Bull. Mater. Sci., Vol. 6, No. 2, May 1984, pp. 317-325. (.©Printed in India.

Pyroelectric materials

M R SRINIVASAN
Materials ScienceGroup, Tam Institute of FundamentalResearch,Bombay400005, India
Abstract. Thispaper presentsa reviewof pyroelectricmaterials.The physicsof pyroclcctric
materials is described.Variouspyroelectricmaterialsare compared for their efficiencyin the
detection of radiation. Current trends in pyroelectricmaterialtechnologyare highlighted.
Keywords. Pyroelectricity,ferroelectricity;piezoelectricity.

1. Introduction

It has been known for a long time that some crystals exhibit electric polarization when
their temperature is changed. This phenomenon is known as pyroelectricity and arises
out of the temperature dependence of spontaneous polarization in these crystals. The
pyroelectric polarization, however, does not persist for a long time and gets neutralized
due to imperfect insulation and the presence of free charges in the crystal. While all the
ferroelectric crystals are necessarily pyroelectric, the converse need not be true. A built-
in asymmetry in the crystal structure hindering the polarization reversal or a high
coercive field exceeding the limit of electrical breakdown might prevent a pyroelectric
from being a ferroelectric.
Quantitatively, the pyroelectric effect could be described by the equation
AP¢= plAT,
where AT denotes the change in temperature, AP~(i= 1, 2, 3) are the changes in the
components of polarization P in the three-dimensional space and Pi are the
corresponding pyroelectric coelIicients, p is a vector property and by Neuman's
principle ought to remain invariant under all the symmetry operations of the crystal
(Nye 1962). Hence pyroelectricity can be exhibited only by crystals belonging to the ten
polar classes 1, 2 m, 2 mm, 3, 3 m, 4, 4 mm, 6 and 6 mm.
It may be pointed out here that in some crystals polarization may appear under the
influence of hydrostatic pressure and this is a special case of piezoelectric effect. Since
hydrostatic pressure is a scalar-like temperature this polarization linearly proportional
to the hydrostatic pressure can also be treated on par with pyroelectricity. Thus the
piezoelectric effect under hydrostatic pressure can occur only in the above mentioned
ten polar classes. In general the piezoelectric effect has the mathematical character of a
third rank tensor and is restricted only to 20 non-centrosymmetric classes excluding the
class 432. Thus the symmetry permits all the pyroelectric crystals to be piezoelectric
while the converse is not true. Therefore it is possible to imagine two possible
contributions to pyroelectricity in the following way. Experimentally to observe
pyroelectricity one can heat the crystal and observe the change in polarization. The
experiment could be performed in two ways. Either the shape and size of the crystal can
317
318 M R Srinivasan

be kept fixed during the heating or the crystal may be released so that thermal
expansion can occur quite freely. Obviously the magnitude of the effect observed in-the
two cases will be different. In the first case the crystal is clamped and the observed effect
may be regarded as the primary pyroelectricity. In the second case, in addition to the
primary etfect, there is pyroelectric effect due to the variation of piezoelectrically-
induced polarization with temperature. This is known as the secondary pyroelectricity.
The secondary pyroelectricity is found to contribute substantially to the total effect.
Hence for a free crystal the total pyroelectric effect could be written as
(dP/dT)o, • = (d P/dT)s, e + (d P/dS)r,E (dS/dT)~,, E (1)
where o, S and E denote the stress, strain and electric field respectively. Equation (1)
could be written as
p~.e = pS.~+ (d P / d o ) r ' ~ ( d o / d S ) r "~ (dS/dT)o, E
= pS, E + AT,
, , i ~ ,F.rT,
~ j ~ , , ~E' j ~~O,E , (2)

where pO,r is the total pyroelectriccoefficient;pS,~ is the primary pyroelectric coefficient;

do~, C~nm and ~ denote respectively the piezoelectric moduli, elastic compliance
coefficients and coeffic;.ents of thermal expansion. Clearly the secondary pyroelec:ric
coefficient is given by the product of dijh, Cjnm and ~ . Though secondary
pyroelectricity is due to piezoelectricity, only those piezoelectric crystals which belong
to the ten polar classes are permitted by crystal symmetry to exhibit secondary
pyroelectricity.
So far we have tacitly assumed that the temperature of the crystal is the same at all
points. Uneven heating causes temperature gradients, which by thermal expansion give
rise to non-uniform stresscs and s~raip,s. Under such conditions it is possible for
piezoelectric crystals like ~-quartz, which does not belong to the ten polar classes, to
exhibit secondary pyroelectricity. This secondary pyroelectricity due to non-uniform
heating is.called the tertiary pyroelectric effect. The tensorial pyroelectricity refers to
the production of quadrupole or higher electric moments on heating.

2. Theory of pyroelectricity

In principle, in order to understand pyroelectricity in any material, one has to consider

the various mechanisms of the spontaneous polarization (such as ionic, electronic,
orientational or surface charge) and study their variation with temperature. The
pyroelectric behaviour is fully described by the temperature variation of pyroelectric
coefficient. Actually, it is convenient to consider the temperature variation of primary
and secondary pyroelectric coefficients separately. The temperature variation of
secondary pyroelectric coefficient is governed by those of piezoelectric moduli, elastic
compliance constants and thermal expansion coefficients. Both piezoelectric moduli
and elastic compliance constants are not strong functions of temperature, in those
crystals which do not undergo any phase transition. Hence the temperature variation of
secondary pyroelectric coefficient is akin to those of thermal expansion coefficients.
Conceptually, the origin of primary pyroelectric effect and its variation with
temperature is more difficult to visualise and has invoked considerable discussion in the
literature.
 Pyroelectric materials 319

The lattice dynamical theory of primary pyroelectricity for the ease of ionic crystals
was first formulated by Max Born in the year 1945 (Born 1945). In this paper, Born had
indicated that primary pyroelectric coefficient would be proportional to T, though in his
later treatise on lamce dynamics he predicted the T a law for the pyroelectric coefficient
(Born and Huang 1954). Recently Szigeti has brought out clearly many salient features
of the theory, especially the r~,le of mechanical and electrical anharmonicity in primary
pyroeleetricity which in fact was overlooked by Born (Szigeti 1975, 1976).
In the case of ionic crystals, there are two important mechanisms of polarization.
One being responsible for absorption in the infrared i.e. the lattice or ionic polarization
and the other in the ullraviolet i.e. the electronic polarization. In the very simplest and
crude model known as the rigid ion model, the electron cloud around the ion is assumed
to be rigid and consequently there is no contribution from electronic polarization. For
such a model, the total dipole moment of the crystal is given by

M = ~e,Q°,
a

where Q, are the active normal coordinates i.e. those which produce uniform
polarization in the direction of M. :q are suitable normalizing constants.
However, in reality the electrons are deformed during the lattice vibrations as they
experience short range forces and also the dipolar field due to ions. The electron
deformation is not just linearly proportional to lattice displacements but involves
higher terms as well. Hence when one takes into consideration the electronic
polarization also, the dipolemoment of the crystal is given by

M = ~_~~taQa + ~ fljj, QjQ), +. (3)

a jj

An important point to note is that % includes both the effects due to lattice
displacements and also the first order effects due to electron deformation. The above
expression for the dipolemoment clearly implies the presence of electrical anharmo-
nicity in the crystal.
For the ease of harmonic crystal, the potential energy is proportional to the square of
the normal coordinates. In reality crystals are not harmonic and the mechanical
anharmonicity has to be taken into account. For such a crystal the potential energy
involves the cubic and higher powers of the normal coordinates

w = eoj Qj + ~ bij.r,QiQj, Qj,, +. (4)

j jj'j"

The primary pyroelectric coefficients/~.E could be expressed in terms of coefficients

which appear in the expansion of dipolemoment and potential energy (i.e. (3) and (4)).

~j( ~°~abajj'~CJ (5)

Here Cj is the contribution of the jth mode to the specific heat and is given by
~ioJ~(O~j/OT), ~j being the average occupation number of the phonons with energy hwj.
From the above expression, it could be shown that under Debye's approximation
(which actually involves invoking some properties of fljj and boj~ it., ,',helong wavelength
limit), the primary pyroelectric coefficient obeys T 3 law. But as ha~ been rightly pointed
320 M R Srinivasan

out by Szigeti, in the case of complex polyatomic crystals, optic modes can have very
low frequencies and hence the T 3 behaviour is observed only at very low temperatures.
At this stage, it is also necessary to emphasize the role of both electrical and mechanical
anharmonicity in primary pyroelectricity. Since the primary pyroelectric effect does not
involve thermal expansion, for quite some time, it was regarded that only the electrical
anharmonicity was responsible for primary pyroelectricity. It may be mentioned that
thermal expansion is the result of mechanical anharmonicity. However, from the
expression for primary pyroelectric coefficient, it is evident that the part played by
mechanical anharmonicity cannot be neglected.

3. Pyroelectric detector and the figure of merit of pyroelectric materials

If the temperature change in a pyroelectric material is caused by the absorption of

radiation, the observation of pyroelectric effect is equivalent to the detection of the
radiation (Lecomte 1948). Basically, a pyroelectric detector consists of a thin
pyroelectric crystal oriented with the electrode surface normal to its polarization
vector. On absorbing the incident radiation the temperature of the pyroelectric material
changes, which in turn changes the electric polarization. Thus pyroelectric material
forms a capacitive element, the capacitance varying with temperature. The capacitor
will deliver a current
i = pA (dT/dt),
where P is the pyroelectric coefficient, A is the electrode surface area and (dT/dt) is the
rate of temperature change.
Since the pyroelectric current is proportional to the rate of temperature change, the
device gives a steady output only if there is continuous variation in the temperature of
the pyroelectric crystal due to modulation, pulsing or chopping of incident radiation.
When the incident radiation is not modulated, pulsed or chopped, the device gives an
output only till the time the pyroelectric crystal attains a steady temperature.
It must be noted that if tile input radiation is slow compared to the thermal
relaxation time of the crystal, the temperature essentially remains constant and the
current response is low. For input frequencies higher than the thermal relaxation
frequency full current response results because the absorbed energy goes into tile
heating of the crystal. The thermal relaxation frequency is given by
fthermal = G / 2 n n ,
where G is the thermal conductance between the crystal and the surroundings and H is
the total heat capacity of the crystal. For a given crystal material and size, the thermal
relaxation frequency is governed by conductance. Since the pyroelectric detector is
made to respond to rate of change of temperature, it differs fi'om other thermal
detectors. The fact that the crystal does not have to come to thermal equilibrium as in
the case of thermocouples and thermistors allows the pyroelectric de~ector to respond
faster than other thermal detectors and still be sensitive to a wide spectrum of radiation
wavelengths.
The performance of the pyroelectric detector is usually described by ~oltage
responsivity (i.e. the pyroelectric voltage output per unit energy of incident radiation)
and detectivity (i.e. the inverse of noise equivalent power (NEP) where NEP is defined as
 Pyroelectric materials 321

the incident energy required to produce a signal equal to that of detector noise). The
voltage responsivity as well as the detectivity depend not only on the particular
pyroelectric material but also on the characteristics of the amplifier employed for
detecting the pyroelectric voltage. The detailed theory of pyroelectric detector has been
discussed by several workers (Lang 1971; Liu 1978). It is sufficient for our purpose to
know that the voltage responsivity is proportional to p(ec)-1 while the detectivity is
proportional to p(c(etan6)l/2) -l, where p is the pyroelectric coefficient, e is the
dielectric constant, tan 6 is the loss tangent and c is the specific heat. Hence p(ec)- ~and
p(c(e tan ~)~/2)-~ could be regarded as the figures of merit of pyroelectric materials.
Clearly, in the case of pyroelectric materials which are also ferroelectric the dielectric
constant and the pyroelectric coefficient are strong functions of temperature and so is
their figure of merit. It could be shown that the figure of merit is maximum at the
temperature corresponding to the stability limit of the ferroelectric phase (Liu 1978).

4. Pyroelectric materials

An important component of pyroelectric detector technology pertains to the prepar-

ation of suitable pyroelectric materials. Since every pyroelectric material has its own
advantages as well as disadvantages, a wide choice of pyroelectric materials is necessary
not only for improving the detector performance but also to cater to specific
applications of detectors. With the discovery of ferroelectricity in a large number of
materials, the choice of pyroelectric materials has increased. In the beginning, the
materials that were examined for pyroelectric detection included triglycine sulphate
and barium titanate, which were among the few well-known ferroelectrics at that time.
In recent years, a number of titanates, niobates, tantalates, zirconates, germanates and
polymers have been studied for pyroelectric detection.
Apart from the study of new pyroelectric materials, another aspect of material
technology pertains to the modification of material properties with a view to enhance
the detector performance. Dielectric constant and dielectric loss are quite sensitive to
dopants and also to radiation damage and hence can be monitored to some extent.
Dopants also increase the coercive field and thus can be employed for obtaining single
domain crystals, which are superior to multidomain ones. A number of mixed
pyroelectric materials like strontium barium niobate, potassium tantalate niobate, lead
zirconate titanate etc exhibit ferroelectricity over a wide range of composition of their
constituents. In these crystals, it is possible to alter the curie temperature and
pyroelectric coefficient by varying the composition. This property also has been made
use of for improving the detector efficiency. Table 1 summarises the pyroelectric
materials studied so far along with their figures of merit for voltage responsivity and
detectivity.
Thechoice of the pyroelectric material is mainly determined by (a) its figure ofmerit,
(b) the detector size, (c) availability and durability of the pyroelectric material,
(d) environment in which the material has to operate, (e) the radiation levels to be
detected, (f) the purpose for which the detector is employed, (g) the maximum
ambient temperature of operation and the range over which stable operation is desired.
The importance of factors (b) and (c) in the choice of pyroelectric material does not
need much of an explanation. It should be possible to grow large crystals of pyroelectric
material and fabricate them into thin slices. The durability is also an important factor.
322 M R Srinivasan

Table I. Figure of merit for various pyroelectric materials.

p FM-R FM-D*
Material T(°C) ~ ( x 10 - s ) c s p ( x 10 -1°) ( x 10 -4) Reference
(1) (2) (3) (4) (5) (6) (7) (8) (9)

TGS 0 20 1'3 2'3 -- 2'4 x 10 l° 2"8 8-8 Beerman

(~ = 49°C) 25 35 4'0 2"5 1-66 1'7 x 101° 4"6 20-9 (1975)
40 100 12.0 2'8 -- 1"6 X 10 9 4"3 17.1
-25 20 0"44 -- --- 1"1 x 10 l° 0"96 2"0
DTGS 0 12 0.84 2"4 -- 5"0 x 101° 2'9 7.8 Beerman
(~ = 61°C) 25 18 2-7 2'5 1'7 5"0 x 101° 6"0 24.2 (1975)
50 60 lif0 2"8 -- 1"2 x 101° 6"0 39"2
-25 11 0"42 -- -- 1'8 x 101° 1"6 2.4
LATGS 30 35 7"0 2"55 t'7 -- 7"8 -- Garn and
(~ = 49.5°C) Sharp (1974)
DLATGS 25 22 2"5 2"55 1-7 -- 4'4 -- Garn and
(~ = 49.2°C, Sharp (1974)
61.2°C)
TGFB 30 15 2-1 1"68 1"66 -- 8'3 Garn and
(T~= 73.8°C, Sharp (1974)
(73 °C)
DTGFB 25 12 2'5 1'87 1"7 -- 11"1 Garn and
(~ = 74.5°C) Sharp (1974)
TGSE 21 400 3'0 1"79 1'7 -- 0"4 Garn and
(~ = 22°C) Sharp (1974)
BaTiO3 25 135 1"9 -- -- -- 0.47 m Beerman
( ~ = 135oc) 60 200 7"0 3.0 6"0 6'0 x 106 1.17 0"57 (1975)
100 400 20.0 -- -- -- 1'67
PbTiO3 25 142 2-7 3"2 7"78 5"3 x 10s 0"59 1"90 Beerman
(T~ = 492°C) (1975)
PbTiO3 25 200 6 3"2 7"78 -- 0"93
(Matsuohita)
(T, = 470°C)
LiNbO3 25 30 0"4 -- -- 9'8 x 101° 0"48 -- Beerman
(T~ = 1210°C) 100 31 0"5 2'8 4-64 6 x 101° 0"58 4.5 (1975)
-25 51 2"1 3"0 -- 5'3 x 10 l° 1-4 16.1
LiTaO3 0 52 2-2 3"0 -- 4"4 x 101° 1.4 15'4 Beerman
( ~ = 618°C) 25 54 2-3 3'16 7"45 3-6 x 101° 1-35 13"8 (1975)
50 56 2-5 3"3 -- 2'9 x 10 l° 1.3 12.9
100 60 2-7 3-5 -- 2'0 x l01° 1"3 10'9
PZT4 25 1400 3.7 3"0 7-5 3"2 x 10a 0-087 2'2 Beerman
(clevite) (1975)
(T, = 328°C) -25 1410 4'6 -- -- 6'2 × 107 0"109 1'21
pzro- 1306 0 1570 4-6 -- -- 5-9x l0 T 0"098 1"18 Beerman
(Gulton) 25 1730 4-6 3'0 7.6 5"7 x 107 0-089 1-16 (1975)
(To = 290°C) 50 1900 4"6 -- -- 5'4 X 10 7 0"081 1'13
100 2330 4"6 -- -- 4"9 x 107 0.066 1"07
PZT.HST-41 25 1800 2'0 3"0 7"6 4"5 x 107 0-037 0"45 Beerman
(~ = 270oc) -25 640 4.0 -- -- 7.8 × 107 0.24 1"36 (1975)
PZT5 -- 2100 4.7 3.0 7.75 -- 0"075 -- Garn and
(~ = 365°C) Sharp (1974)
 Pyroelectric materials 323

Tablel. (contd)

Material (1) (2) (3) (4) (5) (6) (7) (8) (9)
PLZT - 515 3"5 -- 8'0 -- -- -- Garn and
(0/65/35) S h a r p (1974)
i T s = 38o~c~
PLZT -- 600 3.9 -- 7"95 -- -- -- Garn and
(2/65/35) S h a r p (1974)
(T~ = 310~C)
PLZT 0 740 4.8 ..... 4-8 x 107 0.25 1.28 Beerman
(4/60/40) 25 870 5.6 2-6 7-85 3"0 x 107 0'25 1"18 (1975)
( T o = 218°C) 50 1020 6.3 . . . . 2"0 × 107 0-24 1,09
100 1400 7-6 . . . . 1 3 × 107 0-21 1.06
PLZT - 680 5.2 - - 7.90 . . . . Beerman
(4/65/35) (1975)
(T~ = 225°C)
PLZT 25 1400 10.0 2'6 7.83 5 1 × 107 0-23 2.8 Beerman
(6"5/65/35) (1975)
( ~ = 145"'C)
PLZT - 1860 13.0 -- 7.82 . . . . . Garn and
(7/65/35) S h a r p (19747
( ~ = 145~C)
PLZT - 3800 l 7.0 -- 7.80 . . . . Garn and
(8/65/35) S h a r p (1974)
(~ = lOffC)
Srtl - x}Ba~Nb~06 Beerman
(SB~) (1975)
x = 0-52 25 380 6.5 2.1 5.2 1.6 × 10 s 0"81 4-0
( ~ = 115'C)
x = 0"33 25 1800 11 ..... 1-2 x 108 0'29 5"8
(~ = 62C)
x = 0"25 -25 1600 5.6 - -- 9.0 x 10 ~ 0.17 2.5
(T~ = 30"C) 0 2500 12.0 . . . . 5.1 x 107 0.23 4.1
25 5000 37.0 . . . . 1-9 x l 0 T 0-32 7.7
50 12000 10.0 . . . . 6.5 x 106 0.04 1.2

( S r l - x B G h - 1 3 r / 2 ~ R ~ N b 2 0 6 (R: saN) w h e r e x ~ 0-5 a n d y = 0.02 Liu (1974)

La: s a n 25 1630 11.7 2-61 5.328 1.15 x 10 s 0.27 4-72

(T~ = 62"C)
N d : SaN 25 1600 11"8 2"66 5"326 1"76 x 108 0"28 5"94
(To = 68°C)
Sm: s a n 25 1050 9.2 2,39 5323 1.79 x 10 s 0"37 5"19
(T~ = 89~'C)
G d : SBN 25 854 8"35 2'34 5'328 3"21 x 10 s 0-42 6"42
(To = 92 C )
Lu: SaN 25 448 5"3 2-45 5"327 4,74 × 10 s 0-48 4"68
(T,. = 117"C)
Li2SO4H20 --25 12"1 1"0 ..... 1"3 × 1011 3"6 160 Beerman
0 12"2 1'07 ...... 1"1 x 1011 3"8 15.7 (1975)
25 12'3 1-17 2'3 2.06 9"8 × 101° 4'1 15.9
50 12.4 1"6 ..... 6"6 x 10 t° 5"6 17.9
Pb~Ge3Ol l 40 1"1 -- -- 5.10 '1 -- Watton et al
(T~ = 178~C) (19767
324 M R Srinivasan

Table 1. (contd)

Material (1) (2) (3) (4) (5) (6) (7) (8) (9)
Pb4.TsBao.25Ge3011 60 2-0 -- -- 10t4 -- -- Watton et al
(T~= 101°C) (1976)
Pb4.7oBao.3Ge3Oll 81 3'2 -- -- 10j'* -- -- Walton et al
(T, = 70°C) (1976)
Na NO2 25 7-4 0"5 2"1 2.2 -- 3"2 -- Garn and
(T, = 163°C) Sharp (1974)
Polyvinyl- -- 5"33 if01 . . . . . . . . Garn and
chloride (pvc) Sharp (1974)
Polyvinyl -- 5"0 0.10 2"3 1-38 -- ff87 -- Garn and
fluoride (PvF) 0-18 Sharp (1974)
Polyvinylidene - - 11-0 0-30 2-4 1.76 -- -- -- Garn and
fluoride (PvF2) Sharp (1974)
Polyacryl- -- 7.7 0.005. . . . . . Garn and
nitrile (PAW) 0-01 Sharp (1974)
Poly (4-nitro- -- 6.2 0-006 . . . . . . . Garn and
styrene) Sharp (1974)
atactic
Polyacry. . . . 0-01 . . . . . . . . Garn and
lamide (P^A) Sharp (1974)

(1) T(°C): Temperature in °C at which measurements are made; (2) ~: Dielectric constant;
(3) p: Pyroelectric coeflicient in Coul/cm2/K°; (4) c: Volume specific heat in J/em3/K°;
(5) s: Density in g/era3; (6) p: Ac resistivity in ohm-cm; (7) FM-R: Figure of merit for voltage
responsivity given by p(Ec)- J; (8) FM-D*: Figure of merit for detectivity given by p p~/2c- t;
(9) Ref.: Reference

Table I clearly indicates that TGS and Li2SO4 • H 2 0 are g o o d materials for pyroelectric
detection. But the main problem is their hygroscopic nature, which restricts the
operating conditions. Therefore LiTaO3 is generally preferred. The problem is rather
crucial for vidicon applications, where the material ought to be not only physically and
chemically stable in the temperature and pressure conditions o f the tube but also have
low thermal diffusivity, to obtain g o o d spatial resolution. That is why PVF2 is gaining
m u c h attention for vidicon applications on account o f its low thermal diffusivity as
c o m p a r e d to that Of TGS and LiTaO3, and also the ease with which thin target materials
can be prepared.
Similarly for application in laser optics, the choice o f the material is governed by its
vulnarability to radiation damage. Invariably crystals like TGS are not employed.
Materials with high curie temperature are f o u n d to be resistant to radiation damage.
The p h e n o m e n o n o f surface damage by laser pulses is not very well understood. Thus
the LiTaO3 detector is f o u n d to be satisfactory in the 1 m W - 3 W average power levels
from 10"6/~m C.W CO2 laser, but is damaged by a single 1/3 J pulse with a duration o f
200 msec. But a PZTO 1306 detector is not damaged even after fifty pulses. It has been
observed that polycrystalline ceramic pyroelectric materials are more resistant to
radiation damage than single crystals. Strontium barium niobates are another class o f
pyroelectric materials which are extensively employed in laser detection.
The temperature at which the detector has to be operated also determines the choice
o f the pyroelectric material. For better sensitivity o f the detector, the curie temperature
 Pyroelectric materials 325

must be as close to that of the operating temperature (in such a case, it also becomes
necessary to take special care to avoid depoling in the material). In this sense, it is useful
to have pyroelectric materials whose curie temperature could be changed by changing
the composition, doping etc. This accounts for the interest in materials like SBN,~:TN,
PbsGesOlt, XGS-rGSE and PLZT ceramics. Of late ceramic-plastic composites and
improper ferroelectrics are found to have good potential of becoming efficient
pyroelectric materials. (Harrison and Liu 1980; Newnham et al 1980; Shaulov et al
1980). The Pz'r-plastic composite has a figure of merit six times more than that of Pz'r.
The absence of dielectric divergence in improper ferroelectrics makes it possible to have
a high figure of merit. The rare earth molybdates, propionates and boracites are found
to be superior to TGS.

Acknowledgement

The Author is thankful to Prof R Vijayaraghavan and Dr M S Multani for their kind
interest.

References

Bccrman H P 1975 Infrared Phys. 15 225

Born M 1945 Re~. Mod. Phys. 17 245
Born M and Huang K 1954 Dynamical theory of crystal lattices (Oxford: Clarendon Press)
Garn L E and Sharp E J 1974 IEEE Trans. Parts Hybrids Packa#e P H P I 0 208
Harrison W B and Liu S T 1980 Ferroelectrics 27 125
Lang S B 1971 Source book on pyroelectricity (New York: Gordon and Breach)
Lccomte J 1948 Proc. Indian Acad. Sci. A28 339
Liu S T 1974 J. Electr. Mater. 3 91
Liu S T 1978 Proc.-IEEE 66 i4
Newnham R E, Skinner D P, Klicker K A, Bhalia A S, Hardimar B and Gururaja T R 1980 Ferroelectrics
27 49
Nye J F 1962 Physical properties of crystals (Oxford: Clarendon Press)
Shaulov A, Smith W A, Loiacano G M, Bell M T and Tauo Y H 1980 Ferroelectrics 27 117
Szigeti B 1975 Phys. Rev. Lett. 35 1532
Szigeti B 1976 Phys. Rev. Lett. 37 792
Watton R, Smith C and Jones G R 1976 Ferroelectrics 14 719