MATH 235, by T.

Lakoba, University of Vermont 123

13 Black-body radiation and Planck’s formula

This concluding Lecture contains a lighter load of mathematics than most of the other Lec-
tures. From the mathematical perspective, we will see some applications of improper integrals
and infinite series, usually studied in a second semester of Calculus. At the same time, this
Lecture will also attempt to describe the historical background of the emergence of Quantum
Mechanics — the area of physics on which most modern technology (and much of science) is
based.
In Section 1, we begin by providing such a historical background. We will only glance
over related mathematical details, but focus on what prompted Max Planck to derive his
famous formula that revolutionized physics. In subsequent sections, we will more carefully
examine his and his scientific opponents’ derivations of that formula and then analyze some of
its elementary consequences. Finally, in the last section, we will mention a curious modification
of Planck’s formula (made by Planck himself), which was confirmed much later in Quantum
Electrodynamics.

13.1 Historical background of Planck’s formula

13.1.1 The historical stage before Planck
In the middle of the 19th century, physicists actively studied absorbtion and emission of radia-
tion by heated objects, or thermal radiation. An object whose temperature is higher than the
ambient temperature was known to emit more radiation that it absorbs from its surroundings.
Coversely, a cooler object absorbs more radiation than it emits. Thus, thermal radiation is the
main mechanism by which an object comes into thermal, or thermodynamic, equilibrium
with its surroundings and hence, in this sense, it is related to Newton’s cooling law mentioned
in Lecture 9.
What is thermal radiation made of? It is now a well established scientific fact that all
macroscopic objects consist of atoms, and atoms themselves consist of charged particles. At
temperatures above absolute zero (which includes all practical cases around us), the atoms are
in a state of constant and chaotic motion. It is also well known now that moving charged
particles emit electromagnetic waves, and frequencies of those emitted waves depend on the
particles’ accelerations. Thus, thermal radiation consists of electromagnetic waves of various
frequencies or, equivalently, wavelengths. We will refer to the wavelength in this section but
will convert to the frequency “picture” later on, in Section 2. The notation for the wavelength
is λ.
Note that the above description has been based on a number of modern concepts, which
were not known or accepted in the middle of the 19th century. For example, the theory of
electromagnetic waves (and the emission of such waves by accelerating charged particles) was
developed in early 1870’s by James Clerk Maxwell and experimentally confirmed in late 1880’s
by Heinrich Hertz, while the first active studies of thermal radiation laws occurred in 1850’s
(see below). So it may be an interesting project in physics history to understand how radiation
was viewed back then. We, however, will not consider this topic here.
By late 1850’s, is has been known that objects made of different materials and kept at
different temperatures emit different amounts of radiation. Let Eλ dλ denote the amount of
radiation energy emitted by an element of the object’s surface within a small solid angle and
into a wavelength interval [λ, λ + dλ]. Also, suppose that the object’s surface is irradiated by,
say, light, which is a form of radiation. Part of this radiation is reflected, and part absorbed.
MATH 235, by T. Lakoba, University of Vermont 124

Let Aλ denote the percentage of the radiation absorbed. It has been known by late 1850’s
that Eλ and Aλ are related in such a way that poor absorbers (i.e., good reflectors) are also
poor emitters.
Then, in 1859, the great German physicist Gustav Kirchhoff stated a law of thermal
radiation named after him:
For an object that is in thermal equilibrium with its surroundings14 , one has
Eλ
= Kλ (T ) , (13.1)
Aλ
where Kλ (T ) is a constant that depends only on the object’s temperature T (and also on the
wavelength), but is independent of the material and shape of the object.
Thus, while objects made of different materials emit and absorb radiation differently, their
ratio of the emission and absorbtion coefficients as defined by (13.1) is independent of the
material.
For an absolutely absorbing (i.e., “black”) body, which absorbs all the incident radiation,
Aλ = 1 for all wavelengths. Then from Eq. (13.1) it follows that

Kλ (T ) = Eλ |black body .
Thus, the meaning of the universal (i.e.,
material- and shape-independent) constant
Kλ (T ) is that it is the density of radiation
(per wavelength interval) emitted by a black
body. Incidentally, since the absorption by
a black body is the maximum and the ratio
of the emitted to absorbed amounts of radi-
ation is the same for all materials, then the
emittance of the black body is also the max-
imum among all objects. This observation
is consistent with a previously noted fact,
known before Kirchhoff, that poor absorbers
are also poor emitters and, conversely, good
absorbers are also good emitters.

At this point it is instructive to ask a question: What is, then, a black body? A good
physical approximation to a black body is a tiny hole made in a cavity with no other openings,
as schematically shown in the figure above. Any ray of radiation entering the hole will be
reflected by the cavity’s walls and eventually absorbed by them. On the other hand, due to
thermal chaotic motion, the molecules of the walls emit some radiation, in addition to also
absorbing it. After a “sufficiently long time”, equilibrium between the walls and the radiation
inside the cavity is reached. In the equilibrium, a detailed balance takes place: at any given
time, the amount of the emitted radiation of a given wavelength, polarization, and direction
equals, on average, the amount of absorbed radiation with the same properties. In other words,
importantly for a later discussion, there is a thermodynamic equilibrium between the walls and
the radiation.
Thus, in practice, radiation from any cavity with a tiny hole very closely approximates
the radiation of a black body. A familiar example of a cavity with a hole is a building with
14
e.g., any object in a classroom where a constant temperature has been maintained sufficiently long, or a
star in outer space
MATH 235, by T. Lakoba, University of Vermont 125

windows. The windows appear darker than the outside walls. One can say that they appear
approximately black, as most light that gets into a window never reflects back to the observer;
in other words, it is absorbed, regardless of its wavelength.
The above description makes no reference to the material of the inner walls of the cavity.
That is, it implies that no matter what this material may be, a tiny hole in a cavity will radiate
approximately as an ideal black body. You may want to know that while this statement is
widely accepted in modern physics, its is still not entirely undisputed. For example, the paper
by P.-M. Robitaille posted alongside this Lecture defends an alternative point of view, namely
that the radiation of an actual cavity depends both on the material of the cavity’s inner walls
and also on the angle at which this radiation is measured. (However, the quantitaive measure
of this dependence has remained unclear to me.)
Returning now to the Kirchhoff’s radiation law, it is evident that the total energy of radiation
emitted by the black body into all wavelengths is the function of only the temperature T but
not of the material the body is made of:
∫ ∞
Kλ (T )dλ = u(T ). (13.2)
0

In 1879, a Slovene physicist, mathematician, and poet Jožef Stefan, who lived and worked
in Vienna, Austria, formulated the Stefan–Botlzmann law for the total radiated energy15 :

u(T ) = const · T 4 . (13.3)

By the mid-1890’s, some experimental data for the spectral density Kλ (T ) of the emitted
radiation were collected for short (see below) wavelengths. Data for longer wavelengths were
not available at that time due to experimental limitations. A number of researchers proposed
several different analytical forms for Kλ (T ), basing their considerations on two criteria: (i)
provide a fit for the available data; and (ii) satisfy the Stefan–Boltzmann law (13.3).
In 1895, a German physicist, an experimentalist Friedrich Paschen proposed an expression

Kλ (T ) = b · λ−γ · e−a/(λT ) (13.4)

with some constants a, b and with γ ≈ 5.7 to fit his own latest experiments. To this, another
German physicist, a theoretician Wilhelm Wien pointed out that one must have γ = 5 to satisfy
the Stefan–Boltzmann law. Indeed, the intergration of (13.4) using u-substitution yields:
∫ ∞ ∫ ∞
x = λT
Kλ (T )dλ = b · λ−γ · e−a/(λT ) dλ
dλ = dx
0
∫ ∞ ( )−γ
0 T
x dx
e− x ·
a
= b
T T
∫
0
∞
x−γ e− x dx · T γ−1 = const · T γ−1 ,
a
= b· (13.5)
0

which coincides with (13.3) and (13.2) for γ = 5. In an attempt to obtain a better agreement
with the theory, Paschen redid his experiments and fit the data to expression (13.4) with
γ ≈ 5.2.
15
Stefan deduced this law from experimental measurements made by an Irish physicist John Tyndall. Ludwig
Boltzmann, who was a former student of Stefan and by the 1880’s had become one of the greatest physicists of
his time, used a thermodynamics framework to derive this law theoretically in 1884. We will encounter other
contributions by Boltzmann later in this Lecture.
MATH 235, by T. Lakoba, University of Vermont 126

In 1899, a German physicist Max Planck rederived Wien’s formula (i.e., (13.4) with γ =
5) from phenomenological thermodynamical considerations. However, later in that year, two
German physicists Otto Lummer and Ernst Pringsheim obtained new experimental data for
longer wavelengths (λ = 12 ÷ 18 µm) showing that there was a systematic deviation from the
Wien–Planck’s formula. In 1900, Planck gave two different derivations of another, new formula
that matched the latest experimental data.

13.1.2 Planck’s discovery

In 1900, Max Planck was 42 years old and had an established name in thermodynamics. In
particular, it was he who stated the Second Law of Thermodynamics in the for well-known
today: “If a system evolves in thermal and mechanical isolation from the ambient environment,
then its entropy increases.” (Entropy is a measure of disorder of the system. Originally, the
Second Law was formulated in different forms by the mathematical physicists Rudolph Clau-
sius, of Germany, and William Thompson (Lord Kelvin), of Ireland, in the 1850’s). A corollary
of Planck’s formulation of the Second Law is this: Since a system in thermal and mechanical
isolation is known to evolve toward thermodynamic equilibrium, then in this equilibrium, the
system’s entropy must be a maximum. It is this connection between a thermodynamic equilib-
rium and the entropy that motivated Planck’s interest in the black-body radiation theory. We
explain this in some detail now.
At the end of the 19th century, there was much criticism of Thermodynamics, based on a
simple question: How can the time-reversible laws of Mechanics lead to time-irreversible laws
of Thermodynamics? (For example, the evolution of any system towards a thermodynamic
equilibrium is a time-irreversible process.) The main figure in Statistical Thermodynamics was
the great Austrian physicist Ludwig Boltzmann, who advocated the irreversibility of Thermo-
dynamics, but could not explain the irreversibility satisfactorily. (This was not Boltzmann’s
fault; a satisfactory argument for such an explanation was developed only in the second half of
the 20th century — this is the theory of dynamical chaos.)
Planck wanted to resolve the above Mechanics-versus-Thermodynamics paradox. He turned
to the black-body radiation probem for two reasons. First, the black-body radiation is in
thermodynamic equilibrium with its source (see the paragraph after the figure in Section 13.1.1),
so its entropy should be at a maximum (according to the Second Law of Thermodynamics).
Second, radiation was perceived as a continuous substance, as opposed to gases, that consist of
discrete molecules. So Planck hoped to find the cause for the irreversibility in the fact that a
system can be viewed as continuous rather than discrete. What happened instead was that at
the end, Planck had to assume that radiation was also discrete!
Thus, as we said earlier, in 1900 Planck first retrieved the Wien’s law (Eq.(13.4) with γ = 5),
but after Lummer and Pringsheim pointed out its deviation from their experimental data, he
modified his derivation to arrive at
b · λ−5
Kλ (T ) = (13.6)
ea/λT − 1
(for some other constants a and b than in (13.4)). His first derivation of this formula, done
in October 1900, was based solely on phenomenological Thermodynamics and required no
assumptions about microscopic properties of radiation. Remarkably, the agreement between
his new formula (13.6) and the most recent experimental data was very good.
Next, Planck attempted to rederive this “good” formula using microscopic considerations
of Statistical Mechanics developed by Boltzmann. He succeeded and reobtained such a formula
MATH 235, by T. Lakoba, University of Vermont 127

in December 1900. He assumed that the radiation could be subdivided into discrete chunks
(quanta) of energy ε. This was a common trick of continuum mechanics: assume discreteness
and then pass to the continuous limit. Probably, this was also Planck’s intention. But he found
that his formula (13.6) could be obtained only if ε was taken to be a specific function of the
radiation’s frequency ν:
ε = hν, (13.7)
where he found a value for the constant h to be close to its value
h = 6.626 · 10−27 erg · s
cm2
that we accept today. (1 erg = 1g · s2
.) This constant h is the famous Planck constant.

13.1.3 Subsequent historical development

In 1906, Planck wrote “Lectures on the Theory of Thermal Radiation”, where, among other
things, he gave a detailed derivation of his formula. It is often said that since the time Planck
derived that formula in 1900 using the assumption of discrete nature of radiation, he was trying
(unsuccessfully) to get rid of that assumption. However, according to a comprehensive account
of Planck’s discovery in [T.S. Kuhn, “Black-body theory and quantum discontinuity 1894-1912,”
Clarendon Press, Oxford/Oxford University Press, New York, 1978; p. 126], this was not the
case. Overcoming the discreteness assumption was not a central theme of Planck’s research.
While in his original microscopic derivation in December 1900 he assumed that the molecules
of the walls emitted quanta of energies nε = nhν, when he first published that derivation in
1901, he wrote that the molecules could emit radiation with energies lying between nhν and
(n+1)hν. Both derivations gave an almost identical result, as we will show in the last section of
this Lecture. Thus, Planck did not consider the radiation being intrinsically discrete, although
he had to introduce discreteness into his derivations.
The realization that Planck’s derivation did mean that the radiation actually consists of
discrete quanta, appeared gradually as a result of contributions to this issue by a number of
notable scientists, including: Hendrik A. Lorentz (of Holland), Paul Ehrenfest (of Austria),
Max von Laue (of Germany), and Albert Einstein (at that time, Einstein was young and little-
known). In particular, Lorentz gave a new derivation of Planck’s formula in 1910, and Einstein
gave yet another, totally different derivation of it in 1916 when he wrote a ground-laying paper
on induced and spontaneous radiation.
On the other side of the barricades in 1900–1905 were two prominent opponents of Planck’s
formula: Lord Rayleigh16 and an English physicist James Jeans17 . In 1900, Rayleigh proposed
a formula
Kλ (T ) = bλ−4 · T · e−a/(λT ) , (13.8)
where the pre-exponential factor had some phenomenological explanation, while the e−a/(λT )
was brought in “by brute force” to provide agreement wih the experiment. In 1905, Jeans used
Maxwell’s electrodynamic equations to rigorously derive that
Kλ (T ) = bλ−4 · T, (13.9)
16
John William Strutt, 3rd Baron Rayleigh, was the great English physicist who, with William Ramsay,
discovered the element argon, an achievement for which he earned the Nobel Prize for Physics in 1904. He also
made fundamental contributions to the theories of wave propagation, optics, acoustics, and fluid dynamics, where
many phenomena now bear his name. Rayleigh was a doctoral advisor of J.J. Thompson and G.P. Thompson,
who will be mentioned later in this Lecture. Lord Rayleigh was one of the very few members of higher nobility
who won fame as an outstanding scientist.
17
Jeans was knighted in 1928 for his contributions in Astronomy.
MATH 235, by T. Lakoba, University of Vermont 128

where he also derived a value for constant b. He argued that for small λ, for which his formula
was in blatant contradiction with both the experiments and common sense (as we will show in
a later section), Maxwell’s equations were not applicable for some unknown reason. It may be
interesting to note that later Jeans not only converted to the Quantum theory, but also became
one of its first proponents in England.
After 1911, the black-body radiation theory was overshadowed by a newly emerged topic of
specific heat calculation and measurement for solid-state substances. Planck’s formula, however,
found experimental confirmation there as well.
A critical development that eventually propelled Planck’s discovery into its prominent place
occured in 1913. In that year, a young (and later, the great) Danish physicist Niels Bohr related
Planck’s hypothesis of discretness of radiation with two then-unexplainable phenomena inside
the atom: the atom’s stability and radiation spectra emitted by atoms. A couple years before
that, in 1911, Ernst Rutherford, based on the results of his experiments carried out at the
University of Manchester, proposed the planetary model of an atom. (An earlier model,
proposed by J.J. Thompson18 in 190419 , considered an atom as a pudding, with electrons being
included there as raisins.) There was a problem with Rutherford’s planetary model, however.
An electron rotating about the nucleus has centripetal acceleration. According to Maxwell’s
electromagnetic theory, any accelerating charged particle must emit radiation. Therefore, a
rotating electron would constantly emit radiation and hence lose energy, so that eventually
it would fall into the nucleus. Rutherford was well aware of this problem, but insisted that
in spite of it, an atom still had to look like the Solar System. Niels Bohr, who was a young
researcher in Rutherford’s lab, came up with a geniously simple solution: An electron cannot
emit continuously, but only by quanta. Therefore, when it orbits the nucleus, the electron does
not emit at all (because it cannot emit part of a quantum), and hence the atom is stable. The
only possibility for an electron to emit a quantum is when it goes (for whatever reason) from one
stationary orbit to another. Calculations that Bohr did using this principle yielded the first-
ever theoretical explanation of experimentally observed atomic radiation spectra. Thus, Bohr’s
ingenious idea of the connection between Planck’s quantum hypothesis and atomic physics
paved the way to the creation of Quantum Mechanics.

13.2 Density of states of the radiation

In this and the next Sections, we will set the stage for the derivation of Planck’s formula (13.6),
which will follow in Section 13.4.
In Section 13.1 we referred to the wavelength of radiation, as it was done in the actual
development of this theory in the 19th century. In this and the following Sections we will
“change the variable” from the wavelength λ to the angular frequency ω of the radiation.
18
Joseph John Thompson, the great British physicist, made fundamental contributions to the theory of
electrical conductance in gases and discovered the electron in 1897. His numerous awards included the Nobel
Prize, the Knighthood, and the Order of Merit. He was a mentor to seven future Nobel laureates, one of whom
was Rutherford. J.J.’s son, Sir G.P. Thompson, also won the Nobel Prize in 1937 for proving the wavelike
properties of electrons.
19
In the same year, a Japanese physicist Hantaro Nagaoka proposed the first, although incorrect, planetary
model of the atom. It was based around the analogy of the theory of the stability of the Saturnian rings,
proposed by Sir James Clerk Maxwell in 1859.
MATH 235, by T. Lakoba, University of Vermont 129

Any wave can be characterized by: the wave-

length λ, the speed c, the period τ = λ/c, the
frequency ν = 1/τ = c/λ, or, equivalently,
by the angular frequency
2πc
ω = 2πν = . (13.10)
λ
In what follows we will refer to both the fre-
quency ν and the angular frequency ω as sim-
ply the frequency, since this will not lead to
a confusion of the two.

Let us recall that our final goal is to derive Eq. (13.6) for the radiation spectral density
Kλ (T ). Since we now are using the frequency instead of the wavelength in our description, we
need to relate Kω (T ) with Kλ (T ). This is done as follows. The energy of radiation emitted
within a wavelength interval [λ, λ + ∆λ] can be written in two ways:
Kλ (T )dλ = Kω (T )dω. (13.11)
The l.h.s. of this equation is merely the definition of the spectral density Kλ (T ). The r.h.s.
expresses the fact that λ and ω are related by a one-to-one function (13.10). Then, Eqs. (13.11)
and (13.10) provide a relation between Kλ (T ) and Kw (T ):
dω 2πc
Kλ (T ) = Kω (T ) = Kω (T ) · 2 . (13.12)
dλ λ
(Note that this change of variables is analogous to those we did in Lecture 4.)
To begin the derivation of Kω (T ), consider a large box with some dimensions Lx , Ly , Lz ,
as shown below. The density of the radiated energy then equals:
 
number of frequencies in the
( )
 interval [ω, ω + dω] in the box  average energy of one radiation
 
K (T ) dω = 
| ω{z } volume of the box · mode of frequency ω
.
average energy
per ω-interval
per volume
(13.13)

The qualifier “average” above is used because

the radiation is in equilibrium with its source
on average over some macroscopic interval of
time. In this Section we will estimate the first
term on the r.h.s. of (13.13). (It is possible to
compute this term exactly using the Maxwell
equations for the radiation, but this will not
be required for our purposes.)
MATH 235, by T. Lakoba, University of Vermont 130

First we note that

( )
number of frequencies in the dZ
= dω, (13.14)
interval [ω, ω + dω] in the box dω
where Z(ω) is the number of frequencies up to ω that can exist in this box. From Maxwell’s
equations describing the electromagnetic radiation in vacuum (and our box is assumed to
contain vacuum, i.e., no matter), it can be shown that allowed frequencies of the radiation
propagating in any one direction are spaced evenly (see an explanation after Eq. (13.17) below).
Then,
( ) ( ) ( )
ω ω ω
Z(ω) = const · · · , (13.15)
ω ω ω
| min,x
{z } | min,y
{z } | min,z
{z }
number of waves number of waves number of waves
in x-direction in y-direction in z-direction
where ωmin,x is the minimum frequency of radiation that can propagate in the box in the
x-direction; and similarly for ωmin,y and ωmin,z .
This minimum frequency exists because
there is the maximum wavelength,

λmax,x = 2Lx (13.16)

that can exist between the walls located Lx

units apart. (The illustrating figure on the
left assumes that the wave is zero at the
walls, but a similar result can also be ob-
tained for other boundary conditions.)

Now, from (13.16) and (13.10),

2πc πc
ωmin,j = = , j = {x, y, z}. (13.17)
2Lj Lj
From the Figure above one can conclude that the next two largest wavelengths are 2Lx /2 and
2Lx /3, where, respectively, two and three semi-periods of the wave fit between the walls. The
corresponding frequencies, in analogy with (13.17), are 2ωmin,x and 3ωmin,x . This illustrates the
statement, made before Eq. (13.15), that the frequencies of the radiation in a box are spaced
evenly.
Next, from (13.15) and (13.17) one has:
ω3
Z(ω) = const ·
(πc)3 /(Lx Ly Lz )
ω3
= const · · Lx Ly Lz . (13.18)
(πc)3
An exact solution of Maxwell’s equations for the radiation’s electromagnetic field with two
transverse polarizations yields the value of the constant in (13.18) to be:
π
const = . (13.19)
3
MATH 235, by T. Lakoba, University of Vermont 131

Then, according to (13.14) and (13.18), (13.19):

( )
number of frequencies in ω 2 dω
= 2 3 · Lx Ly Lz , (13.20)
[ω, ω + dω] in the box π c
and finally, ( )
number of frequencies in
[ω, ω + dω] in the box ω 2 dω
= 2 3. (13.21)
volume of the box π c
We have derived the first term on the r.h.s. of (13.13). To proceed with the derivation of the
second term, we need some elementary background in the statistical mechanics of gases. This
is given in the next Section.

13.3 Elementary background from the Maxwell–Boltzmann kinetic

theory of gases
In gases, molecules move chaotically, and so different molecules usually have different velocities,
both in their direction and in magnitude. Therefore, one can speak about the probability to
find a molecule within a given interval of velocities at a certain location in space. Let

p(vx , vy , vz )dvx dvy dvz (13.22)

be the probability to find a molecule with the x-, y-, and z-components of its velocity being
within the intervals [vx , vx + dvx ], [vy , vy + dvy ], [vz , vz + dvz ].
A fundamental law derived by James Clerk Maxwell and Ludwig Boltzmann at the end of
the 19th century states that for gases, or for any other systems of particles, in thermodynamic
equilibrium,
p(vx , vy , vz ) = const · e−E/(kT ) , (13.23)
where T is the temperature, k is the Boltzmann constant, and E is the energy of the particle.
For example, if one neglects the potential energy of the particles (such as the gravitational
potential energy of the molecules in a gas), then
m 2
E = Ekinetic = (v + vy2 + vz2 ). (13.24)
2 x
A corollary of the Maxwell–Boltzmann law that is important for us is the so-called energy
equipartition theorem:
In thermodynamic equilibrium, the average energy corresponding to each “degree of freedom” of
a particle is 12 kT .
This fact is derived in courses on thermodynamics and statistical physics.20 The “degrees
of freedom” in this theorem are closely related to the coordinates q introduced in Lecture 7
in connection with the Lagrangian, and their time derivatives, q̇. An example important for
further development of our theory is that of a harmonic oscillator, considered in Lecture 6 – 8.
The energy of such an oscillator, which equals the sum of its kinetic and potential energies, is
mq̇ 2 βq 2
Eoscillator = + ,
2 2
20
A more precise statement of this theorem requires that the energy depend on the “degrees of freedom”
quadratically.
MATH 235, by T. Lakoba, University of Vermont 132

where β is some proportionality constant. Thus, an oscillator has two “degrees of freedom”,
q and q̇. If an oscillator comes, e.g., by means of collisions with surrounding particles, into
thermodynamic equilibrium with the ambient matter, then, according to the equipartition
theorem, its average energy is
1
⟨Eoscillator ⟩ = 2 · kT = kT, (13.25)
2
where the factor ‘2’ occurs because the oscillator has two “degrees of freedom”. The result
expressed by Eq. (13.25) was widely accepted by physicists in the late 19th century.
We are now prepared to derive an expression for the second term on the r.h.s. of Eq. (13.13)
and thereby obtain Planck’s formula.

13.4 Rayleigh–Jeans and Planck’s formulae

First, however, we will obtain the Rayleigh–Jeans Eq. (13.9) (even though this formula both
is incorrect and, historically, was derived after the correct Planck’s formula). We will do so to
highlight the key difference between the two formulae.
Recall that to complete the derivation of the spectral density Kω (T ) in Eq. (13.13), we
need to find the average energy of one radiation mode of frequency ω. By the end of the
19th century it was well-known to physicists that a mode of radiation of a given frequency is
mathematically equivalent to an oscillating string, or, in other words, to an oscillator. This
follows directly from Maxwell’s equations of the electromagnetic theory. Therefore, in 1905,
Jeans made the following connection: (i) a radiation mode is analogous to an oscillator; (ii) the
average energy of an oscillator that is in thermodynamic equilibrium with its surroundings is
kT (see Eq. (13.25)); (iii) hence the average energy of one frequency mode of the black-body
radiation, which is known to be in thermodynamic equilibrium with its source (see Section
13.1.1), equals kT , independently of the mode’s frequency! Substituting this result into (13.13)
and also using our earlier result (13.21), one obtains the Rayleigh–Jeans formula:
ω2
Kω (T ) = · kT. (13.26)
π 2 c3
To verify that Eqs. (13.26) and (13.9) are equivalent, we utilize the relation between Kλ (T )
and Kω (T ) given by Eq. (13.12) (see also (13.10)):
( 2 )
ω 2πc
Kλ (T ) = 2 3
· kT · 2
π c λ
( ( )2 )
2πc
2πc 8π
= λ
2 3
· kT · 2 = 4 kT. (13.27)
π c λ λ

This is the Rayleigh–Jeans formula in the form equivalent to (13.9).

We will now give a derivation of Planck’s formula, Eq. (13.6). It should be noted that the
way we will do it is not how Planck originally derived his formula in 1900 or even re-derived it
in 1906 (see Section 13.1.2).
Let us assume that a radiation mode consists of discrete quanta, with the energy of each
quantum being (see (13.7)):
h
ε = hν ≡ ~ω, ~≡ , (13.28)
2π
MATH 235, by T. Lakoba, University of Vermont 133

where ω is the frequency of the mode. A mode can have several quanta in it. If it contains n
quanta, then the energy of such a mode is nε = n~ω. Then the problem of finding the average
energy of a mode with frequency ω is mathematically equivalent to that of finding the average
number of quanta in such a mode. A key and nontrivial step is to realize that the probability
of having a discrete black-body radiation mode with energy E is:

P (E) = const · e−E/(kT ) , (13.29)

where the constant in (13.29) is chosen so as to have:

∑
P (E) = 1. (13.30)
all discrete E
At first glance, Eq. (13.29) may look just like the Boltzmann law (13.23). However, the
nontrivial part of the connection between these two formulae is that while (13.29) defines
a probability (i.e. a finite number), Eq. (13.23) defines the probability density, so that the
corresponding probability (13.22) is an infinitesimally small number. More importantly, note
that the integration variables in (13.22) are vx , vy , vz , and not E. Therefore, if we were to
compute the probability density p(E) from

p(E)dE = p(vx , vy , vz )dvx dvy dvz ,

which is a counterpart of Eq. (13.11), then the proportionality constant between p(E) and
p(vx , vy , vz ) would depend on E (similarly to how the proportionality constant between Kλ (T )
and Kω (T ) in (13.12) depends on λ). Consequently, the probability density of having a radiation
mode with a continuously changing energy E would not be

p(E) ∝ e−E/(kT ) .

Nevertheless, for the probability (as opposed to a probability density) of having a radiation
mode with discrete values of E, the Boltzmann law (13.29) holds.
Returning to the problem of finding the average energy of a black-body radiation mode, we
write this average energy, ⟨E⟩, using the standard formula of the probability theory for average
values: ∑
⟨E⟩ = E · P (E), (13.31)
all discrete E
where E = n~ω. We now need to determine the constant in the formula (13.29) for P (E). This
follows from the normalization equation (13.30):

∑
∞
const(13.29) · e−n~ω/(kT ) = 1,
n=0

whence
1
const(13.29) = . (13.32)
∑
∞
−n~ω/(kT )
e
n=0

Note that the inclusion of the term with n = 0 into the sum in (13.32) means, physically, that
one has to allow for a nonzero probability for the mode to contain no quanta. In other words,
MATH 235, by T. Lakoba, University of Vermont 134

there is a nonzero probability to have no radiation with a given frequency at any instance of
time.
Substituting (13.28), (13.29), and (13.32) into (13.31), one has:

∑
∞
(n~ω)e−n~ω/(kT )
n=0
⟨E⟩ = . (13.33)
∑
∞
−n~ω/(kT )
e
n=0

Such an expression is calculated using the following trick. Denote

~ω
− ≡ x;
kT
then (13.33) becomes:

∑
∞
d ∑ nx
∞
nx
ne e
dx n=0
⟨E⟩ = ~ω n=0 = ~ω
∑
∞ ∑
∞
nx
e enx
n=0
(∞ ) n=0 (∞ )
d ∑ d ∑
= ~ω ln enx = ~ω ln (ex )n
dx n=0
dx n=0 geometric series
( )
d 1 e x
~ω
= ~ω ln = ~ω = −x .
dx 1 − ex 1−ex e −1

Finally,
~ω
⟨E⟩ = . (13.34)
−1
e~ω/(kT )
Substituting this expression, along with (13.21), into (13.13), one obtains Planck’s formula:

ω2 ~ω
Kω (T ) = 2 3 · ~ω/(kT ) . (13.35)
π c e −1
The equivalence of (13.35) and (13.6) is established following the lines of the calculations in
Eq. (13.27).
The above derivation of Planck’s formula is attributed to H.A. Lorentz (1910). The original
Planck’s derivations of this formula, that used the concept of entropy, may be found in the
book by T.S. Kuhn, “Black-body theory and the quantum discontinuity 1894–1912,” which we
have referenced in Section 13.1.3.

13.5 Corollaries of Rayleigh–Jeans and Planck’s formulae

As we mentioned in Section 13.1.3, the Rayleigh–Jeans formula (13.26) is in blatant contra-
diction with common sense. Indeed, let us use that formula to compute the total amount of
energy radiated by a black body into all frequencies:
∫ ∞ ∫ ∞
kT
u= Kω (T )dω = 2 3 ω 2 dω = ∞. (13.36)
0 π c 0
MATH 235, by T. Lakoba, University of Vermont 135

Paul Ehrenfest called this problem of the Rayleigh–Jeans formula the ultraviolet catastro-
phe. (A connection here is the following. The divergence of the integral in (13.36) occurs for
large ω; large frequencies ω correspond to small wavelengths λ (see (13.10)); and finally, the
color of light changes from red to violet as the light’s wavelength decreases.) Jeans, of course,
was well aware of this problem, and so before he eventually converted to the Quantum theory,
he had been insisting that Maxwell’s electromagnetic equations, on which the derivation of the
term ω 2 /(π 2 c3 ) in (13.21) was based, were not applicable for very short wavelengths for some
unknown reason.
We will now show that Planck’s formula (13.35): (i) predicts a finite value for the total
radiated energy and (ii) for small frequencies reduces to the Rayleigh–Jeans formula (so, in
that region, both formulae are consistent with the experiment). We will begin with item (ii).
Let ~ω ≪ kT . Then, using the first two terms of the Maclaurin series for e~ω/(kT ) in Eq.
(13.34), one has:
~ω
⟨E⟩ |~ω≪kT ≈ ( ~ω
) = kT ; (13.37)
1 + kT −1
this is precisely the value for the average energy of a radiation mode that Jeans used in deriving
(13.26).
Now let us compute the total amount of radiated energy using Planck’s formula:
∫ ∞ ∫ ∞ 2
ω ~ω dω
u(T ) = Kω (T )dω = · hω/(kT ) .
0 0
2 3
π c e −1

Following the approach used in Eq. (13.5), we make a variable substitution

~ω kT kT
y= , ω=y , dω = dy ·
kT ~ ~
in the above integral to obtain:
∫ ∞
(kT )4 y 3 dy
u= 3 2 3 . (13.38)
~π c 0 ey − 1

It remains to show that the improper integral in (13.38) converges. First, we check that there
is no divergence at y ≈ 0:
y3 y3
≈ = y2;
ey − 1 y≪1 (1 + y) − 1
this shows that the integrand is a continuous, and hence, integrable, function at y = 0. At the
other end of the integration interval, i.e. for y → ∞, one has

y3
≈ y 3 e−y ,
e −1
y

which decays faster than any power of y. Hence, the integral converges at that limit also.
Finally, the exact value of the integral can be found with Mathematica21 to be (π 4 /15). Sub-
stituting this in (13.38) yields the total radiated energy being in agreement with the Stefan–
Boltzmann law (13.3):
k4π2
u(T ) = · T 4. (13.39)
15~3 c3
21
An analytical technique to compute this integral is considered in the homework.
MATH 235, by T. Lakoba, University of Vermont 136

It is interesting to point out that the constant in the Stefan–Boltzmann law is expressed via
three physical fundamental constants: the Boltzmann constant k, the Planck constant ~, and
the speed of light c.
To conclude this Section, we will point out a corollary of Planck’s formula which is encoun-
tered in everyday life. Recall that Kλ (T )dλ is the energy emitted by a black body into the
wavelength interval dλ. The human eye, however, reacts not to the energy but to the number of
photons which the eye detects per unit time. The number of photons (per wavelength interval)
is related to the radiation energy density in a simple manner:
( )
density of
radiation energy Kλ (T ) const · λ−4
Nλ (T ) = ( ) ≡ = hc/(λkT ) . (13.40)
energy of hc/λ (13.35), (13.6) e −1
one photon (13.28), (13.10)

Using the y-substitution equivalent to the one we used before Eq. (13.38), we have:
( hc )4 ( kT )4 ( 4 )
const · λkT · hc ′ y
Nλ (T ) = = const · T ·
4
, (13.41)
e hc/(λkT ) −1 e −1
y

where const′ =const·(k/(hc))4 .

Equation (13.41) and the figure on the left

show that the maximum number of radiated
photons occurs for the value of y which max-
imizes the last fraction in (13.41), whence
( )
ch
λmax ≈ /T. (13.42)
3.9k

This is called the Wien’s displacement

law, which Whilhelm Wien derived from his
formula (13.4) in 1894.

This law says that the wavelength of the maximal observed emittance decreases (“is displaced”)
in inverse proportion to the temperature. Thus, as the body is heated, it glows first in infrared
(very long wavelengths, over 1 µm), then in red, then in blue, and then in ultraviolet. For
example, the approximate temperatures of (the surfaces of the) three well-known stars are:
Betelgeuse — 3400o K, our Sun — 5800o K, and Sirius — 9500o K. Betelgeuse and Sirius are
known as the Red and Blue stars, respectively, while our Sun’s color is yellow. You will be
asked to use Eq. (13.42) to verify these color assignments in a homework problem. Moreover,
four cautionary notes regarding a correspondence between star temperatures and their observed
radiation spectrum are in order.
First, while star temperatures can be approximately deduced from Wien’s law, they are not
deduced from the location of the peak of the spectral density (i.e., from (13.41) or (13.42)). A
detailed discussion of this can be found in [D. Cenadelli, M. Potenza, M. Zeni, “Stellar temper-
atures by Wiens law: Not so simple,” Am. J. Phys. 80, 391 (2012); doi: 10.1119/1.3699958].
There are also other methods to measure a star’s temperature, e.g., by measuring the strengths
of absorption lines corresponding to various atoms (e.g., hydrogen) in the star’s radiation spec-
trum.
MATH 235, by T. Lakoba, University of Vermont 137

Second, radiation originates at various layers of the star, and the temperature of each layer
usually depends on the depth of the layer beneath the surface. Thus, the radiation spectrum of
a star is a superposition of several Planckian functions (13.41) each corresponding to a different
temperature. If these temperatures are not too far apart, the overall spectrum can be thought
of a Planckian function corresponding to some average temperature.
Third, even if the star emits a Planckian spectrum, the observer on Earth will detect it
with distortions, due to the fact that different wavelengths are absorbed and scattered by the
atmosphere differently. Generally, shorter wavelengths are absorbed or scattered more, which
shifts the peak of the observed spectral density towards longer wavelengths.
The above three reasons, and how they may affect the observed radiation spectrum, are
relatively straightforward to understand, at least conceptually. Our last note will be about an
issue whose resolution is not known to this instructor. Notice that if one expresses the photon
density (13.41) in terms of frequency rather than wavelength, one obtains from (13.40) and
(13.12) a different function than (13.41):

Nω (T ) ∝ ω 2 /(e~ω/(kT ) − 1) ∝ y 2 /(ey − 1).

If we maximize this function instead of (13.41), that we would have obtained a different nu-
merical coefficient in Wien’s law: instead of “3.9” in (13.42), one would obtain the maximum
at y ≈ 1.6. This appears to imply that the star colors depend on relative to which variable,
wavelength or frequency, we maximize the number of photons. This, of course, cannot be true.
While I am not aware of a concrete resolution of this paradox, I suspect that it should be sought
in a detailed mechanism of how the eye perceives different colors. Some references that may
contain relevant information are listed on the course website.

13.6 Planck’s formula and Quantum Electrodynamics

We conclude this Lecture with a curious historical fact. In 1913, Planck published the second
edition of his Lectures on the Theory of Thermal Radiation (see Section 13.1.3), where he gave
a new derivation of his formula. Conceptually, that derivation was similar to the one presented
in Section 4, but he made a modification that slightly changed the final result. Namely, back
in 1900, he assumed that the radiation’s energy could be subdivided into discrete values (which
later turned out to be the true quanta of light): 0, ε = hν, 2ε, 3ε, etc. In 1913, he said that the
energy of the mode could lie in the intervals [0, ε), [ε, 2ε), [2ε, 3ε), etc, and that the equilibrium
state of the radiation was reached by transitions among those intervals. Note that this is
almost like having no discreteness at all! With this modification, Planck’s earlier derivation
goes through with only a little change: the average energy of the radiation in the interval [0, ε)
is 21 ε; in [ε, 2ε) – ε + 12 ε; in [2ε, 3ε) – 2ε + 12 ε, etc. All this does22 is shift the average energy
of the radiation mode by 12 ε:
~ω 1
⟨E⟩ = ~ω/(kT ) + ~ω . (13.43)
e −1 2
However small this added term on the r.h.s. is, it turned out to have observable implications in
the theory of specific heats of solids. And, remarkably, it was detected by experiments of that
time that the 12 ~ω-term indeed needed to be there!
22
You will show this in a homework problem.
MATH 235, by T. Lakoba, University of Vermont 138

Equation (13.43) has another highly nontrivial implication. The average number of photons
(i.e. the quanta of light) per mode is:

average energy per mode ⟨E⟩

nω = =
energy of one photon ~ω
1 1
= ~ω/(kT ) + . (13.44)
e −1 2
In absolute vacuum, the temperature is zero (since there are no particles in vacuum, there is
no thermal motion and hence, by definition, the temperature must vanish). Then, since

lim e~ω/(kT ) = e∞ = ∞
T →0

for all ω ̸= 0, then in (13.44),

1 1
nω |T =0 = 0 + = . (13.45)
2 2
This says that in vacuum, there is, on average, 21 photon for each possible frequency of radiation,
and hence vacuum is not really empty! This result was proved several decades later in Quantum
Electrodynamics. Thus, Planck’s incorrect derivation of (13.43) predicted a correct result.
 Blackbody radiation and Plank’s law
“blackbody” problem:
calculating the intensity of radiation at a given wavelength emitted by a body at a specific temperature

Max Planck, 1900

quantization of energy of radiation-emitting oscillators:
only certain energies for the radiation-emitting oscillators
in the cavity wall are allowed

Albert Einstein, 1905 Niels Bohr, 1913

extended quantization of energy quantum model of the atom
of radiation-emitting oscillators
to quantization of light → photons

explained the photoelectric effect

 Blackbody radiation and Plank’s law
blackbody is an object that absorbs all electromagnetic radiation falling on it
an consequently appears black
the opening to the cavity is a good approximation of a blackbody:
after many reflections all of the incident energy is absorbed

radiation is in thermal equilibrium with the cavity (e.g. oven cavity)

because radiation has exchanged energy with the walls many times
similar to the thermal equilibrium
of a fluid with a container
spectral energy density u( f , T )
is energy per unit volume per unit frequency of the radiation within the blackbody cavity

u( f , T ) depends only on temperature and light frequency

and not on the physical and chemical makeup of the blackbody

all the objects in the oven, regardless of their chemical nature, size, or shape,
emit light of the same color

blackbody is a perfect absorber and ideal radiator

 towards “ultraviolet”
Max Planck, 1900
catastrophe
spectral energy density of blackbody radiation
classical
Rayleigh-Jeans law

radiatedd intensity
8π hf 3 1
u( f , T ) = quantum
c 3 e hf kBT − 1 Planck law

h = 6.626 × 10−34 J ⋅ s is Planck's constant

k B = 1.380 × 10−23 J/K is Boltzmann's constant
f

limiting behaviors
at high frequencies hf k BT >> 1 at low frequencies hf k BT << 1
1 1 1 k T
≈ e − hf k BT
= ≈ B
e hf k BT
−1 e hf k BT
−1 1 + hf k BT + ... − 1 hf
8π hf 3 1 8π hf 3 − hf 8π hf 3 1 8π hf 3 k BT 8π f 2
u( f , T ) = ≈ e k BT
u( f , T ) = ≈ = 3 k BT
c 3 e hf kBT − 1 c3 c 3 e hf kBT − 1 c 3 hf c

Wien’s exponential law, 1893 classical Rayleigh-Jeans law

Rayleigh-Jeans law is the classical limit obtained when h→0

Max Planck:
blackbody radiation is produced by vibrating submicroscopic electric charges,
which he called resonators
the walls of a cavity are composed of resonators vibrating at different frequency

Classical Maxwell theory:

An oscillator of frequency f could have any value of energy and could change its amplitude
continuously by radiating any fraction of its energy

Planck: the total energy of a resonator with frequency f could only be an integer multiple of hf.
(During emission or absorption of light) resonator can change its energy only by
the quantum of energy ∆E=hf
Ε Planck: allowed energy levels of a resonator
E = nhf where n = 1, 2,3,... 4hf
∆E = hff 3hf
photon
h t absorption
b ti
2hf
photon emission
hf
0
all systems vibrating with frequency f are quantized
and lose or gain energy in discrete packets or quanta ∆E=hf

Consider a pendulum
m=0.1 kg, l=1 m, displaced by θ=100
g ((1 − cos θ ) = ( 0.1 kgg ) ( 9.8 m/s2 ) (1 m ) (1 − cos100 ) = 1.5 × 10−2 J
E = mgl

1 g 1 9.8 m/s2
f = = = 0.5 Hz
2π l 2π 1m
∆E = hf = ( 6.63 × 10−34 J ⋅ s )( 0.5 s −1 ) = 3.3 × 10−34 J quantum of energy
∆E 3.3 × 10−34 J
= −2
= 2.2 × 10−32 classical physics:
E 1.5 × 10 J
quantization
ti ti isi unobservable
b bl andd
energy loss or gain looks continuum
∆E is large ← f is large
quantum
qua tu physics
p ys cs
E is
i smallll ← m is
i smallll
energy change of atomic oscillator sending out green light

∆E = hf =
hc
=
( 6.63 × 10 −34
J ⋅ s )( 3 × 108 m/s )
3 68 × 10−19 J = 2.3
= 3.68 2 3 eV
λ 540 × 10−9 m
−19
a more appropriate unit of energy for describing atomic processes 1 eV = 1.602 × 10 J
 average energy
emitted per oscillator towards “ultraviolet”
Max Planck: number of catastrophe
oscillators with classical
u( f , T )df = EN ( f )df frequency
Rayleigh-Jeans law

radiated intensity
between f and f+df
8π f 2 quantum
N ( f )df = 3 df Planck
a c law
aw
c
1 Plank
E = hf hf kBT distribution function
e −1
f
8π f 2 hf
u( f , T )df = 3 hf kBT df in classical Rayleigh-Jeans theory E = k BT
c e −1 8π f 2
u( f , T )df = 3 k BTdf
8π hc 1 c
u(λ , T )d λ = 5 hc λ kBT dλ for f→∞ classical theory predicts unlimited
λ e −1 energy emission in the ultraviolet region
c
f = “ultraviolet
ultraviolet catastrophe
catastrophe”
λ
in quantum theory “ultraviolet catastrophe” is avoided:
E tents to zero at high f because the first allowed energy level E=hf is so large for large f
comparedd to
t the
th average thermal
th l energy available
il bl kBT
that occupation of the first excited state is negligibly small
spectral energy density u( f , T ) c power density e f = J ( f , T )
is energy per unit volume per unit frequency J ( f , T ) = u ( f , T ) is power emitted per unit
of the radiation within the blackbody cavity 4 area per unit frequency
8π hc 1
u (λ , T ) = hc λ k BT
Wien’s displacement law, 1893:
λ 5
e −1 the wavelength marking the maximum power
emission of a blackbody, λmax, shifts towards
shorter wavelengths with increasing temperature
λmax ~ T −1
λmaxT = 2.898 × 10−3 m ⋅ K

Stefan-Boltzmann law, 1879:

The total ppower pper unit area emitted at all frequencies
q by
y a blackbody,
y, etotal,
is proportional to the forth power of its temperature
∝
etotal = ∫ e f df = σ T 4
0

σ = 5.67
6 × 10−8 W ⋅ m −2 ⋅ K −4 ← The
h Stefan-Boltzmann
f l constant

c ∝ ∝ 2π hc 2 2π k B 4T 4 ∝ x3 2π 5k B 4 4
etotal = ∫ u(λ , T )d λ = ∫ 5 hc λ k T dλ = ∫ dx = T = σT 4
4 0 0 λ
(e B
− 1) c 2h 3 0 e −1
x 2 3
15c h
x = hc λ k BT = π 4 15
 c
for perpendicular derivation of J ( f ,T ) = u( f , T )
radiated energy 4

radiated power

δx=cδt area A
because half the power will be going in the –x direction
1 volume
radiated power = [ energy density ] × =
2 time
1 A⋅δ x 1 A⋅δ x c
= u( f , T ) = u( f , T ) = u( f , T ) A
2 δt 2 δx c 2

for any angle 1 volume

θ
perpendicular radiated power = [energy density] × × cos θ =
2 time
1 A⋅δ x 1 A⋅δ x c
= u( f , T ) cos θ = u( f , T ) cos θ = u ( f , T ) A cos 2
θ
2 δt 2 δ x ⋅ cos θ c
−1
2

c
radiated power u ( f , T ) A cos 2
θ c
averaging over θ J ( f ,T ) = = 2 = u( f , T )
and dividing by A A 4
A to get cos2 θ = 1 2
power density
Estimate the surface temperature of the Sun and find λmax for the Sun emission.
The Sun radius RS=7.0x108 m
The Earth-to-Sun distance R=1.5x1011 m
The total power from the Sun at the Earth etotal=1400 W/m2
etotal ( RS ) = σ T 4 ← Stefan-Boltzmann law
t t l ( RS ) ⋅ 4π RS = etotal
t t l ( R ) ⋅ 4π R
2 2
etotal ← conservation of energy
14 14
⎛1 ⎞ ⎛ 1 etotal ( R ) ⋅ R 2 ⎞
T = ⎜ ⋅ etotal ( RS ) ⎟ =⎜ ⋅ ⎟
⎝σ σ 2
⎠ ⎝ R S ⎠
14
⎛ ( ) ( ) ⎞
2
1 1400 W/m 2
⋅ 1.5 × 1011
m
T =⎜ ⋅ ⎟ = 5800 K
( m)
−8
⎜ 5.67 × 10 W/m ⋅ K2 4
7.0 × 10 8 2
⎟
⎝ ⎠
2.898 × 10−3 m ⋅ K 2.898 × 10−3 m ⋅ K eye’s
λmax = = = 500 × 10−9 m = 500 nm sensitivity
T 5800 K peak

Wien’s displacement law

 Planck, 1900: oscillators in the walls of the blackbody are quantized

Einstein, 1905: light itself is composed of quanta of energy

photon – a quantum of electromagnetic radiation (a quantum of light)

E 1 photon
x E=hf
x
E 2 photons
E=2hf

B
3 photons
E=3hf

Ephoton = hf

the photoelectric effect

 Einstein, 1906:
in addition to carrying energy E=hf
a photon carries a momentum p=E/c=hf/c directed along its line of motion

Peter Debye, Arthur Holly Compton, 1923:

scattering of x-rays photons from electrons could be explained by treating photons as
particles with energy hf and momentum hf/c and
by conserving energy and momentum of the photon-electron pair in a collision

pparticle properties
p p of light
g

x-rays – are electromagnetic

waves with short wavelengths
x-rays were discovered by
Wilhelm Roentgen in 1895:
Frequency (f)
X-rays are generated
X t d when
h
high-speed electrons strike a metal
target and give up some of their
energy when they interact with the
hc hc
h t energy hf =
photon ≈ ≈ 12 keV
k V orbital
bi l electrons
l off an atom
λ −10
10 m