Phys. Letters 7, 156 (1965).
3 H. B. Huntington, The Elastic Constants of Crystals (Academic
Press, New York, 1958).
4R. Staneley, Geophys. Suppl. to Monthly Notes, Roy. Astron. Soc. 5,
343 (1949).
5 R. Staneley, Proc. Roy. Soc. 232A, 447 (1955).
BD. Berlincourt, H. Jaffe, and L. R. Shiozawa, Phys. Rev. 129,
1009 (1963).
D. Lewis and H. Pearson
Battersea College of Technology
London, England
(Received 22 September 1965)
the end product is in a form which can be used most
readily for diffractometer specimens. However, the
effects of filing, so far as plastic deformation is con-
cerned, vary widely from one metal to another and
appear to be largely related to the recrystallization
temperature of the metal. Taking the extreme ex-
amples of tungsten (2,400°F) and aluminum
(175°F), Williamson and HalF have obtained strain
values of 0.62% and 0.045% respectively. The value
for aluminum was increased to 0.14% by impurities
and also by filing the pure specimen in liquid
nitrogen.
Since the magnitude of strain produced by filing
varies so widely, it cannot be regarded as an ideal