Laser in Physics, Chemistry, Biology and Medicine

the inducing photon. The probability 𝑃'% that a molecule is induced to emit a photon per second is
analogous to (2.5):

𝑃'% = 𝐵'% ∙ 𝜌(𝜈) (2.6)

𝐵'% is the Einstein coefficient of the induced emission. An excited molecule can also use its excitation
energy spontaneously, i.e. emit by light emission without an external, inducing field. The spontaneously
emitted photon can increase the occupation number of any mode with a suitable frequency 𝜈 = (𝐸' −
𝐸% )/ℎ but any direction of the 𝑘d⃗ -vector by one. In the case of isotropic emission, the probability for all
these modes is the same.

The probability 𝑃'% (spontaneous) per second that a photon ℎ is spontaneously emitted by an excited
molecule is independent of the external field and depends only on the structure of the molecule and the
transition 𝐸' → 𝐸% under consideration:

𝑃'% (spontaneous) = 𝐴'% (2.7)

𝐴'% is called the Einstein coefficient of spontaneous emission or spontaneous transition probability for the
transition |2⟩ → |1⟩.

In thermal equilibrium, the Boltzmann distribution applies to the occupation numbers 𝑁/ of the energy
level 𝐸/ (without light field)

𝑔/ 𝑁 $ 1!
𝑁/ = Ž • 𝑒 >? (2.9)
𝑍

where 𝑔/ is the statistical weight of the state 𝐸/ (for a state 𝐸/ of a free atom with angular momentum 𝐽/ is
𝑔/ = 2𝐽/ + 1), 𝑁 is the total number of molecules per volume and 𝑍 is the partition function 𝑍 =
∑C@D& 𝑒
$@A;/>?
. For the ratio 𝑁' /𝑁% , one obtains from (2.9):

𝑁' 𝑔% $(1 $1 )/>? 𝑔% $A;/>?

= 𝑒 " # = 𝑒 (2.10)
𝑁% 𝑔' 𝑔'

Figure 2.3 Definition of absorbed intensity and the spectral absorption profile upon incidence of spectrally continuous
radiation
This means that with the same statistical weights of the two states 𝐸% and 𝐸' , the Einstein coefficients for
induced emission and absorption are the same!

2.1.3 Beer’s law

Usually, the absorption of light when passing through matter is not described as a decrease in amplitude
but as a decrease in intensity, since only intensities - but not amplitudes - can be measured directly. If a

14
Laser in Physics, Chemistry, Biology and Medicine

plane wave with the intensity 𝐼(𝜔) runs in the z-direction through a homogeneous medium (Fig. 2.3), the
intensity is reduced by:

𝑑𝐼 = −𝛼 ∙ 𝐼 ∙ 𝑑𝑧 (2.14)

on the distance 𝑑𝑧. The absorption coefficient 𝛼[𝑐𝑚$% ] indicates the fraction 𝑑𝐼/𝐼 absorbed on the
route 𝑑𝑧 = 1𝑐𝑚. If 𝛼 is independent of 𝐼 (linear absorption), the integration of (2.14) gives Beer's law of
absorption with 𝐼(𝑧 = 0) = 𝐼& :

𝐼 = 𝐼& 𝑒 $G9 (2.15)

The absorption coefficient 𝛼 is in many cases proportional to the imaginary part of the refractive index (𝑛H
real part, 𝜅 imaginary part) and has a frequency dependence in the vicinity of the oscillator natural
frequency for molecules at rest, which is described by a Lorentz profile for 𝑛 − 1 ≪ 1, as generally in the
case of resonances.

𝑁𝑞' 𝛾𝜔'
𝜅=
2𝜀& 𝑚 (𝜔&' − 𝜔 ' )' + 𝛾 ' 𝜔 '

𝑁𝑞' 𝜔&' − 𝜔'

𝑛H = 1 + (2.16)
2𝜀& 𝑚 (𝜔&' − 𝜔 ' )' + 𝛾 ' 𝜔 '

The equations (2.16) are also called dispersion relations.

In the neighborhood of the natural frequencies the equations simplify to:

𝑁𝑞' 𝛾
𝜅=
2𝜀& 𝑚𝜔& (𝜔& − 𝜔) + (𝛾/2)'
'

𝑁𝑞' 𝜔& − 𝜔
𝑛 =1+
2𝜀& 𝑚𝜔& (𝜔& − 𝜔)' + (𝛾/2)'

2.1.4 Interaction between radiation and molecular system

As we have already established, in laser spectroscopy we want to observe a reaction of the atomic or
molecular system as a result of the radiation of electromagnetic radiation - and that as a function of the
wavelength of this radiation.

The starting point for most of the following considerations is the simplified two (three) level system for the
quantum mechanical description of the atom (molecule) (Figure 2.5). A laser radiation (electromagnetic
radiation) 𝐸d⃗ falls on an atom which is described by the energy level diagram shown in Figure 2.5. For the
sake of simplicity, we will limit ourselves to monochromatic radiation of form 𝐸 = 𝐴& cos(𝜔𝑡 − 𝑘𝑧). In
addition, the frequency 𝜔 of the radiation should be close to the resonance frequency 𝜔IJ of the
transition 𝑎 → 𝑏.

15
Laser in Physics, Chemistry, Biology and Medicine

Figure 2.5 Scheme of the interaction of electromagnetic radiation of frequency 𝜔 and wave number 𝑘 with a
molecular system. The radiation only causes changes in the system with regard to the energy levels 𝐸& and 𝐸' , since
the irradiated frequency is so narrow-band that the effective cross-sections for all other possible transitions are
practically zero. The damping constant 𝛾' describes all possible relaxation processes, i.e. transitions from level 𝐸' to
lower levels that are not triggered by radiation 𝜔 (e.g. spontaneous emission, non-radiative transitions). An analogous
constant 𝛾& must be defined if 𝐸& is not the ground state.

In the following, we want to restrict ourselves to the case of the so-called dipole approximation (𝐸 is
constant in space). As is known from electrodynamics, one can determine the force that a field exerts on a
charge, provided that the field does not vary greatly in the area of interest (in our case we require that the
phase of the field does not vary greatly across the atom changes, i.e. 1 >> dimensions of the atom),
developed into a Taylor series. The corresponding potential is also obtained through integration as a series
development. If we now calculate the (interacting) energy of the field with the atom consisting of several
charges, we get a multipole expansion for the energy. The contribution we are interested in here to the
energy, which is caused by the external 𝐸-field, is the dipole interaction of the form:

−𝑑⃗ ∙ 𝐸d⃗ = −𝑑⃗ ∙ dddd⃗

𝐴& cos(𝜔𝑡) = −𝑑⃗ ∙ dddd⃗
𝐴& p𝑒 /=4 + 𝑒 $/=4 q (2.17)
':
Where we have already taken into account that is 𝑘𝑧 = K
≪ 1.

Explicitly, between two atomic states 𝑚, 𝑛 we can write 𝑑⃗ as follows:

𝑑⃗ = ™ 𝜌(𝑟⃗)𝑟⃗𝑑+ 𝑟 = ™ −𝑒𝜓. (𝑟⃗)∗ 𝑟⃗𝜓" (𝑟⃗)𝑑+ 𝑟 = −𝑒⟨𝑛|𝑟⃗|𝑚⟩ (2.18)

2.2 Coherent and incoherent excitation

2.2.1 Rate equations (incoherent excitation)

When treating the excitation with rate equations, one starts from the simplifying assumption that the
change in population is proportional to the product of an effective cross section and the current population
of the state. The effective cross-sections can depend on the radiation intensity or can also be pure
relaxation constants. (See equations 2.11, 2.12).

The rate equations describe the incoherent transition of the system from Boltzmann's thermodynamic
equilibrium to the time-independent state of equilibrium under the action of radiation. It is characteristic
that even in a system without relaxations, the upper and lower states can at most be equally occupied. The
incoherent case corresponds to the classic superposition of the measurable quantity, i.e. here the

16
Laser in Physics, Chemistry, Biology and Medicine

population of the different energy levels (with electrons). Possible interferences of the states (wave
functions) are neglected from the start.

Neglecting transient processes (i.e. 𝑡 ≫ 𝜏), the probability of the induced excitation (and also emission) 𝑃IJ
is a product of the laser intensity and a constant value (the Einstein coefficient 𝐵IJ ). The probability of the
spontaneous emission 𝛾 = 1/𝜏 is proportional to the Einstein coefficient 𝐴IJ . As already discussed, the two
Einstein coefficients are related to one another via the equation:

𝑐+
𝐵IJ = 𝐴IJ (2.19)
8𝜋ℎ𝜈3

With these approximations one can then describe the change in the occupation of the individual states by
simple rate equations, whereby we will initially be particularly interested in the stationary case (𝑡 → ∞). For
𝑖 atomic states we get 𝑖 rate equations for the occupation number 𝑛. (𝑡) of the individual states of the
form:
/ /
𝑑
𝑛 (𝑡) = Ÿ 𝑛" (𝑡)𝑅". (𝑡) − 𝑛. (𝑡) Ÿ 𝑅." (𝑡) (2.20)
𝑑𝑡 .
"M. "M.

where 𝑅". (𝑡) is the transition probability from state 𝑚 to state 𝑛. This can itself consist of several
contributions. For our considerations we can restrict ourselves to the following contributions to 𝑅". :

• Induced emission and absorption: proportional to

𝐵". × Laser intensity × Absorption cross section (optional)

• Spontaneous emission: proportional to

/
1
𝛤. = = Ÿ 𝐴." (2.21)
𝜏
"[1$ O1% ]

The term 𝐴." describe the spontaneous emission between the individual states.

• Cross excitation: in some cases, crosses can act as damping. You can treat this like an additional
contribution to 𝛾". .

Solution of the rate equations

The rate equations 2.20 corresponding to a special case can be written down very easily if the Einstein
coefficients and the available intensities (or spectral densities) are known. This applies to the transition
between discrete energy levels, but can also be used for transitions into the continuum (ionization).

Equations 2.20 result in a differential equation system that can be solved analytically in very simple cases.
However, it is very easy to solve numerically even in relatively complex situations. Very convenient
methods are available today for this purpose.

E.g. Mathematica provides a good way to solve the rate equations. In certain cases an analytical solution
can be obtained with it. In most of the more complex cases, it is probably more convenient to do a
numerical integration of the equations. This is not only possible with Mathematica, but also with a variety

17
Laser in Physics, Chemistry, Biology and Medicine

of the so-called simulation languages. As an example we consider a four-level system, where we can choose
the individual Einstein coefficients. A two, three or four level system can thus be represented.

For the numerical value of the Einstein coefficients, one should note that in the case shown, the product
𝐵/> ∙ 𝜌 of the coefficients for the induced transitions is actually changed.

The following is a simple Mathematica notebook that simulates a four-level system (Figure 2.6). By
choosing different numerical values for the Einstein coefficients, a two-, three- or four-level system can be
simulated. In addition, the length of the pump pulse and 𝑄-switching can be selected.

Figure 2.6 Definition of the energy levels in a four-level system.

2.2.2 Time-dependent Schrödinger equation (coherent excitation)

An exact description of the excitation process with a coherent light source requires the use of the
Schrödinger equation and thus the calculation of the probability amplitudes of the states in which the
system is at time 𝑡. We can then calculate the occupation of the individual states (as obtained from the rate
equations) from the square of the probability amplitudes. This is the only way to ensure that we do not
neglect coherent phenomena (interference).

The time-dependent Schrödinger equation is now:

𝜕𝜓
𝐻𝜓 = −𝑖ħ (2.22)
𝜕𝑡

Where we assume that the system first exists without an external field up to time 𝑡 = 0, while it is
described by the eigenvalues 𝐸. and the eigenfunctions 𝜑. (𝑟⃗, 𝑡) up to this point in time:

𝜑. (𝑟⃗, 𝑡) = 𝜓. (𝑟⃗)𝑒 $/1% 4/ħ = 𝜓. (𝑟⃗)𝑒 $/=% 4 (2.23)

At time 𝑡 = 0, the external field is switched on. Since the eigenstates of the atom are not (substantially)
changed by the external field, the state of the system at time 𝑡 can be represented by the superposition of
the unperturbed eigenfunctions, the coefficients 𝑐. (𝑡) being dependent on time.
C

𝜓. (𝑟⃗, 𝑡) = Ÿ 𝑐. (𝑡)𝜓. (𝑟⃗)𝑒 $/=% 4 (2.24)

.D%

This sum is converted into the following expression for our two-level system (discussed in detail here):

18
Laser in Physics, Chemistry, Biology and Medicine

𝜓(𝑟⃗, 𝑡) = 𝑎(𝑡)𝜓I (𝑟⃗)𝑒 $/=& 4 + 𝑏(𝑡)𝜓J (𝑟⃗)𝑒 $/=' 4 (2.25)

The further calculation results from inserting into the Schrödinger equation (2.22):

RSJ#4/4S4T 𝜕
→ 𝐻ª𝑎(𝑡)𝜓I (𝑟⃗)𝑒 $/=& 4 + 𝑏(𝑡)𝜓J (𝑟⃗)𝑒 $/=' 4 « = −𝑖ħ p𝑎(𝑡)𝜓I (𝑟⃗)𝑒 $/=& 4 + 𝑏(𝑡)𝜓J (𝑟⃗)𝑒 $/=' 4 q
𝜕𝑡

After some transformations we obtain:

𝑑⃗ ⦁ 𝐴⃗& p𝑒 /=4 + 𝑒 $/=4 q² ª𝑎(𝑡)𝜓I (𝑟⃗)𝑒 $/=& 4 + 𝑏(𝑡)𝜓J (𝑟⃗)𝑒 $/=' 4 « =

¬𝐻U + ®¯¯¯¯¯°¯¯¯¯¯±
V
)

𝜕 𝜕
= −𝑖ħp𝑎̇ (𝑡)𝜓I (𝑟⃗)𝑒 $/=& 4 + 𝑏̇ (𝑡)𝜓J (𝑟⃗)𝑒 $/=' 4 q − 𝑖ħ Ž𝑎(𝑡) ª𝜓I (𝑟⃗)𝑒 $/=& 4 « + 𝑏(𝑡) ª𝜓J (𝑟⃗)𝑒 $/=' 4 «•
𝜕𝑡 𝜕𝑡
W/4A Y( ZD1! Z
→ − 𝑖ħp𝑎̇ (𝑡)𝜓I (𝑟⃗)𝑒 $/=& 4 + 𝑏̇(𝑡)𝜓J (𝑟⃗)𝑒 $/=' 4 q = 𝑎(𝑡)𝑉´𝜓I (𝑟⃗)𝑒 $/=& 4 + 𝑏(𝑡)𝑉´𝜓J (𝑟⃗)𝑒 $/=' 4

⦁\% (H⃗)〈.DI,J〉`a.4TbHI4/U.
→

𝑖
𝑎̇ (𝑡) = − ª𝑎(𝑡) ∙ 𝑉II + 𝑏(𝑡) ∙ 𝑉IJ ∙ 𝑒 /(=& $=')4 «
ħ
𝑖
𝑏̇(𝑡) = − ª𝑏(𝑡) ∙ 𝑉JJ + 𝑎(𝑡) ∙ 𝑉JI ∙ 𝑒 $/(=& $=' )4 «
ħ

Taking into account the dipole matrix element 𝑅/> , which causes 𝑉// = 0 and with the following definitions:

𝑉/> = ™ 𝜓/∗ ∙ 𝑉@ ∙ 𝜓> ∙ 𝑑+ 𝑟 = −𝑒 ∙ 𝐸d⃗ ™ 𝜓/∗ ∙ 𝑟⃗ ∙ 𝜓 ∙ 𝑑+ = −𝑒 ∙ 𝐸d⃗ ∙ 𝑅/> (2.26)

𝐸J − 𝐸I
𝜔JI = (2.27)
ħ
𝑅/> 𝑒𝐴&
ℜ/> = 𝑒 ∙ 𝐴& ∙ = ™ 𝜓I∗ ∙ 𝑟⃗ ∙ 𝜓J ∙ 𝑑𝜏 (2.28)
ħ ħ

we get the following differential equation system (please note: 𝑉´ = 𝑑⃗ ⦁ 𝐴⃗& p𝑒 /=4 + 𝑒 $/=4 q; the time factor
in equations 2.26, 2.28 not yet taken into account):

𝑖
𝑎̇ (𝑡) = ℜIJ ª𝑒 $/(='& $=)4 + 𝑒 $/(='& `=)4 « ∙ 𝑏(𝑡)
2
𝑖
𝑏̇(𝑡) = ℜIJ ª𝑒 /(='& $=)4 + 𝑒 /(='& `=)4 « ∙ 𝑎(𝑡)
2

Furthermore one can show that the terms with 𝑒 ±/(='& `=)4 vanish near the resonance (rotating wave
approximation). The differential equation system then simplifies to 2.29.

Furthermore, relaxation processes are not yet taken into account in the above equations, but can be taken
into account by inserting phenomenological damping terms, which ultimately results in the final form of the
differential equation system for the time-dependent probability amplitudes (or actually the time-
dependent weighting factors of the eigenfunctions):

19
Laser in Physics, Chemistry, Biology and Medicine

𝑖 1 1
𝑎̇ (𝑡) = ℜIJ ª𝑒 $/(='& $=)4 « ∙ 𝑏(𝑡) + 𝛾J 𝑏(𝑡) − 𝛾I 𝑎(𝑡)
2 2 2
𝑖 1 1
𝑏̇(𝑡) = ℜIJ ª𝑒 `/(='& $=)4 « ∙ 𝑎(𝑡) − 𝛾J 𝑏(𝑡) + 𝛾I 𝑎(𝑡) (2.29)
2 2 2

For the case without damping, (2.32) can be solved analytically, which leads to the known Rabbi frequency.
For |𝑏(𝑡)|' one obtains the solution given by Rabbi (equations 2.30 to 2.32).
d 𝜇 '
ℜ'IJ 𝑒 $'4 ¶sin 2 𝑡 ¶
𝑃IJ = |𝑏(𝑡)|' = (2.30)
𝛾 '
(𝜔JI − 𝜔)' + j m + ℜ'IJ
2

𝛾'
𝜇 = =~𝜔JI − 𝜔 + 𝑖 € + ℜ'IJ (2.31)
2

𝑒𝐴&
ℜIJ = ™ 𝜓I∗ ∙ 𝑟⃗ ∙ 𝜓J ∙ 𝑑𝜏 (2.32)
ħ

As can be seen from the solution function by forming the limit, the population (∝ |𝑏(𝑡)|' ) of the excited
state increases linearly with 𝑡 ' for small intensities (for the case of resonance). With this we can easily
establish the connection with the Einstein coefficients introduced earlier.

For the (realistic) case, where the attenuation (in the next example e.g. the spontaneous emission from
level 𝑏 to level 𝑎) is taken into account, the following time behaviour of |𝑎(𝑡)|' and |𝑏(𝑡)|' results. It
should be noted that the coherent solution on the one hand results in an occupation of the individual
coherently excited states in a certain time range, which oscillates between a maximum value and a
minimum value, but on the other hand heads towards the incoherent equilibrium state for long periods of
time.

Figure 2.8. Temporal behavior for the occupation of the ground and excited state for small times and small intensities.

20
Laser in Physics, Chemistry, Biology and Medicine

Figure 2.9. Temporal behavior for the occupation of the ground and excited state for greater intensities.

We will come back to these sizes several times. At this point we want to summarize the following:

1. The material-dependent dipole matrix element essentially determines the strength of the
interaction between the laser radiation and the medium (atoms-molecules-solids, etc.).
2. The response of the system to the laser radiation depends according to a resonance curve
[equation 2.30] (Lorentz curve) on the frequency difference between the exciting radiation and the
resonance frequency of the system. Through this interaction, the system is brought in the most
general form from a state of low energy to a higher energy (excitation, ionization, etc.). The further
development of the system can take place through a wide variety of energy channels and thus
cause a wide variety of physical effects.

The properties of the laser radiation also allow processes with very small dipole matrix elements on the one
hand, and the existing power can be optimally adapted to the resonance of the system on the other.

2.2.3 Transition probabilities

Taking into account the following relationships between intensity 𝐼 of the laser radiation and the
corresponding spectral density 𝜌(𝜔):

𝐼 = 𝑐 ™ 𝜌(𝜔)𝑑𝜔 = 𝑐𝜀& 𝐸 ' = 𝑐𝜀& 𝐸& cos ' (𝜔𝑡 − 𝑘𝑧)

𝐸& @@@@@@@@@@@@@@@@@@@
𝐴& = → 𝐼 ̅ = 𝑐𝜀& 𝐸&' cos ' (𝜔𝑡 − 𝑘𝑧) = 2𝑐𝜀 𝐴'
& &
2

𝐴'&
𝜌(𝜔) = 𝜀& (2.33)
2

we get for the transition probability 𝑃IJ (𝑡) (under the simplifying assumption that 𝐼 is small and can
therefore be neglected in 𝑚):

𝑡
𝑒𝑅 '
IJ
sin ~(𝜔JI − 𝜔) 2€
𝑃IJ (𝑡) = ™|𝑏(𝑡)|' 𝑑𝜔 = ™ 𝜌𝜔 ¹ 𝜔JI − 𝜔 º 𝑑𝜔 (2.34)
2𝜀& ħ'
2

21
Laser in Physics, Chemistry, Biology and Medicine

The integral can be carried out under the assumption that 𝜌(𝜔) is constant in the region of resonance, as is
the case for thermal, white light sources, and results in the value 𝜌(𝜔IJ ) ∙ 2𝜋𝑡. We thus obtain for the
ef&'
transition probability 𝑃IJ or the transition probability per unit of time e4
:

'
𝜋𝑒 ' 𝑅IJ
𝑃IJ (𝑡) = 𝜌(𝜔IJ ) ∙ 𝑡
𝜀& ħ'
'
𝑑𝑃IJ 𝜋𝑒 ' 𝑅IJ
= 𝜌(𝜔IJ ) = 𝜌(𝜔IJ ) ∙ 𝐵IJ (2.35)
𝑑𝑡 𝜀& ħ'

We also get the well-known relationship for the Einstein coefficient 𝐵IJ from equation 2.35:
'
𝜋𝑒 ' ∗ +
𝐵IJ = »™ 𝜓J 𝑟
⃗ 𝜓I 𝑑 𝑟» (2.36)
𝜀& ħ'

3. STRUCTURE AND REALIZATION OF A LASER –

ESSENTIAL PROPERTIES

Figure 3.1. Basic elements of every laser.

Each laser can be reduced to four basic elements (see Figure 3.1):

1. Laser medium - optical amplifier;

22