JOURNAL OF MATERIALS SCIENCE $ (1973) 1 7 3 7 - 1 7 4 3

The P - T - x phase diagram of PbTe and

PbSe
B. J. S E A L Y * , A. d. CROCKER-~
Zenith Radio Research Corporation, (UK) Ltd, Stanmore, Middlesex, IlK

Pressure-temperature-composition (P-T-x) phase diagrams have been calculated for PbTe

and PbSe in the temperature range 250~ to the melting point. Various thermodynamic
quantities have been estimated which are consistent with electrical parameters measured
at room temperature. However, refinements in the band structure and its temperature
dependence are needed before an accurate calculation of intrinsic carrier concentration
at high temperatures becomes possible.

1. Introduction been performed and agree with previous deter-

The determination of phase diagrams, such as minations of the phase widths of PbTe and PbSe
the temperature-composition (or T-x) and (Figs. 1 and 2) [20]. However, difficulty was
pressure-temperature (or P - T ) projections, and experienced with n-type PbTe which is clear
the estimation of reaction constants and enthal- from the scatter of results in Fig. 1. The straight
pies for the lead chalcogenides have been the line portions of the phase extremity in this
subject of numerous publications [1-19]. Because figure may be represented in terms of the net
of inherent difficulties in vapour-phase equilibra- carrier concentration [ n - p ] by the equation
tion and the long annealing (precipitation) [21]
times required at low temperatures, most experi- d[ln(n - p)]/d(1/T) = - A H p b / k f (1)
mental data exist only for high temperatures.
Thus prediction of the equilibrium composition where AHa,b is the heat involved in reversibly
of evaporated thin films deposited onto heated transferring an atom of lead from solid PbTe
substrates (for 200 ~< T ~< 400~ is not
TABLE I Slope of phase width for T > 400~
immediately possible. The present work was
undertaken in order to predict and control the Slope PbTe PbSe
growth of epitaxial PbTe (PbSe) films and to Hpz(eV) 0.66 0.50
understand the mechanisms involved in post- Hx2(eV)* 0.52 0.30
growth annealing experiments. *X2 = Te2, Se2.

2. M ethod to the coexistent liquid (Table I). The parameter

In order to calculate P - T - x diagrams, various f is given by
experimental data are required as a function of
temperature, (i) the phase width or existence
range, (ii) the intrinsic carrier concentration, f= [(I n -7 l]-:'
(iii) the partial pressure of the constituent In - p I/2K ''''>
elements, (iv) the enthalpy of formation and + {[(I n -- IS'-"}
sublimation energy of the compound. The charge
state of the predominant defects is also required. where ni is the intrinsic carrier concentration and
Most quantities are available in the literature, Ks' the ionized Schottky constant. In the region
although some values are rather uncertain. where the compound is in equilibrium with
essentially pure Pb (or pure Xz where X 2 = T%,
2.1. Phase width or T-x projection S%, Sz), the slope may be either z~Hi, b/k or
Annealing and precipitation experiments have AHpb/2k depending upon whether I n - p I is
*Present address: Department of Physics, University of Surrey, Guildford, Surrey.
tPresent address: Department of Physics, Thames Polytechnic, London.
9 1973 Chapman and HallLtd. 1737
 B. J. SEALY, A. J. CROCKER

-- ' ~ ~ 0.8 I ---- I TABLE II Band parameters used to calculate the np

product
~a~ ~ , ~ 1"O-
Parameter PbS PbSe PbTe
9 ~. 1-2-
Ego (OK), (eV) 0.28 0.14 0.18
N 1"4-
.Eg (max), (eV) 0.45 a 0.34 9 0.35b
dEg/dT% (eV K -1) 4 x 10-4 4 x 10-4 4 x 10-4
,, [5[ A~--~ \x
o 6
# 111
~ [7]
~[12]
-[361
9 THIS INVESTIGATION
1"6-

1.8-

2"0-

2"2-
/ Ev (300K), (eV)

rnaa/rno (hole) (300K)

rna2*/mo (hole) (300K)
me*/mo (electron) (300K)
0.38t

0,40a
--
0.40a
0.38%
0.375
0.35e
0.95
0.30e
?Estimated to give reasonable values of ni at high
0.100

0.35r
2.10
0.25f

I I I I temperatures.
1019 IO18 1017 IO18 1019 1020
ELECTRON CONCENTRATION, HOLE CONCENTRATION, 5[23].
n77 (cm -3) P77(cm-3) (a) [37].
(b) Mean of three results, 0.39 eV [37]; 0.34 eV [38];
Figure 1 Phase width of PbTe. and 0.33 eV, [39].
(c) [401.
(d) [1]
(e) [391
(f) [41]
09
v" ~"
1.1
function of temperature for PbSe (Fig. 3) used
4, * to calculate the internal structure o f the P - T - x
93 - - ~ phase diagram, agree with experiment, and agree
o [13] A '5--~
with theory only below 300~ F o r PbTe there
9. [161 are no available experimental values at high
v [14] ~ 1"7--~ temperatures for a similar comparison to be
[ 9 ] ~
made.
9 THIS INVESTIGATION \ 1"9-

#1- -~ 10 2o
10 20 1019
I
iO 18
I I
IO 16 1019
[
10 20
to& 8&& 650 5So 4&o 3oo
ELECTRON CONCENTRATION~ HOLE CONCENTRATION~
n 77 (cm -3} P77(cm-3) u

lO19 ~2
Figure 2 Phase width of PbSe. z
O

intermediate in value between 2ni and 2Ks '1 ;2 or

5
large compared with both. This criterion explains 1018 U
Z
the kinks in the locus o f the phase extremity for O
U
p-type Pbq-e and n-type PbSe, (Figs. 1 and 2). (2:
<
2.2. Intrinsic carrier concentration IO17

The intrinsic carrier concentrations o f PbTe and Z

PbSe have been calculated using a two valence _z
band model for the former and a single valance
q I I iO 16
b a n d for the latter, all bands being assumed 1.O 2-0 3.0
parabolic. The parameters used in the calcula- RECIPROCAL TEMPERATURE
1 0 ~ / r (K~
tions are shown in Table II. Experimental
evidence for a second valence b a n d in PbSe has Figure 3 ni versus 1/T for PbTe and PbSe.
been obtained [22, 23]. However, assuming a
single valence b a n d results in values of ni 50
greater than experimental ones for PbSe [20]. 2.3. The three phase line
The presence of two valence bands produces an The envelope o f the P - T diagram is the line
even greater discrepancy. The values of n~ as a along which three phases coexist in equilibrium
1738
THE P-T-x PHASE DIAGRAM OF PbTe AND PbSe

and is called the "three phase line". The liquid- 4

vapour-solid and vapour-solid-solid lines are of 2 OHm,-/ ~~ ~ ~ ~ Zt
.o.
particular interest here. Part of the three phase vZL
OOH
line has been determined experimentally for 2 o
PbTe [17, 18] at high temperatures but at lower =-2 Q r _
temperatures where there are no available data, ~o_4
the three phase line has been represented by the
pressure of tellurium (PT%) over pure tellurium 5-6
,pc
and the pressure of lead (P~b) over pure lead [24]. -~ n =p
This is a reasonable assumption at low tem- -10
peratures where PbTe is in equilibrium with
I I I
essentially pure lead on the lead rich extreme -12 o.s o.9 1.0 1.1 1.2 1.3 1.4 1.s 1.6 1.7 1.8 1.9
R E C I P R O C A L T E M P E R A T U R E , 103]7-(K)
and pure tellurium on the tellurium rich extreme
of the existence range. In order to obtain the Figure 5 Calculated P - T - x phase diagram for PbSe.
three phase line at the lead rich extreme of the Experimental points, Z1 [9]; O H , [13].
existence range in terms of PTe=, Equation A4
was used in conjunction with values of KPbTe
after Bis [17] (see Fig. 4). PbTe (i) T > 7 0 0 K , p = [Vpb'] p-type
(ii) T ~< 700K, p = n~ p-type
(iii) All T, n = ni n-type
PbSe (i) All T, p = [Vpb'] p-type
(ii) All T, n = [Pb~'] n-type
Knowing Kox, Kg and ni as a function of
temperature (Figs. 6 and 7) it was possible to
calculate n- and p-type isocarrier concentration
-p lines (Figs. 4 and 5) whose intersection gives
values of the ionized Schottky or Frenkel
2-a constant (Figs. 6 and 7).
x . \ -.<.-p
-1o
-12 ' ~ ? . ~~ . 1 3, Discussion
m 3.1. Schottky verus Frenkel disorder
I , , I , I , ~ , 1'~-,, |
O-8 0'9 90 1'1 1'2 1'3 1'4 1"5 1"6 1'7 1"8 1"9 Bloem and Kr6ger [2] assumed Schottky dis-
R E C I P R O C A L T E M P E R A T U R E ~10 3/7-(K)
order to predominate in PbS at high temperatures
Figure 4 Calculated P - T - x phase diagram for PbTe. but lead interstitials have been deduced as being
present in all three lead chalcogenides from
tracer diffusion experiments as a function of
The three phase line for PbSe was obtained non-stoichiometry [25-29]. Density and lattice
from values of Kpbse calculated by Ohashi and parameter measurements also indicate that lead
Igaki [13] and using published data for the interstitials and not selenium vacancies are the
vapour pressure of pure lead and selenium as a most abundant defect in lead-rich PbSe [30]. In
function of temperature [24] (see Fig. 5). fact, the diffusion experiments indicate that the
defect structure consists of Frenkel defects on the
2.4. Internal structure of the P-T-x phase cation sublattice, i.e. lead interstitials, for lead-
diagram rich [25-29] and lead vacancies for chalcogen-
In order to calculate the internal structure of the rich lead chalcogenides [25, 31, 32].
P - T diagram various points at the phase width The analysis of Frenkel disorder in a binary
(i.e. maximum ionized vacancy or ionized compound is identical to the Schottky defect
interstitial concentration) were plotted on the case. Thus, we assume that the n-type con-
three phase line, each point corresponding to a ductivity of lead rich material is owing to singly
given defect concentration, temperature and ionized interstitials.
partial pressure (PTe2). From these values and
Equations A1, A2 and A4, values of Kox and 3.2. PbTe
KR were calculated using the following electro- There are very few published experimental
neutrality conditions: results to compare with the present calculations
1739
 B. J. S E A L Y , A. J. C R O C K E R

,~176
r 9-

1,~ ~ 1039 I042

E .
~10L/ \
15 "1
,,o'
'~ lO39~,~ ~ T

1038 10 41 'ue

0"9 t.1 1'3 , 1'5 1.7 1.9 2"1

RECIPROCAL TEMPERATURE~
103IT(K)
Figure 6 Reaction constants of PbTe as a function of
temperature.
of the internal structure of the P-T-x phase
diagram. Values of Ki from Fujimoto and Sato
[11 ], (FS), are appreciably lower than those
calculated here (Fig. 3). However, by re-drawing
their diagrams of vacancy concentration versus
KoxPTe~I/2 to fit their experimental points, it was
possible to show that FS results are consistent
with the authors'. Table III compares reaction
constants obtained from the adjusted experi-
mental results of FS [11] with those calculated
here.
Most reaction constants obtained from FS are
less than the authors', but the high temperature
results (750~ agree reasonably well. The
differences in the Ks values, if not experimental,
can be accommodated by adjusting parameters
in the calculations, e.g. effective masses, energy
gap, temperature dependences, non-paraboIicity
of energy bands, variations in scattering mecha-
nisms, values of which are uncertain at elevated
V I ! I \ I 1
I I
0"9 1.0 1.1 1-2 1'3 1.4 t.5 1.6 1.7
RECIPROCAL TEMPERATURE~
1031T(K)
Figure 7 Reaction constants of PbSe as a function of
temperature.
temperatures. At 600~ it is difficult to reconcile
the differences. However, the FS Kox value is less
than the authors' because the formers' carrier
concentration is not the equilibrium value at the
phase boundary (see Fig. 1), i.e. [Vl,b'] is too
low, presumably because of an insufficiently long
annealing time.
3.3. PbSe
Although there are more experimental data
available for PbSe than for PbTe, the phase
width measurements have not, until now, been
related to the internal structure (experimental)
of the P-T-x diagram. Neither have the results
been extrapolated to low temperatures where
they should give useful guidance to the prepare-
tion of PbSe films. Thus, the PbSe results
presented here represent a more detailed analysis
over a wider temperature range than do previous

T A B L E I I I Comparison of experimental (FS) [11 ] and calculated reaction constants for PbTe

Temperature (~ Reactionconstants
Ks ( c m - 0 K1 (cm -3) Kox (cm -6 T o r r - 0

FS T-x FS T-x FS T-x

750 8.5 X 101~ 8 • 1017 6 • 1018 1019 1.3 • 1038 1.8 X 1038
600 2.0 >( l 0 t~ 1017 2 X 10 Is 6 X 10Is 6.4 X 10 a7 1.7 X 1038

1740
THE P-T-x PHASE DIAGRAM OF PbTe AND PbSe

investigations. PbTe and PbSe are compared with those for

The calculated internal structure of the P - T PbS (Table IV), all values being estimated at
diagram compares well with experiment (Fig. 5) high temperatures (,-~ 800 to 1100K). For PbTe
although a number of inconsistencies were there is agreement with the results of Bis [17].
apparent at high temperatures on the p-side. The PbSe results (Table IV) are numerically
These were reconciled by adjustment of Kox less than those calculated by Zlomanov [9], but
values so that experimental and calculated his data and those of Ohashi and Igaki [13] are
internal structure coincided. But this adjustment consistent with the calculations reported here
suggested that the three phase line at high (see Fig. 5).
temperatures [13] was incorrect, so a new line
was estimated to be consistent with experiment. TABLE IV Comparison of some reaction enthalpies
Although this recalculated line disagrees with for the lead chalcogenides (T ~ 800 to
the results of Ohashi and Igaki by about an ll00K)
order of magnitude at some points, it is preferred
because a departure from ideal behaviour is most Enthalpy (eV) PbTe PbSe PbS*
likely at these temperatures, i.e. the vapour Hs (or HF) 2.9? 3.0? 2.5
pressure of selenium over p-type PbSe is probably Ei 0.58 0.78 1.0
less than that over pure selenium, suggesting a Hs' (or HF' ) 2.3 2.2 1.78
H~ - 1.0 - 0.94 - 1.34
decrease in activity coefficient to a value less than
Hox 0.53 0.30 0.64
unity; cf. PbS [2], ZnTe [33], CdTe [33],
HgTe [34], PbTe [18] where the activity of T% *Results taken from [1], which uses Ea' = Eb' = 0.14 eV.
(S~) has decreased considerably at high tem- +.Calculated assuming Ea' = Eb' ~ 0, i.e. Hs = Hs' +
Ei.
peratures.

3.4. General discussion Because of the variations reported in some

There are a number of points which have not
been mentioned so far which warrant some
discussion. The first point is that there is a small
error present in the position of the lead rich phase
boundary of the P - T diagram because no
allowance was made for the temperature
dependence of the enthalpy of formation of
PbTe and PbSe.
The concentration of ionized defects at
temperature T K may be written as [1 ]
Ka = 4( 2rrmok T/h2) ~12(mal */mo) ~/2
e x p ( - Ea/kT) (3)
If we assume the effective mass mdl* oc I"1/a,then
Equation 3 becomes
Ka oc T~/~ e x p ( - Ea/kT) oc e x p ( - E a ' / k T ) ( 4 )
where Ea and Ea' are the actual and effective
ionization energies of the defects. For PbS in
the range 1000 to 1200K, Kr6ger [1] assumed
Ea ,-~ 0.01 eV giving Ea' "~ 0.14 eV. For the lead
chalcogenides there is doubt as to the value of
Ea because recent band structure calculations
suggest that there is no ionization energy of
defects in PbTe [35]. Furthermore, there has
been no experimental observation of the freeze-
out of carriers at low temperatures (,-~ 4.2K).
Thus no attempt has been made to calculate
values of KB~v and KAy from the respective
Kox and K~ values (see Appendix).
Finally the enthalpies of several reactions for
experimentally determined parameters and the
assumptions required for the calculations, the
results reported in this paper may need some
adjustment as and when new and more reliable
data appear. However, it is thought the results
will be of help for example, in predicting the
electrical properties and the annealing charac-
teristics of epitaxial thin films.
Appendix
Equilibrium between the solid and the gas
phase
If a crystal compound AB is in equilibrium with
a gas phase the interaction between the solid
and gas may be represented by the following
equations assuming Schottky disorder pre-
dominates [1, 2]
A(g) ~-AA:' + VBx + HAy,
[VBx] = PAKAv (A1)
89 ~ BBx + VAx + HB~v,
[VAx] = P]321/ZKB~v (A2)
where g denotes the gaseous form of A and B,
and P the vapour pressure. The notation follows
that of KrSger [1 ]. The enthalpy, H, of a reac-
tion is related to the reaction constant, K, by the
equation,
K = K ~exp(- H/kT) (A3)
1741
 B. J. S E A L Y , A. J. C R O C K E R

where the pre-exponential, K ~ can be t e m p e r a - electroneutrality c o n d i t i o n for n-type PbSe is

ture dependent. T h e r e l a t i o n between the h e a t n = [ P b i ' ] then E q u a t i o n A 7 will b e c o m e :
o f f o r m a t i o n , HAB a n d the p a r t i a l pressures o f the
A(g) ~ AAx + Ai" + e' + H m ,
c o m p o n e n t species is
[ A i ' ] n = P•Km (A7a)
P A P ~ 1'~ = KAI3 = Ks/KAvKB2v (A4)
the values o f KR a n d K m being identical.
Ks being the S c h o t t k y constant. T h e i n c o r p o r a -
tion o f a t o m s f r o m the gas phase can p r o d u c e Acknowledgements
c h a r g e d defects,
T h e a u t h o r s wish to a c k n o w l e d g e the help a n d
VB" ~- VB" + e' + Eb, [ V B ' J n = Kb (A5) advice o f D r M i n o G r e e n a n d D r L. M. Rogers.
and
VA x ~ Va' + h" + Ea, [VA']p = Ka (A6) References
1. F. A. KR(SGER, "The Chemistry of Imperfect
Ea a n d Eb are the i o n i z a t i o n energies a n d n a n d p
Crystals" (North Holland, Amsterdam, 1964).
are the concentrations o f electrons a n d holes 2. J. BLOEM and F. A. KR6GER, Z. Phys. Chem. 7
respectively. The c o m b i n a t i o n o f E q u a t i o n s A1, (1956) 1.
A2, A 5 a n d A 6 gives 3. J. R. S T U B B L E S a n d c . E. B I R C H E N A L L , Trans. Met.
A(g) ~ AAx + VB" + e' + HR, [VB'] n Soe. AIME 215 (1959) 535.
= PAKR (A7) 4. A. J. STRAUSS, ibid239 (1967) 794.
5. R. r. BREBRICK and R. S. ALL~AIER, J. Chem. Phys.
89 ) ~ - B B x + VA' + h" + H o x , [VA']p 32 (1960) 1826.
= Pz21'~Kox (A8) 6. R. F. BREBRICl( and E. GUBNER, ibid 36 (1962)
where writing KR = KA + Kb a n d K o x = KBV 1283.
+ Ka is often m o r e convenient. U s i n g the mass 7. w. w. SeANLON, Phys. Rev. 126 (1962) 509.
8. H. ABRAMS, Ph.D. Dissertation, Lehigh University,
a c t i o n e q u a t i o n for the np p r o d u c t ( E q u a t i o n
USA (1968).
A9) we m a y o b t a i n further relationships, e.g. 9. V. P. ZLOMANOV, O. V. MATVEEV and A. V.
O ~ - e ' + h" + E~, n p = K~ (A9) NOVOSELOVA, Vestnik Moskov Univ. 5 (1967) 81;
Ks' = KsKaK~)/K, (A10) 6 (1968) 67.
10. K. IGAKI and N. OHASHI, Bull. Univ. Osaka
and
Prefecture 11 (1962) 89.
Hs' =Hs+ Ea + E b - - E~ (All) 11. M. FUJIMOTO and Y. SATO, Jap. J. Appl. Phys. 5
the primes (Equations A10 a n d A l l ) referring (1966) 128.
to the ionized S c h o t t k y c o n s t a n t (enthalpy). 12. T. L. K O V A L ' C H 1 C a n d I U . P. M A S L A K O V E T S , Soy.
Electroneutrality ( E q u a t i o n A12) a n d the Phys. Tech. Phys. 1 (1956) 2337.
equality o f the n u m b e r o f A a n d B sublattice 13. N. OHASHI and K. IGAKI, Trans. Jap. Inst. Met. 5
sites i m p o s e d by the crystal structure ( E q u a t i o n (1964) 94.
14. R. F. BREBRICK and E. GUBNER, J. Chem. Phys. 36
A13) represent o t h e r i m p o r t a n t c o n d i t i o n s :
(1962) 170.
n + [Vx'] = p + [VB'] (A12) 15. A. V. NOVOSELOVA, V. 1". ZLOMANOV and o. v.
[AAx] + [V.~:'] + [VA'] = [BBX] + MATVEEV, Inorg. Mat. 3 (1967) 1154.
16. A. R. C A L A W A , T. C. H A R M A N n M. F I N N and 1".
[VBx] + [VB']. (A13)
YOUTZ, Trans. Met. Soe. AIME 242 (1968) 374.
A t high t e m p e r a t u r e s the d o m i n a n t neutrality 17. R. F. BIS, Naval Ordnance Lab. Tech Report, 61-
conditions n = [ V s ' ] a n d p = [VI,(] a p p l y to 128 (1961); J. Phys. Chem. Solids 24 (1963) 579.
PbS for lead- a n d sulphur-rich material, 18. R. F. BREBttlCK and A. a. STRAUSS, Y. Chem. Phys.
respectively [2]. O n substitution into the a b o v e 40 (1964) 3230.
equations we o b t a i n : 19. E. C. ABBOTT, Ph.D. Thesis, University of Delaware
K, Kb (1965).
n~ - ( A 14) 20. ~. L SEALY, Ph.D. Thesis, University of Surrey
Ps~ ~/~ Ks~v (1972).
p~ = K o x Ps~ 1'~ (A15) 21. R. E. BREBRICK, J. AppL Phys. Suppl. 33 (1962) 422.
22. M. N. VINOGRADOVA, I. M. RUDNIK, L. M.
n ~" = Ke~ P~b. (A16) SYSOEVA and N. V. K O L O M O E T S , Sov. Phys.
The i n t r o d u c t i o n o f F r e n k e l d i s o r d e r o f the Semieond. 2 (1969) 892.
c a t i o n sublattice does n o t affect the calculations 23. M. N. VlNOGRADOVA, ibid3 (1969) 231.
b u t different n o m e n c l a t u r e is required to dis- 24. R. n. HONIG and D. A. KRAMBR, R.C.A. )~ev. 30
tinguish F r e n k e l f r o m S c h o t t k y disorder. I f the (1969) 285.
1742
THE P-T-x PHASE DIAGRAM OF PbTe AND PbSe
25. M.P. GOMEZ, D. A. STEVENSONand R. A. HUGGINS,
J. Phys. Chem. Solids 32 (1971) 335.
26. T. D. GEORGE and J. B. WAGNER, J. Appl. Phys. 42
(1971) 220.
27. M. S. SELTZER and J. B. WAGNER, Jr. Chem. Phys.
36 (1962) 130; J. Phys. Chem. Solids 24 (1963) 1525.
28. G. SIMKOVICH and J. B. WAGNER, ,]. Chem. Phys.
38 (1963) 1368.
29. K. R. ZAN~O and J. B. WAGNER, ,]. AppL Phys. 39
(1968) 5686.
30. H. G O B R E C H T a n d A. R I C H T E R , J. Phys. Chem.
Solids 26 (1965) 1889.
31. M.S. SELTZER and s. B. WAGNER, ibid 26 (1965) 233.
32. Y. BAN and J. B. WAGNER, J. AppL Phys. 41 (1970)
2818.
33. A. S. JORDAN and g. R. ZUPP, J. Electrochem. Soc.
116 (1969) 1285.
34. R. F. B R E B R I C K and A. J. S T R A U S S , J. Phys. Chem.
Colloids 25 (1964) 1441.
35. N. J. PARADA and G. W. PRATT, Phys. Rev. Letters
22 (1969) 180.
36. x. D. GEORGE and J. B. WAGNER, J. Phys. Chem.
Solids 30 (1969) 2359.
37. A. r . GIBSON, Proc. Phys. Soc. 65 (1952) 378.
38. R. N. T A U B E R , A. A. M A C I t O N I S and I. B. CADOFF,
J. Appl. Phys. 37 (1966) 4855.
39. L. H. ROGERS and A. J. CROCKER, to be published.
40. L. M. ROGERS, Brit. Y. Appl. Phys. 18 (1967) 1227.
41. G. W. JOHNSON, J. J. Electron Control 12 (1962) 421.
Received 24 May and accepted 22 June 1973.
1743