NUCLEAR INSTRUMENTS AND METHODS IO3 (I972) 85-91; ©
GENERALIZED SEMIEMPIRICAL EQUATIONS
FOR THE EXTRAPOLATED RANGE OF ELECTRONS
T. TABATA, R. ITO and S. OKABE
Radiation Center of Osaka Prefecture, Sakat, Osaka, Japan
Received 28 March 1972
The extrapolated or practical range R~ of monoenergetic
electrons in the energy regmn 0 3 keV-30 MeV for the absorbers
of atomic number 6-92 has been found to he expressed by a
single sem~emplrxcalequation of the form
Rex = al [(1/a2)ln(1+ a2r) -- as~/( 1 + aara5) ],
where r ~s the mctdent kinetic energy in units of the rest energy
of the electron, and the parameters a~ (i= 1, 2. . . . . 5) are given
by simple functions of atomic number Z. Values of nine con-
stants to express a~ have been determined by least-squares fit to a
1. Introduction
The extrapolated or practical range Rex o f electrons ~s
useful for various purposes in m e a s u r e m e n t a n d utdiza-
tion of electrons, for example, design o f detection
systems, simple energy d e t e r m i n a t i o n , a n d estzmation
of effects associated with deep p e n e t r a t i o n of electrons.
It is usually defined as the thickness of material at
which the extension of the hnearly decreasing region
of the transmission curve (number-thickness curve)
becomes zero. A t low energies, Rex is frequently
determined from h n e a r extrapolation of the n u m b e r -
energy curve measured for a given thickness of the
absorber, a n d the two definitions have been f o u n d to
yield consistent results. F o r absorbers of high atomic
n u m b e r , the t r a n s m i s s i o n curve often does n o t show
the linear region, a n d the extrapolation is then made
o f the t a n g e n t at the steepest p o i n t o f the curve*.
A n u m b e r o f authors have given empirical equations
to express the relation between Rex a n d the incident
kinetic energy T of the electron. Earlier works on this
p r o b l e m were m a i n l y restricted to the case of A1
absorbers, a n d have been reviewed by K a t z a n d
Penfold2). Later works 1'3-15) are briefly reviewed
in the appendix of the present paper. A l t h o u g h for-
m u l a t i o n of the dependence of Re~ on atomic n u m b e r Z
* Because of ambiguities accompanying the extrapolatmn in
such a case, Dupouy et al. ~) have proposed to use the cha-
racteristic thmkness Re instead of Re~; it ~s defined as the
thickness at which the transmission coefficient has fallen to 5 %.
While Rc is numerJcally almost the same as Re,, in the case of
the AI absorber, the difference is appreciable for absorbers of
higher atomic number. The present authors consider that each
of Rex and Re Is useful m different context.
85
NORTH-HOLLAND PUBLISHING CO.
total of 232 experimental points reported in 18 references; the
data used have been confined to those obtained from number-
transm~ssmn measurement The rms deviation of the equation
from the experimental data as 4 5700for energies above 1 MeV
and 8.4% for the entire energy regmn. Although the latter value
is rather large, it is mainly due to large fluctuation of the ex-
perimental data for energies below 1 MeV. An approximate
inverse relation which expresses z as a function of R~ IS also
gwen.
a n d atomic weight A o f the a b s o r b e r has been attempt-
e d t l - ~ * ) , the results are applicable only in limited
energy regions.
The purpose of the present paper is to describe a
semiempirical expression for Rex(T; Z, A) a n d an
expression for T(Re,,; Z, A), the latter being a n approx-
Imate inverse functxon of the former. Each of these
expressions, given in a single form, covers w-der
regions of T a n d Z t h a n any of previous equations.
2. Formulation
In order to derive the functional form of a generaliz-
ed e q u a t i o n for Rex, we first consider an expression
for the c o n t i n u o u s slowing d o w n a p p r o x i m a t i o n range
R o. It is defined by
Ro = St~t1 d T , (1)
0
where Stot is given by
Sto t = Seo I + Sra d ; (2)
Sco~ is the mass stopping power due to collision, a n d
Sraa is the one due to ra&atlon. W e b e r 9) has shown
that Sco~ can be well a p p r o x i m a t e d by
S~o,={C~[1-Cz(I+C3T)-2]} -', (3)
where C, ( i = 1, 2, 3) is a c o n s t a n t for a given absorber.
O n the other h a n d , the ratio S,.,d/S~ol is given approx-
imately by 16)
Sraa/S~ot "" TZ/1600mc2 _ ~tT, (4)
86
w h e r e m c 2 is the rest energy of the electron. Substituting
eqs. (2)-(4) into eq. (1), we obtain the relation
Ro~-Cl[(1/ct)ln(1 + a T ) - C2T/(1 + C3Z)],
where the terms of order u/Ca (.~ 1 0 - 3 Z ) have been
neglected compared with those of order unity*. The
relation between Rex and T is expected to be fitted
with a function similar to the right hand side of eq.
(5), though the behavior of Re~ and Ro differs shghtly
owing to the influences of multiple scattering detours
and energy-loss straggling exerted upon the former.
Actually a good fit to the existing experimental data
(the sources of the data used are described in section 4)
has been obtained by introducing a single additional
parameter. The fitting function can be written as
Rex = a l [ ( 1 / a 2 ) l n ( 1 + a 2 z ) - - a 3 z / ( 1 +a4zas)],
where a, (i= 1,2 ..... 5) is a constant for a given absorber
and z = T / m c 2. When a 2 z ~ 1 and a5 = l, the relation
(6) reduces to Weber's equation [eq. (A3) in the
appendix].
For the purpose of finding the dependences of a,
on Z and A, the values of a, for representative elements
have been determined by least-squares fit to the
experimental data. From an analysis of the results, the
following relations have been found:
al = b l A / Z b2,
a 2 = b3Z ,
a 3 = b4- bsZ ,
a4 = b 6 - b7Z,
a5 = ba/Z bg,
where the symbols b, (i= 1, 2 . . . . , 9) denote constants
independent of absorber material. In eq. (6a). the
constant b2 is close to unity, and the approximate
proportionality of a~ to A / Z is reasonable considering
the known behavior of the corresponding parameter
C1 in eq. (5). The form of eq. (6b) for a2 is also ex-
pected from the dependence of ~ on Z [see eq. (4)].
Since it is difficult to solve eq. (6) for ~, we formulate
an approximate inverse relation for convenience.
It is of the following form:
z= cx(exp{Rex[C2+C3/(l+c4R~-Sx)]/cl}--l),
* An expression approximately equal to the right-hand side
of eq. (5) can be obtained also by using a more rigorous
expression for Seo~, 1 e. the expression given by Rohrhch
and Carlson 17) in whxch only the logarithmic term subtracted
by the density effect correction is replaced by a constant.
T. TABATA et al.
where c, (i= 1, 2 ..... 5) is a constant for a given absorb-
er. The functional form of eq. (7) can be obtained by
replacing the variables z in the second term between
(5)
brackets of eq. (6) with Rex multiplied by a constant
and by solving for the remaining r. The dependences
of c, on Z have been determined by the same method
as used in finding eqs. (6a)-(6e), and are given by
cl = da/Z, (Va)
c 2 = d2Za3/A, (7b)
¢3 = d 4 - dsZ, (7c)
c4 = d6/Z ~7, (Vd)
c5 = d s / Z ag, (7e)
(6) where the symbols d, (i--1, 2, ..., 9) denote constants
independent of absorber material. The forms of eqs.
(Va) and (Vb) are expected from the fact that the
parameters ci and c2 are approximately the reciprocals
of a 2 and a j, respectively.
Hereafter, the expression (6) in which eqs. (6a)-(6e)
are substituted are simply called eq. (6), and similar
implication is carried by eq. (7).
3. Determination of the constants
Values of the constants b, and d, have been sought
(6a) by the least-squares fit with eqs. (6) and (7) to the
experimental data, respectively. Since the result of a
(6b) preliminary fit has shown that the data for Z = 4 cannot
(6c) be well fitted with eq. (6), the final computation has been
made by using a total of 232 data points for Z > 6 .
(6d) Uncertainties of the experimental values are not
shown in all the sources of the data used. Therefore,
(6e)
each data point has been given equal significance in
the sense that a,/(Rex), is assumed to be a constant
independent of i, where (Rex), is the ith experimental
value of Rex, and a, is its uncertainty or the standard
deviation about the "true" value. Strictly speaking,
this assumption is not always correct, but is the most
simple and convenient for obtaining the best fit to the
measurements of the quantity whose value varies over
more than a few orders of magnitude.
4. Sources of the data
The sources of the experimental data used are
(7) presented in table 1. Only the data obtained from
number-transmission measurement have been em-
ployed, and those obtained from ionization-in-depth
measurement have been excluded, because it has been
p o i n t e d o u t 24'26) that the latter technique leads
to shghtly different results. Although the previous
 EQUATIONS FOR THE EXTRAPOLATED RANGE OF ELECTRONS 87

TABLE 1
Sources of the data used to determine the values of the constants m the empirical equahons
The energy region and the absorber atomic numbers of each experiment are also shown.
The data for Z = 4 have been excluded in the final computation of least-squares fit.

Author T (MeV) Z Ref.

Kanter, Sternglass n 0.000764).00498 6, 13, 28, 47, 79 18

Kanlcheva, Burtsev h 0.0014-O.0049 79 4
Cosslett, Thomas 0.0025-0 015 13, 29, 47, 79 5
Schonland 0.01024).0876 13 19
Schonland 0.0185-0.0285 13, 29, 47, 79 20
Lane, Zaffarano 0 0273 13 21
Eddy 0.14654). 576 13 22
Sehger ~ 0 1594).96 13, 47, 50, 79, 82 23
Agu et al. e 0.254).75 4, 13, 29, 47, 79 24
Mdler, Hendncks 0.25-1 13 25
Gnmshaw 0.36-1.86 13 26
Nakal et al. 1-1.8 4, 13, 47 27
Nakai et al b 1-2 13, 29, 47 28
Miller 1-3 13, 22, 23, 29, 79 29
NakaJ et al. 1 5, 2 6 30
Ebert et al. b 4-12 6, 13, 29, 47, 73, 92 13
Tabata et al 4.09-23.5 4, 13, 29, 47, 79 14
Harder, Poschet 4.17-30 6, 13, 29, 48, 82 3l

"~ Numerical data obtained through private commumcatlon from the authors have been used.
b Values have been read from the graphs m the paper.
e Values of Re~/Ro have been read from the graph except m the case of Z = 13, for which numerical data are given, and have been
multlphed by the values of Ro given in refs. 32, 33

a u t h o r s 6.s'~°) have also used the data for A1203, TABLE2

Al-coated F o r m v a r , a n d Al-coated collodion in Values of the constants b~ and d~ (Re~ m g/cm2).
d e t e r m i n i n g the empirical constants for AI, the data
t b~ d~
for such composite absorbers have n o t been used
in the present treatment. The data of Viatskm and
1 0 2335+0.0091 (2.98 +0.30) x 10~
M a k h o v 34) have n o t been adopted because of rather
2 1.209 +0015 6.14 +0.29
large systematic deviations from the other data 4' 5,18,19). 3 (1.78 +0.36)× 10-4 1.026 +0.020
Since the data of T a b a t a et al. 14) for the effectively 4 09891+00010 (2.57 +0.12) × 10'z
semiinfinite m e d i u m show no significant difference 5 (3 01 +0.35)× 10-4 0.34 +0.19
from the data for the usual fimte-slab geometry 13'31), 6 1.468 +0.090 (1.47 +0.19)× I0a
7 (1 180 +0.097)×10 -z 0.692 _+0.039
they have been included. 8 1 232 +0.067 0.905 _+0.031
9 0 109 +0017 0.1874_+0.0086
5. Results and discussion
Values of the constants b, a n d d, determined are
hsted in table 2. The relations given by eq. (6) with these
values of b, are c o m p a r e d with the experimental data (~rms = 1/n) Rex(Z,)/(ReO~-l] 2 , for eq. (6), (8)
1
in figs. 1 a n d 2. In fig. 1, experlmental data for Al203
(refs. 18, 35) are also plotted at the lowest energies,
, for eq. (7), (9)
where the data for pure AI are absent. It can be seen
from these figures that eq. (6) well reproduces the
experimental data presently available in the energy where n is the n u m b e r o f experimental data, z~ is the ith
region 0.3 keV-30 MeV. A measure of goodness of experimental value of z, a n d Rex(L) a n d z[(Rex),]
fit m a y be given by the relative rms deviation t~rm s of are the values of Rex a n d z calculated from eqs. (6)
the empwical e q u a t i o n from the experimental data. a n d (7), respectively. Values of 6 .... are shown in table
It is defined by 3. W h e n the entire energy region is considered, 6rms
88 T. TABATA et al.

I0" I I0 10 2
, r , I ''"I , , , I ~,,,I t , ilrq]

I0 ___- 10-2

_ 10 -3

%-.
E
u
AI i
I0" _ 10-4

n~
Cu

10 .2 _ i0-~
(;u

AI

1 0 - 3 ~ i 0 -o
I 0 -4 I 0 -3 I 0 -2 I0-'
"r (MeV)

Fig. 1. Comparison of the semlempirlcal relation (6) (hnes) with experimental data (solid circle: AI, open clrcule: Cu). Use the scales
as indicated. The three experimental points at the lowest energies are for A12Oa.

is rather large. However, this is mainly due to large
fluctuation of experimental points for T < I MeV
as can be seen from figs. 1 and 2, and 6r,,s becomes
considerably small when only the data for T>_ 1 MeV
are taken into account (see the parenthesized values
in table 3). Although there is also a trend that 6rms IS
smaller for eq. (7) than for eq. (6), this can be attributed
to the superficial effect of the fact that the logarithmic
gradient g=(~/Rex)dRex/dz of the z-Rex curve is
greater than unity in larger part of the energy region
treated here (the relative deviation in Re~ is equal
to the corresponding relative deviation in z multiplied
by g).
The energies calculated from eqs. (6) and (7) for a
given value of Rex agree with each other within a few %
in most cases. Although the maximum relative devia-
tion 6 ..... of eq. (7) from eq. (6) exceeds 10% in the
cases of the lowest and the highest values of Z as
shown in table 4, it is considered to be compatible with
the uncertainty in the interpolation or the extrapolation
of the presently available data.
The relations (6) and (7) can be applied also to the
 E Q U A T I O N S FOR THE E X T R A P O L A T E D R A N G E OF E L E C T R O N S 89

I0-' I I0 I0 2
I ' ' ' I ''"I ' ' I ''"1 ' ' ' LT'~ "~

I0 I 0 -2

IO s

E
0

I0-; Ag 10 .4
>¢
\
n~

10 -2 IO-S
Au

Ag

10-s ~ 10-6
I 0 -4 10 3 I 0 -2 I 0 -I
T (MeV)
Fig. 2. Comparison of the serniernpirlcal relation (6) (lines) with experimental data (sohd cwcle: Ag, open circle: Au and Pb). Use the
scales as indicated.

absorbers which are mixtures or compounds. In this and f, is the fraction by weight of the constituent
case, Z and A should be replaced by 11,12) element with atomic number Z, and atomic weight A,.

Ze. = Y~AZ,, (10) Appendix

In this appendix, previous works concerned with
equations for the extrapolated range Rex and the
ACff = Z¢ff(Z/A)-~d, (11) characteristic thickness R c are reviewed (the definition
of Rc is given in the footnote on the first page of this
respectively, where paper).
Equation for R~x. Earlier works have been well
summarized by Katz and Penfold2), so that we do not
(Z/A)¢fr = ~f,Z,/A,, (12)
l repeat them here. The equations proposed by them for
90 T. TABATA et al.

TABLE 3 Rex=CIT[1-C2/(1-t-C3T)], (A3)

Number n of the data points used for the fit and
relative rms deviation 0rms of the empirical equa- where C, (i= 1,2, 3) is a constant. Kobetich and Katz ~°)
tions from the experimental data. The numbers
have made this equation applicable to the energy
m the bottom row pertain to the overall data
including absorbers other than those hsted, and interval 0.3 keV-20 MeV by small adjustment of the
the numbers m parentheses are for T_- > 1 MeV. constants.
For light materials and 3 < T < 15 MeV, Markus tl)
Absorber n t)rms (~o) has proposed an equation of the same form as eq.
eq. (6) eq. (7)
(A2) with Bl' and B2' proportional t o (Z/A)ecf - 1 ,
where (Z/A)eff is given by eq. (12) in section 5.
~3AI 82 (32) 6.5 (3.0) 4.3 ~2 5) Expressing B1 by a power function of atomic
zgCu 38 (27) 5.0 (4 3) 3.7 (3.6)
47Ag 32 06) 9 4 (4.0) 6.0 (2.8)
number Z, Rao lz) has made eq. (A1) apphcable to the
79Au 32 (10) 13.8 (5 9) 8.8 (4.3) regions 100 k e V < T < l MeV and 1 3 < Z < 8 2 . Ebert
Overall 232 (122) 8.4 (4.5) 5.8 (3.4) et al. 13) have expressed their experimental data for
4 < T < 12 MeV and 6 < - Z < 9 2 with an equation of the
same form as eq. (A2) in which 1/BI' and 1/B2' are
TABLE 4
linear functions of Z. Tabata et al. 14) have fitted their
Maximum relaUve deviation Omax m energy T of eq. (7) from experimental results for 4 < T < 2 4 MeV and 4 < Z < 79
eq. (6). with an equation containing six parameters.
Equation for R e. Dupouy et al. 1) have formulated
Absorber ,5.... (%) an empirtcal equation to express the transmissmn
1 keV < T< 1 MeV 1 MeV-< T <
- 30 MeV
curves observed for 50 keV<T<_l.2 MeV and
13<Z_<79, and have derived from it an expression
~C - 13.0 8.2
for Rc as a function of T, Z, the atomic mass, and the
laAl - 6.1 - 18
29Cu - 3.4 - 5.0 density of absorber material. Kobetlch and Katz t5)
4TAg - 4.3 3.9 have fitted the values of R e, given by the equation of
79Au - 12.7 6.1 Dupouy et al., with eq. (A3) expressing C~ as functions
92U - 20.3 - 11.6 of Z. The resultmg equation has nine empirical
constants.

the A1 absorber have been the most famihar, and are
expressed in the form
Re~ = BIT B2-B31nT, for 0.01 < T < 2 . 5 MeV,
Rex = B'~T-B'2, for T > 2 . 5 MeV,
where T is the incident kinetic energy of the electron,
and B, (1=1, 2, 3) and B / (i=1, 2) are constants.
Young 3) has fitted the power law to his experimental
data for A1 in the energy region 0.5-11 keV. Kanicheva
and Burtsev 4) and Cosslett and Thomas s) have found
that the power law is applicable also to metals other
than A1 for 2 < T<_ 15 keV.
Using the published experimental data for
0 . 2 < T < 3 . 5 MeV, Varshni and Karnatak 6) have
reevaluated the constants in the Flammersfeld equa-
tion 7) for AI. Kahre and Varshni 8) have proposed a
modified Flammersfeld equation, extending the region
of validity down to 2 keV.
Weber 9) has semiempirically estabhshed an equation
valid for AI in the energy region 3 keV-3 MeV. It is
of the form
References
1) G. Dupouy, F. Perrler, P. Verdmr and F. Arnal, Compt.
Rend. 260 (1965) 6055.
2) L. Katz and A. S. Pen(old, Rev. Mod. Phys. 24 (1952) 28.
(A1) 3) j. R. Young, J. Appl. Phys. 27 (1956) 1.
4) I. R. Kamcheva and V. V. Burtsev, Flz. Tverd. Tela 1 (1959)
(A2) 1250 [Enghsh transl.: Sovmt Phys.-Sohd State 1 (1960) 1146].
5) V. E. Cosslett and R. N. Thomas, Brit. J. Appl. Phys. 15
(1964) 1283.
6) y p. Varshm and R. C. Karnatak, Ann. Physlk 2 (1959)413.
7) A. Flammersfeld, Naturwlssenschaften 33 (1946) 280; Z.
Naturforsch. 2a (1947) 370.
8) S P. Kahre and Y. P. Varshm, Ann Physlk 7 (1961) 220.
9) K.-I-[. Weber, Nucl. Instr. and Meth. 25 (1964) 261.
10) E. J. Kobetmh and R. Katz, Phys. Rev. 170 (1968) 391.
11) B Markus, Strahlentheraple 116 (1961) 280.
12) B N S. Rao, Nucl. Instr. and Meth. 44 (1966) 155.
13) p. j. Ebert, A. F. Lauzon and E M. Lent, Phys Rev. 183
(1969) 422.
14) T. Tabata, R. Ito, S. Okabe and Y. Fujlta, J. Appl. Phys.
42 (1971) 3361
15) E J Kobetmh and R. Katz, Nucl. Instr. and Meth. 71
(1969) 226.
16) H A. Bethe and W. Heltler, Proc. Roy. Soc. (London)
A146 (1934) 83.
17) F. Rohrhch and B. C. Carlson, Phys. Rev. 93 (1954) 38.
18) H. Kanter and E. J. Sternglass, Phys. Rev. 126 (1962) 620.
 E Q U A T I O N S FOR THE E X T R A P O L A T E D R A N G E OF ELECTRONS 91

19 B. F. J. Schonland, Proc. Roy. Soc. (London) A108 (1925) Ann. Rep. Japan Assoc. Rad. Res. Polym. 3 (1961) 11.
187. 28) y. Nakai, K. Matsuda, T. Takagaki and K. KJmura, Ann.
20 B. F. J. Schonland, Proc. Roy. Soc. (London) A104 (1923) Rep. Japan Assoc. Rad. Res. Polym. 5 (1964) 7.
235. ~9) W. E. Mdler, NASA Report TN D-5724 (1970).
21
R. O. Lane and D. J. Zaffarano, Phys. Rev. 94 (1954) 960. 30) y. Nakal, K. Matsuda, T. Takagakt and K. Klmura, Ann.
22 C. E. Eddy, Proc. Cambridge Phd Soc. 25 (1929) 50. Rep. Japan Assoc. Rad. Res. Polym. 7 (1966) 5.
23 H. H Sehger, Phys. Rev. 100 (1955) 1029. 31) D. Harder and G. Poschet, Phys. Letters 24B (1967) 519.
24 B. N. C. Agu, T. A. Burdett and E. Matsukawa, Proc. Phys. 32) L. V. Spencer, Phys. Rev. 98 (1955) 1597.
Soc (London) 71 (1958) 201, 72 (1958) 727. za) A. T. Nelms, Nat. Bur. Std. Circular no. 577 (1956).
25 W. E Mdler and H. D. Hendncks, NASA Report TN D-4364 a4) A. Ia. Vlatskm and A. F. Makhov, Zh. Tekhn. Flz. 28
(1968). (1958) 740 [English transl.: Soviet Phys.-Tech. Phys. 3
26 J. A. Gnmshaw, Brit. J. Appl. Phys 17 (1966) 639. (1960) 690].
27
Y. Nakal, H. Inomo, K. Matsuda, T. Osuga and K. Klmura, 3~) j. R. Young, Phys. Rev. 103 (1956) 292.