We provide a semiclassical optics derivation of Einstein’s rate equations (ERE) for a two-level system illuminated by a broadband light field, setting a limit on their validity that depends on the spectral properties of the light field. Starting from the optical Bloch equations for individual atoms, the ensemble averaged atomic inversion is shown to follow ERE under two concurrent hypotheses: (i) the decorrelation of the inversion at a given time from the field at later times and (ii) a Markov approximation owing to the short correlation time of the light field. When the latter hypothesis is relaxed, we find effective Bloch equations for the ensemble average in which the atomic polarization decay rate is increased by an amount equal to the width of the light spectrum, which allows its adiabatic elimination for a large enough spectral width. The use of a phase-diffusion model of light allows us to check our results and hypotheses using numerical simulations of the corresponding stochastic differential equations. We take into account both light bandwidth and atomic linewidth, which allows us to discuss the differences existing between rate equations in the limits of large light bandwidth or large atomic linewidth. Only in the former case does one obtain ERE with the correct expression for Einstein’s B coefficient.

1.
H. A.
Lorentz
,
The Theory of Electrons
(
Teubner
,
Leipzig
,
1909
), Chap. 4.
2.
A.
Einstein
, “
Zur Quantentheorie der Strahlung
,”
Phys. Z.
18
,
121
128
(
1917
); an English translation (“Quantum theory of radiation”) appears in The World of the Atom, edited by H. A. Boorse and L. Motz (Basic Books, New York, 1966), Vol. 2, pp. 888–901.
3.
P. W.
Milonni
and
J. H.
Eberly
,
Lasers
(
John Wiley & Sons
,
New York
,
1988
).
4.
B. W.
Shore
,
The Theory of Atomic Coherent Excitation
(
John Wiley & Sons
,
New York
,
1990
).
5.
R.
Loudon
,
The Quantum Theory of Light
(
Oxford U.P.
,
New York
,
2000
).
6.
J. N.
Dodd
,
Atoms and Light: Interactions
(
Plenum Press
,
New York
,
1991
).
7.
It is important to note that standard semiclassical theory does not explain spontaneous emission. This means that the derivation of the A coefficient requires quantization of both matter and the electromagnetic field (see, e.g., Ref. 5). On the contrary, the standard semiclassical theory allows one to derive an expression for the B coefficient, yielding the same result as obtained with the fully quantized theory as we show here. We must remark, however, that there is at least one formulation of the semiclassical theory that accounts for spontaneous emission: Self-field Quantum Electrodynamics, the theory developed by A. O. Barut and collaborators during the 1980s. See
A. O.
Barut
and
J. P.
Dowling
, “
Self-field quantum electrodynamics: The two-level atom
,”
Phys. Rev. A
41
,
2284
2294
(
1990
) and references therein.
8.
S. M.
Barnet
and
P. M.
Radmore
,
Methods in Theoretical Quantum Optics
(
Oxford U.P.
,
New York
,
2003
).
9.
Furthermore, for Δ0 (i.e., for a fully coherent field characterized by a constant Rabi frequency Ω0), Eqs. (30) and (31) are equivalent to Eqs. (5) and (6) for Ωα(t)=Ω0 α upon identifying q¯ with iAσ/Ω0 and W21 with Ω02/Δ. For Ω, it can be shown that Reσ21=0 in Eqs. (5) and (6). Hence, Imσ21 can be identified with q¯ in this case.
10.
This is similar to what happens with the application of Fermi's golden rule to the photo-ionization problem. See
H.
Fearn
and
W. E.
Lamb
, “
Corrections to the golden rule
,”
Phys. Rev. A
43
,
2124
2128
(
1991
).
11.
A random process ξ(t) of zero mean is said to be Gaussian if its nth-order correlation function satisfies
for n even, and zero for n odd. That is, for Gaussian noise of zero mean all moments are known if the second-order moment is known. The Gaussian noise is said to be white when its second order moment satisfies Eq. (46) (see Ref. 12).
12.
L.
Mandel
and
E.
Wolf
,
Optical Coherence and Quantum Optics
(
Cambridge U.P.
,
New York
,
1995
).
13.
P. E.
Kloeden
and
E.
Platen
,
The Numerical Solution of Stochastic Differential Equations
(
Springer
,
Berlin
,
1995
).
14.
G. E. P.
Box
and
M. E.
Muller
, “
A note on the generation of random normal deviates
,”
Ann. Math. Stat.
29
,
610
611
(
1958
).
AAPT members receive access to the American Journal of Physics and The Physics Teacher as a member benefit. To learn more about this member benefit and becoming an AAPT member, visit the Joining AAPT page.