Chapter 14
α Decay
Note to students and other readers: This Chapter is intended to supplement Chapter 8 of
Krane’s excellent book, ”Introductory Nuclear Physics”. Kindly read the relevant sections in
Krane’s book first. This reading is supplementary to that, and the subsection ordering will
mirror that of Krane’s, at least until further notice.
14.1 Why α Decay Occurs
See Krane’s 8.1.
14.2 Basic α Decay Processes
An α decay is a nuclear transformation in which a nucleus reduces its energy by emitting an
α-particle.
A A−4
ZXN −→Z−2 XN′ −2 +42 He2 ,
or, more compactly :
A
X −→ X ′ + α .
The resultant nucleus, X ′ is usually left in an excited state, followed, possibly, by another
α decay, or by any other form of radiation, eventually returning the system to the ground
state.
1
2 CHAPTER 14. α DECAY
The energetics of α decay
The α-decay process is “fueled” by the rest mass energy difference of the initial state and
final state. That is, using a relativistic formalism:
Ei = Ef
mX c2 = mX ′ c2 + TX ′ + mα c2 + Tα
Q = TX ′ + Tα where
Q ≡ mX c2 − mX ′ c2 − mα c2
Q/c2 ≈ m(AX) − m(A−4X ′ ) − m(4He) , (14.1)
where TX ′ and Tα are the kinetic energies of the nucleus and the α-particle following the
decay. Note that atomic masses may be substituted for the nuclear masses, as shown in the
last line above. The electron masses balance in the equations, and there is negligible error
in ignoring the small differences in electron binding energies.
The line of flight of the decay products are in equal and opposite directions, assuming that X
was at rest. Conservation of energy and momentum apply. Thus we may solve for Tα in terms
of Q (usually known). X ′ is usually not observed directly. Solving for Tα by eliminating X ′ :
Q = Tα + TX ′ (14.2)
|~pα | = |~pX ′ |
p2α = p2X ′
2mα Tα = 2mX ′ TX ′
(mα /mX ′ )Tα = TX ′ . (14.3)
Using (14.3) and (14.2) to eliminate TX ′ results in:
Q = Tα (1 + mα /mX ′ ) or (14.4)
Q
Tα = (14.5)
(1 + mα /mX ′ )
Q
Tα ≈ or (14.6)
(1 + 4/A′ )
Q
Tα ≈ or (14.7)
(1 + 4/A)
Tα ≈ Q(1 − 4/A) (14.8)
14.3. α DECAY SYSTEMATICS 3
Equations (14.4) and (14.5) are exact within a non-relativistic formalism. Equation (14.6)
is an approximation, but a good one that is suitable for all α-decay’s, including 8 Be→ 2α.
Equation (14.7) is suitable for all the other α-emitters, while (14.8) is only suitable for the
heavy emitters, since it assumes A ≫ 4.
We see from (14.8) that the typical recoil energy, for heavy emitters is:
TX ′ = Q − Tα ≈ (4/A)Q . (14.9)
For a typical α-emitter, this recoil energy (Q = 5 MeV, A = 200) is 100 keV. This is not
insignificant. α-emitters are usually found in crystalline form, and that recoil energy is more
than sufficient to break atomic bonds, and cause a microfracture along the track of the recoil
nucleus.
Relativistic effects?
One may do a fully relativistic calculation, from which it is found that:
1 Q
Q 1+ 2 mX ′ c2
Tα = , (14.10)
mα Q
1+ + m c2
mX ′ ′
X
mα 1 Q
m
Q 1 + 2 mα c2
TX ′ = X′ . (14.11)
1 + mmα + m Q c2
X′ X′
Even in the worst-case scenario (low-A) this relativistic correction is about 2.5 × 10−4 . Thus
the non-relativistic approximation is adequate for determining Tα or TX ′ .
14.3 α Decay Systematics
As Q increases, t1/2 decreases
This is, more or less, self-evident. More “fuel” implies faster decay.
The “smoothest” example of this “law” is seen in the α decay of the even-even nuclei.
Shell model variation is minimized in this case, since no pair bonds are being broken. See
Figure 14.1, where log10 (t1/2 ) is plotted vs. Q. Geiger and Nuttal proposed the following
phenomenological fit for log10 (t1/2 (Q)):
4 CHAPTER 14. α DECAY
log10 λ = C − DQ−1/2 or (14.12)
log10 t1/2 = −C ′ + DQ−1/2 , (14.13)
where C and D are fitting constants, and C ′ = C − log10 (ln 2). Odd-odd, even-odd and odd-
even nuclei follow the same general systematic trend, but the data are much more scattered,
and their half-lives are 2–1000 times that of their even-even counterparts.
Figure 14.1: Paste in Krane’s Figure 8.1.
Prediction of Q from the semiempirical mass formula
The semiempirical mass formula can be employed to estimate Qα , as a function of Z and A.
Q = B(Z − 2, A − 4) + B(e He) − B(Z, A)
8
≈ 28.3 − 4aV + aS A−1/3 + 4aC ZA−1/3 (1 − Z/3A) − 4asym (1 − 2Z/A)2 + 3ap A−7/4 .
3
(14.14)
A plot of Q(Z, A) using (14.14) is given in Figure 14.2.
These trends, from experimental data, are seen in Figure 14.3. However, there is also evidence
of the impact of the shell closing at N = 126, that is not seen in Figure 14.3.
14.4 Theory of α Emission
Figure 14.4 shows 3 potentials that are used in the estimation of barrier penetration probabil-
ities, for determining the halflife of an α-emitter. The simplest potential, the rectilinear box
potential, shown by the dashed line, although crude, maybe used to explain the phenomenon
of α decay.
The simplest theory of α emission
In this section we solve for the decay probability using the simplest rectilinear box potential.
This potential is characterized by:
14.4. THEORY OF α EMISSION 5
Q vs. A
10
8
Q (MeV)
160 170 180 190 200 210 220 230 240 250 260
A
Figure 14.2: Q(Z, A) using (14.14). The dashed line is for Pb (Z = 82), while the lower
dotted is for Os (Z = 76). The upper dotted line is for Lr (Z = 103). Each separate Z has
its own line with higher Z’s oriented to the right.
V1 (r < RN ) = −V0
V2 (RN < r < b) = VC
V3 (r > b) = 0 , (14.15)
where V0 and VC are constants.
The 3-D radial wavefunctions (assuming that the α is in an s-state), are of the form, Ri (r) =
ui (r)/r take the form:
Figure 14.3: Paste in Krane’s Figure 8.2.
6 CHAPTER 14. α DECAY
Potentials for α−barrier penetration, 190Pb
40
V
simple
Vcoulomb
30 Vigo
Q
20
10
V (MeV)
−10
−20
−30
−40
0 10 20 30 40 50 60 70
r (fm)
Figure 14.4: Potentials for α decay.
u1 (r) = Aeik1 r + Be−ik1 r
u2 (r) = Cek2 r + De−k2 r
u3 (r) = F eik3 r , (14.16)
where
p
k1 = 2m(V0 + Q)/~
p
k2 = 2m(VC − Q)/~
p
k3 = 2mQ/~ . (14.17)
Turning the mathematical crank, we arrive at the transmission coefficient:
14.4. THEORY OF α EMISSION 7
−1
sinh2 [2k2 (b − RN )] k1 k3 k22
1 k1 k3 k1 k3
T = 2+ + + + + + (14.18)
4 k3 k1 4 k22 k1 k3 k3 k1
Note: We solved a similar problem in NERS 311, but with V0 = 0, that is, k1 = k3 .
The factor k2 (b − RN ) ≈ 35 for typical α emitters. Thus, we can simplify to:
16e−2k2 (b−RN )
T = (14.19)
k1 k3 k22 k1 k3
k22
+ k1 k3
+ k3
+ k1
Recall that the transmission coefficient is the probability of escape by a single α-particle. To
calculate the transmission rate, we estimate a “frequency factor”, f , that counts the number
of instances, per unit time, that a α, with velocity vα presents itself at the barrier as an
escape candidate. There are several estimates for f :
f Source Estimate (s−1 ) Remarks
vα /RN Krane ≈ 1021 Too low, by 101 –102
vα /(2RN ) Others ≈ 5 × 1020 Too low, by 101 –102
Fermi’s Golden Rule # 2 ≈ 1024 Too high, by 101 –102
The correct answer, determined by experiment, lies in between these two extremes. However
approximate our result it, it does show an extreme sensitivity to the shape of the Coulomb
barrier, through the exponential factor in (14.19).
Gamow’s theory of α decay
Gamow’s theory of α decayis based on an approximate solution1 to the Schrödinger equation.
Gamow’s theory gives:
" 1/2 Z #
b
2m p
T = exp −2 dr V (r) − Q , (14.20)
~2 RN
where b is that value of that defines the r where V (r) = Q, on the far side of the barrier.
If we apply Gamow’s theory to the potential of the previous section, we obtain:
1
The approximation Gamow used, is a semi-classical approximation to the Schrödinger equation, called
the WKB (Wentzel-Kramers-Brillouin) method. The WKB method works best when the potential changes
slowly with position, and hence the frequency of the wavefunction, k(x), also changes slowly. This is not the
case for the nucleus, due to its sharp nuclear edge. Consequently, it it thought that Gamow’s solution can
only get to within a factor or 2 or 3 of the truth. In nuclear physics, a factor of 2 or 3 is often thought of as
“good agreement”!
8 CHAPTER 14. α DECAY
16
Texact = TGamow . (14.21)
k1 k3 k22 k1 k3
k22
+ k1 k3
+ k3
+ k1
That factor in front is about 2–3 for most α emitters. This discrepancy is usually ignored,
considering the large uncertainly in the f factor.
Krane’s treatment of α-decay
Krane starts out with (14.20), namely:
" 1/2 Z #
b
2m p
T = exp −2 dr V (r) − Q ,
~2 a
where
2(Z − 2)e2
V (x) =
4πǫ0 x
2(Z − 2)e2
V (a) ≡ B =
4πǫ0 a
a = R0 (A − 4)1/3
2(Z − 2)e2
V (b) ≡ Q = . (14.22)
4πǫ0 b
That is, the α moves in the potential of the daughter nucleus, B is the height of the potential
at the radius of the daughter nucleus, and b is the radius where that potential is equal to Q.
(See Krane’s Figure 8.3 on page 251.)
We note that a/b = Q/B.
Substituting the potential in (14.22) into (14.20) results in:
( 1/2 )
2m′α c2 zZ ′ e2 h √
i
p
T = exp −2 arccos( x) − x(1 − x) , (14.23)
Q(~c)2 4πǫ0
where x ≡ a/b = Q/B. Note that the reduced mass has been used:
mα mX ′
m′α = ≈ mα (1 − 4/A) . (14.24)
mα + mX ′
14.4. THEORY OF α EMISSION 9
This “small” difference can result in a change in T by a factor of 2–3, even for heavy nuclei!
Krane also discusses the approximation to (14.23) that results in his equation (8.18). This
comes from the Taylor expansion:
√ p π √
arccos( x) − x(1 − x) −→ − 2 x + O(x3/2 ) .
2
This is only valid for small x. Typically x ≈ 0.3, and use of Krane’s (8.18) involves too much
error. So, stick with the equation given below.
Factoring in the frequency factor, one can show that:
s
a mα c2
t1/2 = ln(2) ×
c 2(V0 + Q)
( 1/2 )
2m′α c2 zZ ′ e2 h √ p i
exp 2 arccos( x) − x(1 − x) . (14.25)
Q(~c)2 4πǫ0
14.4.1 Comparison with Measurements
In this section we employ the simplest form of f and compute the half-life for α-decay as
follows:
s ( 1/2 )
a mα c2 2m′α c2 zZ ′ e2 h √ p i
t1/2 = ln(2) exp 2 arccos( x) − x(1 − x)
c 2(V0 + Q) Q(~c)2 4πǫ0
(14.26)
The data are shown in the following table, where the half-lives of the even-even isotopes
of Th (Z = 90) are shown. The calculations were performed using a nuclear radius of
a = 1.25A1/3 (fm), and V0 = 35 (MeV). The absolute comparison exhibits the same trends
for both experiment and calculations, with the calculations being overestimated by 2–3 orders
of magnitude. This is most likely due to a gross underestimate of f . the relative comparisons
are in much better shape, showing discrepancies of about a factor of 2–3, quite a success
for such a crude theory. We note that small changes in Q result in enormous differences
in the results. In this table Q changes by about a factor of 2, while the half-lives span
about 23 orders of magnitude. The probability of escape is greatly influenced by the height
and width of the Coulomb barrier. Besides this dependence, the only other variation in the
comparison relates to the nuclear radius. This also affects the barrier since the nuclear radius
is proportional to A1/3 . This hints that the remaining discrepancy, at least for the relative
comparison, is related to the fine details of the shape of the barrier, perhaps mostly in the
vicinity of the inner turning point. A more refined shape of the Coulomb barrier would likely
10 CHAPTER 14. α DECAY
yield better results, as well would a higher-order WKB analysis that would account more
precisely, for that shape variation. Additionally, the α-particle was treated as if it were
a point charge in this analysis. A refined calculation should certainly take this effect into
account.
A Q (MeV) t1/2 (s) t1/2 (s) t1/2 (s) t1/2 (s)
abs. meas. abs. calc. rel. meas. rel. calc.
220 8.95 10−5 10−3 5.4 × 10−9 3.6 × 10−9
222 8.13 2.8 × 10−3 2.1 × 10−1 1.5 × 10−6 7.3 × 10−7
224 7.31 1.04 1.1 × 102 5.6 × 10−4 3.8 × 10−4
226 6.45 1854 2.9 × 105 ≡1 ≡1
228 5.52 6.0 × 107 1.2 × 1010 3.2 × 104 4.2 × 104
230 4.77 2.5 × 1012 6.0 × 1014 1.3 × 109 2.1 × 109
232 4.08 4.4 × 1017 1.8 × 1020 2.4 × 1014 6.2 × 1014
Table 14.1: Half-lives of Th isotopes, absolute and relative comparisons of measurement and
theory.
Cluster decay probabilities
If α decay can occur, surely 8 Be and 12 C decay can occur as well. It is just a matter of
relative probability. For these decays, the escape probabilities are given approximately by:
T8Be = Tα2
T12C = Tα3
Tax = Tαz/2 . (14.27)
The last estimate is for a a x cluster, with z protons and an atomic mass of a.
14.5 Angular momentum and parity in α decay
Angular momentum
If the α-particle carries off angular momentum, we must add the repulsive potential associ-
ated with the centrifugal barrier to the Coulomb potential, VC (r):
l(l + 1)~2
V (r) = VC (r) + , (14.28)
2m′α r 2
14.5. ANGULAR MOMENTUM AND PARITY IN α DECAY 11
represented by the second term on the right-hand side of (14.28).
90
The effect on Th, with Q = 4.5 MeV is:
l 0 1 2 3 4 5 6
Tl /T0 1 0.84 0.60 0.36 0.18 0.078 0.028
So, as l ↑, T ↓.
Conservation of angular momentum and parity
α decay’s must satisfy the constraints given by the conservation of total angular momentum:
I~i = I~f + I~α
Πi = Πf × Πα , (14.29)
where i represents the parent nucleus, and f represents the daughter nucleus. Since the
α-particle is a 0+ nucleus, (14.29) simplifies to:
I~i = I~f + ~lα
Πi = Πf × (−1)lα , (14.30)
where lα is the orbital angular momentum carried off by the α-particle. If Ii is non-zero, the
α decay is able to populate any excited state of the daughter, or go directly to the ground
state.
If the initial state has total spin 0, with few exceptions it is a 0+ . In this case, (14.30)
becomes.
~0 = I~f + ~lα
+1 = Πf × (−1)lα , (14.31)
or.
I~f = ~lα
Πf + = (−1)lα , (14.32)
12 CHAPTER 14. α DECAY
Thus the only allowed daughter configurations are: 0+ , 1− , 2+ , 3− , 4+ , 5− , 6+ , 7− , 8+ , 9− · · ·
All other combinations are absolutely disallowed.
The α decay can show these allowed transitions quite nicely. A particularly nice example is
the case where the transition is 0+ −→ 0+ , where the low-lying rotational band, and higher
energy phonon structure are explicitly revealed through α decay. (See Figure 8.7 in Krane.)
Angular intensity of α decays for elliptic nuclei
This is well described in Krane, pages 260–261.
14.6 α-decay spectroscopy
Not covered in NERS312.