Fluorescence Anisotropy

Classical explanation of radiation

We have already discussed on the relation between 𝐸"⃗ and 𝐵
"⃗ given the form of 𝐸"⃗ is
monochromatic plane wave. Now we ask question of how does 𝐸"⃗ behave when we have the
information about the source. That is, how is 𝐸"⃗ generated?

Suppose we have moving charge that follows

𝜌(𝑥⃗, 𝑡) = 𝜌(𝑥⃗)𝑒 !"#$ (1.1)
⃗𝐽(𝑥⃗, 𝑡) = 𝐽⃗(𝑥⃗)𝑒 !"#$ (1.2)
where complex notation is used.
Such oscillatory behavior of the charge can be illustrated in the example showing in Figure 1.

Figure 1. Picture two tiny metal spheres separated

by a distance d and connected by a fine wire; at
time t the charge on the upper sphere is 𝑞(𝑡), and
the charge on the lower sphere is −𝑞(𝑡). Suppose
that we drive the charge back and forth through
the wire, from one end to the other, at an angular
frequency w. (that is, the charge on the metal
sphere oscillate over time)

The electric field from the source Eq.(1.1)-(1.2) (for 𝑥⃗ sufficiently away from the source)
% !"#$ % "&' + ",
𝐸"⃗ (𝑥⃗, 𝑡) = &'( .𝑘 ) (𝒏 × 𝝁) × 𝒏 * + [3𝒏(𝒏 ∙ 𝝁) − 𝝁] 9* ( − * ) : 𝑒 ",* ; (1.3)
%
-⃗
where 𝑟 = |𝑥⃗|; 𝒏 = |-⃗|
; and 𝝁 = ∫ 𝑥⃗ 0 𝜌(𝑥⃗ 0 )𝑑𝑥 0 1 .
At 𝑟 → ∞, the first term dominates and
% !"#$ % "&' )
𝐸"⃗ (𝑥⃗, 𝑡) ≈ &'(%
𝑘 (𝒏 × 𝝁) × 𝒏
*
(1.4)
It should be pointed out that 𝐸"⃗ is perpendicular to 𝒏.

*Note that
(𝒂 × 𝒃) × 𝒄 = (𝒂 ∙ 𝒄)𝒃 − (𝒂 ∙ 𝒃)𝒄
We have
(𝒏 × 𝝁) × 𝒏 = (𝒏 ∙ 𝒏)𝝁 − (𝒏 ∙ 𝝁)𝒏 = 𝝁 − cos (𝜃)|𝝁|𝒏
where 𝜃 is the angle between 𝝁 and 𝒏.
In case 𝒏 = 𝒊 (x-direction), then 𝝁 − cos(𝜃) |𝝁|𝒊 = 𝜇- 𝒊 + 𝜇2 𝒋 + 𝜇3 𝒌 − 𝜇- 𝒊 = 𝜇2 𝒋 + 𝜇3 𝒌.
Note that cos(𝜃) |𝝁| = 𝜇- .

Some textbook uses the spherical coordinate system where 𝒌 coincide with the direction of 𝝁; 𝒓P
is pointing in the direction of 𝒏. The lengths of |(𝒓P × 𝒌) × 𝒓P| is given by
|(𝒓P × 𝒌) × 𝒓P| = Q(𝒌 − cos(𝜃) 𝒏) ∙ (𝒌 − cos(𝜃) 𝒏) = Q1 − cos ) (𝜃) = sin (𝜃)
𝜃V is defined to be the unit vector
(𝒓P × 𝒌) × 𝒓P
𝜃V =
|(𝒓P × 𝒌) × 𝒓P|
or
(𝒓P × 𝒌) × 𝒓P = sin (𝜃)𝜃V
Using this coordinate system, one can express Eq.(1.21) as
4 |𝝁|# ) "(,*!#$) 89: (<)
𝐸"⃗ (𝑥⃗, 𝑡) ≈ % &'
𝑒 *
𝜃V (1.5)
This corresponds to the Lacowicz Eq.(10.10).

Similarly, one can calculate the magnetic field 𝐻"⃗ (𝑥⃗, t)

!"#$ = % "&'
"⃗ (𝑥⃗, 𝑡) = %
𝐻 𝑘 ) (𝒏 × 𝝁) 91 −
+
:
(1.6)
&' * ",*
The energy radiated by an oscillating electric dipole is determined by the Poynting vector 𝑆⃗(𝑥⃗, t),
|𝝁|# ) 89: (<) )
4
𝑆⃗(𝑥⃗, 𝑡) = 𝐸"⃗ × 𝐻
"⃗ ≈ % .
= &' *
𝑒 "(,*!#$) ; 𝒓P (1.7)
The intensity is obtained by averaging (in time) over a complete cycle
) * )
>
"⃗ ≈ 4% |𝝁| )# 89: )(<) 𝒓P
〈𝑆⃗(𝑥⃗)〉 = ∫ 𝑆⃗(𝑥⃗, 𝑡)𝑑𝑡 = 𝐸"⃗ × 𝐻 (1.8)
? 1)' = *
)'
𝑇 is the period #
.

Fun fact:
Eq.(1.8) provide us with intuitive ways of explaining why the sky is blue. Sunlight
passing through the atmosphere stimulates atoms to oscillate as tiny dipoles. The incident solar
radiation covers a broad range of frequencies (white light), but the energy absorbed and
reradiated by the atmospheric dipoles is stronger at the higher frequencies because of the 𝜔& in
Eq.(1.8). It is more intense in the blue, then, than the red. It is this reradiated light that you see
when you look up in the sky.
The redness of sunset is the other side of the same coin: Sunlight coming in at a tangent
to the earth’s surface must pass through a much longer stretch of atmosphere than sunlight
coming from overhead. Accordingly, much of the blue has been removed by scattering and
what’s left is red.

Experimental assessment of fluorescence anisotropy

For most experiments the sample is excited with vertically polarized light. That is,
𝐸"⃗ = 𝐸(𝑥⃗, 𝑡)𝒌 (2.1)
When the emission polarizer is oriented parallel (∥) to the direction of the polarized excitation
the observed intensity is called 𝐼∥ . Likewise, when the polarizer is perpendicular (⊥) to the
excitation the intensity is called 𝐼A . These intensity values are used to calculate the anisotropy:
B !B
𝑟 = B ∥C)B, (2.2)
∥ ,
The anisotropy is dimensionless quantity that is independent of the total intensity of the
sample. This is because the difference (𝐼∥ − 𝐼A ) is normalized by the total intensity, 𝐼D = 𝐼∥ +
2𝐼A .
 Figure 2. Schematic diagram for
measurement of fluorescence
anisotropies.

Excitation Photoselection of Fluorophores

When a sample is illuminated with polarized light, those molecules with absorption
transitions aligned parallel to the electric vector of the polarized excitation have the highest
probability of excitation.
Recall b𝑒E c𝜇̂ |𝑒" ⟩ is known as transition dipole moment, and is denoted as 𝜇⃗E? .
) )
f |𝑒? ⟩c = c𝜇⃗E? ∙ 𝐸"⃗ c
cb𝑒E c𝜇̂ ∙ 𝐸? 𝒏
) )
= c𝜇⃗E? c c𝐸"⃗ c cos ) (𝜃) (2.3)
where
𝜇⃗E? ∙ 𝐸"⃗
cos(𝜃) =
c𝜇⃗E? cc𝐸"⃗ c

The probability rate of absorption is given by

𝑑𝑃?→E
= 𝐵?→E 𝑢i𝜈E" k
𝑑𝑡
where
2𝜋 )
𝐵?→E = ) c𝜇⃗E? c
3ℏ

Hence, excitation with polarized light results in a population of excited fluorophores that are
partially oriented along the z-axis. As a result, excitation with polarized light results in a
population of excited fluorophores that are partially oriented along the z-axis. This phenomenon
is called photoselection.

Similarly, the probability of emission is proportional to

)
f |𝑒? ⟩|) = c𝜇⃗+? ∙ 𝐸"⃗ c
|⟨𝑒+ |𝜇̂ ∙ 𝐸? 𝒏 (2.4)
 Let us define coordinate system in molecular frame where the axes that are fixed in the
body of the molecule 𝒆𝟏 , 𝒆𝟐 , and 𝒆𝟑 (see Fig 3). For convenience, let us consider the special case
where the axis 𝒆𝟑 coincides with the direction of the transition dipole moment 𝜇⃗E? .
On the other hand, space fixed axis is defined to be our usual axis 𝒊, 𝒋, and 𝒌 that is fixed to
space. Thus when the molecule rotates, axis 𝒆𝟏 , 𝒆𝟐 , and 𝒆𝟑 will rotate while 𝒊, 𝒋, and 𝒌 do not.
We can describe the orientation of the transition dipole 𝜇⃗E? in either molecular frame or space
axis frame. For example, 𝜇⃗E? in molecular frame is always pointing at 𝒆𝟑 .

Figure 3. Molecular frame vs

space fixed frame. One can
transform the coordinates from
molecular frame to space fixed
frame by applying successive
rotations rotation around 𝒌 by 𝜙
followed by rotation around 𝒋 by
𝜃.

One can easily transform the coordinates in molecular fixed axis to the coordinates in the space-
fixed axis by the rotation operator 𝑅V (𝜒, 𝜃, 𝜙).
There are two types of rotations, active and passive rotations.
Active rotation refers to when we rotate molecule counterclockwise (i.e., coordinates of
molecule rotate). This is equivalent to transforming the coordinates from space-fixed axes to
molecular fixed axes.
𝑅VJK→LK (𝜒, 𝜃, 𝜙) = 𝑅VMNO9PQ (𝜒, 𝜃, 𝜙) = 𝑅V3 (𝜒)𝑅V2 (𝜃)𝑅V3 (𝜙)
cos(𝜙) − sin(𝜙) 0 cos(𝜃) 0 − sin(𝜃) cos(𝜒) − sin(𝜒) 0
= usin(𝜙) cos(𝜙) 0w u 0 1 0 w usin(𝜒) cos(𝜒) 0w
0 0 1 sin( 𝜃 ) 0 cos(𝜃 ) 0 0 1
When we want the coordinate from molecular axis to space-fiexed frame, we can think of this
action as rotating the axes clockwise or passive rotation, that is inverse of the 𝑅V R=$"S% (𝜒, 𝜃, 𝜙)
!+ !+
𝑅VLK→JK (𝜒, 𝜃, 𝜙) = 𝑅VTU889PQ (𝜒, 𝜃, 𝜙) = 𝑅V3 (𝜙)!+ 𝑅V2 (𝜃)𝑅V3 (𝜒)
cos(𝜙) sin(𝜙) 0 cos(𝜃) 0 sin(𝜃) cos(𝜒) sin(𝜒) 0
= u−sin(𝜙) cos(𝜙) 0w u 0 1 0 w u−sin(𝜒) cos(𝜒) 0w
0 0 1 −sin 𝜃( ) 0 cos(𝜃) 0 0 1
cos(𝜒) cos(𝜃) cos(𝜙) − sin(𝜒) sin(𝜙) sin(𝜒) cos(𝜃) cos(𝜙) + cos(𝜒) sin(𝜙) sin(𝜃) cos(𝜙)
= (− cos(𝜒) cos(𝜃) sin(𝜙) − sin(𝜒) cos(𝜙) − sin(𝜒) cos(𝜃) sin(𝜙) + cos(𝜒) cos(𝜙) − sin(𝜃) sin(𝜙),
− cos(𝜒) sin(𝜃) −sin(𝜒) sin(𝜃) cos(𝜃)
In this note, we refer to 𝑅V(𝜒, 𝜃, 𝜙) = 𝑅VLK→JK (𝜒, 𝜃, 𝜙)
Let us consider the basis ℬ = 𝒆𝟏 , 𝒆𝟐 , and 𝒆𝟑
0
LK
𝜇⃗V = 0𝒆𝟏 + 0𝒆𝟐 + 𝜇V,3 𝒆𝟑 =u 0 w
LK
𝜇V,3
Space fixed axis be coordinate system with the basis ℰ = 𝒊, 𝒋, and 𝒌.
Thus
JK
𝜇V,- 0 sin(𝜃) cos(𝜙)
𝜇⃗V = x𝜇V,2 y = 𝑅V(𝜒, 𝜃, 𝜙) u 0 w = 𝜇V,3 x− sin(𝜃) sin(𝜙)y
JK LK
(2.5)
LK
JK
𝜇V,3 𝜇V,3 cos(𝜃 )
LK LK LK
Similarly, if emission dipole moment 𝜇⃗% is given by 𝜇%,- 𝒆𝟏 + 𝜇%,2 𝒆𝟐 + 𝜇%,3 𝒆𝟑
LK
𝜇%,-
LK
𝜇⃗% = x𝜇%,2 y
LK
𝜇%,3
Then the coordinates in space-fixed axis can be given by
JK LK
𝜇%,- 𝜇%,-
JK
𝜇⃗% = x𝜇%,2 y = 𝑅V (𝜒, 𝜃, 𝜙) x𝜇%,2
LK
y (2.6)
JK LK
𝜇%,3 𝜇%,3

Figure 4. Absorption and emission transition

dipole moment in molecular fixed axes.

In isotropic liquid, the fluorophores are oriented randomly with probability sin(𝜃)𝑑𝜃𝑑𝜙 and the
probability or the angular distribution of molecules excited by vertically polarized light is given
by
Probability density of •inding molecule when 𝜇⃗U is at (𝜒, 𝜃, 𝜙) =
+
𝑃(𝜒, 𝜃, 𝜙) = X') (2.7)
With this transformation one can evaluate

YZ[\. [^ Q_9889[: YZ[\. [^ U\8[ZYO9[: YZ[\. [^ b9:c9:d _[eQNfeQ8

UO (`,<,a) UO (`,<,a) UO (`,<,a)
)' )' '
𝐼A = 𝐶 ˆ ˆ ˆ ‰
|𝜇⃗Š%‹
∙Š
𝒋|Œ) ‰Š‹ŠŒ
|𝜇⃗V ∙ 𝒌|) ‰ŠŠ‹Š
𝑃(𝜒, 𝜃,ŠŒ
𝜙) sin(𝜃)𝑑𝜃𝑑𝜙 𝑑𝜒
? ? ?
LK )
)' )' '
𝜇%,- 0 )

= 𝐶ˆ ˆ ˆ •𝑅V(𝜒, 𝜃, 𝜙) x𝜇%,2
LK
y ∙ 𝒌• Ž𝑅V(𝜒, 𝜃, 𝜙) u 0 w ∙ 𝒌Ž sin(𝜃)𝑑𝜃𝑑𝜙 𝑑𝜒
LK
? ? ? LK
𝜇%,3 𝜇V,3
 X')
= +g 𝐶|𝜇⃗V |) |𝜇⃗% |) (2 − cos(𝜉)) ) (2.8)
where 𝜉 is the angle between the absorption and emission transition dipole.
In HW#3, you will explicitly calculate above integral.

Similarly, we get
)' )' '
𝐼∥ = 𝐶 ˆ ˆ ˆ |𝜇⃗% ∙ 𝒌|) |𝜇⃗V ∙ 𝒌|) P(𝜃, 𝜙, 𝜒)𝑑𝜃𝑑𝜙 𝑑𝜒
? ? ?
+C)N[8) (h)
= +g
(2.9)
and thus
)!N[8) (h)
𝐼A = (2.10)
+g
Taking Eqs.(2.7)-(2.8) and substitute into Eq.(2.2),
1N[8) (h)!+
𝑟= g
(2.11)

Experimental calculation of anisotropy

In practice, for L-shape detection, let us denote 𝐼ii where V stands for the vertical polarization.
Similarly,
𝐼ij where H stands for the horizontal polarization.
𝐼ii = 𝑘𝑆i 𝐼∥ (3.1)
𝐼ij = 𝑘𝑆j 𝐼A (3.2)
where 𝑘 is a proportionality factor to account for the quantum yield of the fluorophore and other
instrumental factor aside from the polarization-dependent sensitivity.
Definition: fluorescence quantum yield 𝜙K describes the fraction of excited state that become
deexcited through fluorescence:
,-
𝜙K = (3.3)
k,- C,./ C,.0/ C,1 [m]o
where 𝑘K is the rate constant of fluorescence; 𝑘9N is the rate constant of internal conversion; 𝑘98N
is the rate constant of intersystem crossing; and 𝑘p is the rate constant of quenching with
quencher concentration [𝑄].
Division of Eq.(2.10) and (2.11) yields
B22 q B B
B
= q2 B ∥ = 𝐺 B ∥ (3.4)
23 3 , ,
where
q
𝐺 = q2 (3.5)
3
If the measured intensities are not corrected for the factor 𝐺, the calculated anisotropy will be
incorrect. For example, assume 𝐺 = 2.0. This means that the emission monochromator passes
vertically polarized light, at the chosen emission wavelength, with twofold greater efficiency
than horizontally polarized light. The 𝐺 factor is dependent upon the emission wavelength, and
to some extend the bandpass of the monochromator.
The 𝐺 factor is easily measured using horizontally polarized excitation. With horizontally
polarized excitation the excited-state distribution is rotated to lie along the observation axis.
When this is done both the horizontally and vertically polarized components are equal and
proportional to 𝐼A .
 That is, recall 𝐼∥ and 𝐼A are defined to be the emission intensities parallel and
perpendicular to the excitation. Let’s suppose that the excitation is polarized in the y direction in
Figure 2. Then 𝐼2 is parallel to the excitation while 𝐼- and 𝐼3 are perpendicular to the excitation.
Therefore, 𝐼- = 𝐼2 = 𝐼A in isotropic solution. Since emission polarizer detects 𝐼- and 𝐼3
components of the emission, we have that 𝐼ji = 𝑘𝑆i 𝐼3 and 𝐼jj = 𝑘𝑆j 𝐼- . Equivalently,
𝐼ji = 𝑘𝑆i 𝐼A (3.6)
𝐼jj = 𝑘𝑆j 𝐼A (3.7)
Division of Eq.(2.15) and (2.16) yields
q B B
𝐺 = q2 B, = B32 (3.8)
3 , 33
When factor is known the ratio can be calculated using
B22 q B B
B r
= q2 B ∥ = B ∥ (3.9)
23 3 , ,
The anisotropy is given by
4∥
!+
4,
𝑟= 4∥ (3.10)
C)
4,
In terms of 𝐼ii and 𝐼ij , Eq.(2.19) has alternative form:
B !rB
𝑟 = B 22C)rB23 (3.11)
22 23

Effect of molecular motion

Most macromolecules of biological interest fall between the two extreme cases: their
rotational motions are not negligible on the fluorescence time scale, but neither can they tumble
fast enough to achieve random orientation. Suppose that a fluorescent probe is rigidly attached to
the macromolecule. The observed polarization will be some intermediate value between
Eq.(2.11) and zero. To compute this value, we must analyze the relative rate of emission and
macromolecular rotational motion. Note that translational motion does not affect fluorescence
polarization. Only motion that changes the orientation dipoles can be observed.
First consider what happens if the sample is excited by a pulse of polarized light, and the
time dependence of 𝐼∥ and 𝐼A is measured. If emission and absorption transition dipoles are
parallel, the earliest photons to be emitted are highly likely to be z-polarized because the
molecules have not had time to reorient. The last photons to be emitted should have random
polarization because by then the system has experienced considerable rotational motion.
Therefore, if anisotropy is measure as a function of the time of fluorescence emission, it will
decay from initial values of 𝑟(𝑡 = 0) = 𝑟s to final values of zero. The rate of decay is a measure
of the rate of rotational motion.
Rotational Brownian motion is described by a diffusion equation quite analogous to that
used for translational motion. If 𝑊(𝜃, 𝜙, 𝑡) is the probability of finding the spherical excited
molecule at 𝜃 to 𝜃 + 𝑑𝜃 and 𝜙 to 𝜙 + 𝑑𝜙 at time 𝑡, then it follows rotational diffusion equation.
tu(`,<,a,$ |`% ,<% ,a% )
t$
= 𝐷Z[O ∇)*s$ 𝑃(𝜒, 𝜃, 𝜙, 𝑡|𝜒s , 𝜃s , 𝜙s ) (4.1)
where
1 𝜕 𝜕 1 𝜕)
∇)*s$ = —sin(𝜃) ˜ + )
sin(𝜃) 𝜕𝜃 𝜕𝜃 sin (𝜃) 𝜕 ) 𝜙
The solution to Eq.(4.1) is
 { v
$
∗ !y
𝑃(𝜒, 𝜃, 𝜙, 𝑡|𝜒s , 𝜃s , 𝜙s ) = ™ ™ 𝑌v,w (𝜃, 𝜙)𝑌v,w (𝜃s , 𝜙s )𝑒 5

vz? wz!v
)vC+
= ∑{ v v v∗
vz? ∑wz!v 𝐷w,? (𝜃, 𝜙, 𝜒)𝐷w,? (𝜃s , 𝜙s , 𝜒s )𝑒
!$/y5
(4.2)
&'
where
+
𝜏v = v(vC+)} (4.3)
678
and 𝑐v,w is a constant determined by the initial distribution; 𝑌v,w (𝜃, 𝜙) is the l,m of spherical
v (𝜃, v (𝜃,
harmonics. 𝐷w,~ 𝜙, 𝜒) is the Wigner rotation matrix. 𝐷w,~ 𝜙, 𝜒) and 𝑌v,w (𝜃, 𝜙) are
related by
2𝑙 + 1 v
𝑌v,w (𝜃, 𝜙) = 𝐷w,? (𝜃, 𝜙, 𝜒)
4𝜋

Using the relation,

)' ∗ '
∫? ∫? 𝑌v,w (𝜃, 𝜙)𝑌v0,w0 (𝜃, 𝜙) sin(𝜃) 𝑑𝜃𝑑𝜙 = 𝛿v,v0 𝛿w,w0 (4.4)
we get
)' )' '
𝐼∥ (𝑡) = 𝐶 ˆ ˆ ˆ |𝜇⃗% (𝑡) ∙ 𝒌|) |𝜇⃗V (𝑡 = 0) ∙ 𝒌|) 𝑃(𝜒s , 𝜃s , 𝜙s )𝑑𝜃s 𝑑𝜙s 𝑑𝜒s
? ? ?
LK )
𝜇%,- 0 )
)' )' ' )' )' '
•𝑅V (𝜃, 𝜙, 𝜒) x𝜇%,2
LK
y ∙ 𝒌• Ž𝑅V(𝜒s , 𝜃s , 𝜙s ) u 0 w ∙ 𝒌Ž
=𝐶ˆ ˆ ˆ ˆ ˆ ˆ LK ×
? ? ? ? ? ?
LK
𝜇%,3 𝜇V,3
𝑃 (𝜒, 𝜃, 𝜙, 𝑡|𝜒s , 𝜃s , 𝜙s )𝑃(𝜒s , 𝜃s , 𝜙s )𝑑𝜃s 𝑑𝜙s 𝑑𝜒s 𝑑𝜃𝑑𝜙𝑑𝜒
(4.5)
)' )' '
𝐼A (𝑡) = 𝐶 ˆ ˆ ˆ |𝜇⃗% (𝑡) ∙ 𝒋|) |𝜇⃗V (𝑡 = 0) ∙ 𝒌|) 𝑃(𝜒s , 𝜃s , 𝜙s )𝑑𝜃s 𝑑𝜙s 𝑑𝜒s
? ? ?
LK )
𝜇%,- 0 )
)' )' ' )' )' '
•𝑅V(𝜃, 𝜙, 𝜒) x𝜇%,2
LK
y ∙ 𝒋• Ž𝑅V (𝜒s , 𝜃s , 𝜙s ) u 0 w ∙ 𝒌Ž
=𝐶ˆ ˆ ˆ ˆ ˆ ˆ LK ×
? ? ? ? ? ?
LK
𝜇%,3 𝜇V,3
𝑃(𝜒, 𝜃, 𝜙, 𝑡|𝜒s , 𝜃s , 𝜙s )𝑃(𝜒s , 𝜃s , 𝜙s )𝑑𝜃s 𝑑𝜙s 𝑑𝜒s 𝑑𝜃𝑑𝜙𝑑𝜒
(4.6)
If one substitutes Eqs.(2.7), (4.2) into Eq.(4.5) and (4.6). One obtains
$
+ & !9 1N[8) (h)!+
𝐼∥ (𝑡) = ¤1 + +g 𝑒 ) 9 )
:¥ (4.7)
$
+ ) !9 1N[8) (h)!+
𝐼A (𝑡) = ¤1 − +g 𝑒 ) 9 )
:¥ (4.8)

Define fluorescence decay time

!+
𝜏K = ¦𝑘K + 𝑘9N + 𝑘98N + 𝑘p [𝑄]§
Eqs.(4.7)-(4.8) will be modified by taking into account of fluorescence decay and get
The overall time-dependence of the 𝐼∥ (𝑡) and 𝐼A (𝑡) as
 $
+ & !9 1N[8) (h)!+
𝐼∥ (𝑡) = ¤1 + +g 𝑒 ) 9 )
:¥ 𝑒 !$/y- (4.9)
$
+ ) !9 1N[8) (h)!+
𝐼A (𝑡) = ¤1 − +g 𝑒 ) 9 )
:¥ 𝑒 !$/y- (4.10)
Substituting in Eq.(2.2) one gets
$ $
) !9 1N[8) (h)!+ !
𝑟(𝑡) = g 𝑒 ) 9 )
: = 𝑟s 𝑒 9: (4.11)
where
3cos ) (𝜉) − 1
𝑟s =
5
and
1
𝜏• = 𝜏) =
6𝐷Z[O
The time averaged or steady-state anisotropy 〈𝑟〉 is given by
'%
$
; ! > >
∫< *($)% 9- t$ = ? @
9) 9- y
〈𝑟〉 = $ = > = 𝑟s / 91 + y - : (4.12)
; ! :
∫< % 9- t$ 9-

or
+ + y-
*
=* +* (4.13)
% % y:
The Eq.(4.12) or Eq.(4.13) is known as Perrin Equation.
Stoke-Einstein relation says that for a spherical molecule,
𝑘• 𝑇
𝐷Z[O =
8𝜋𝜂𝑎1
where 𝑘• is Boltzmann constant; 𝑇 is temperature; 𝜂 is a viscosity of the solvent; and 𝑎 is the
radius of the molecule.

Notice that
)' )' '
𝐼∥ (𝑡) = 𝐶 ˆ ˆ ˆ |𝜇⃗% (𝑡) ∙ 𝒌|) |𝜇⃗V (𝑡 = 0) ∙ 𝒌|) 𝑃(𝜃s , 𝜙s , 𝜒s )𝑑𝜃s 𝑑𝜙s 𝑑𝜒s
? ? ?
has the form of
)' )' '
ˆ ˆ ˆ 𝐴(𝑡)𝐵(𝑡 = 0)𝑃(𝜃s , 𝜙s , 𝜒s )𝑑𝜃s 𝑑𝜙s 𝑑𝜒s = 〈𝐴(𝑡)𝐵(𝑡 = 0)〉
? ? ?
where 𝐴(𝑡) and 𝐵(𝑡) are some physical quantities of interest, and 〈 〉 denotes averaging of the
angular dependence on 𝐴 and 𝐵. 〈𝐴(𝑡)𝐵(𝑡 = 0)〉 measures the time-correlation of the quantities
𝐴 and 𝐵 thus is called time-correlation function.