Chapter 2

Chain Length Distribution of Polymers

2.1 Population Balance Equations

In the following we concentrate on linear chains only which means kfp = 0 and no propagation
of the terminal double bond.

2.1.1 Active Chains

Under the assumption of pseudo steady state for species R•1 ,
dR•1
=0 (2.1)
dt
we can set equation (1.3) equal to zero:
 
RI + (kfm M + kfs S ) R• − R•1 = kp MR•1 + (ktc + ktd ) R• R•1 (2.2)

Solving for R•1 we obtain:

RI + (kfm M + kfs S ) R•
R•1 = (2.3)
kp M + (ktc + ktd ) R• + (kfm M + kfs S )
Since RI is given by equation (1.10):

RI = (ktc + ktd ) R•2 (2.4)

we can simplify equation (2.3) as follows:

α
R•1 = R• (2.5)
1+α
where α is the summation of various characteristic time ratios:
ktc R• ktd R• kfm M kfs S
α = + + +
kp M kp M kp M kp M
τp τp τp τp
= + + + (2.6)
τtc τtd τfm τfs

19
CHAPTER 2. CHAIN LENGTH DISTRIBUTION OF POLYMERS

where τp is the characteristic time of the addition reaction of a single monomer unit, while the
τ at the denominator are the characteristic times of the chain transfer and the termination reac-
tions involved in the system. The order of magnitude of these characteristic times, which refer
to second-order processes, can be computed using pseudo-first order reaction approximation.
So far the monomer addition reaction, r = kp MR•n and the pseudo-first order rate constant is
kp M, so that τp = kp1M . In most applications, we have that τp ≈ 10−3 s while termination
reactions (or at least the dominating one) has τf or τt ≈ 100 s. This means that a polymer chain
leaves about 1 s and in this time it adds about 103 monomer units. Note that the characteristic
time of the process is that of monomer consumption, τM = k 1R• which is of the order of 103
p
s. These are the three time scales dominating the process under examination. From equation
(2.5) we can see that for α  1 ⇒ R•1  R• , as expected.
If we apply the pseudo steady state assumption now to the species R•n (n ≥ 2), we can write:

dR•n
=0 (2.7)
dt

 
kp MR•n−1 = kp M + kfm M + kfs S + (ktc + ktd ) R• R•n (2.8)

⇒ R•n = R•n−1 Φ (2.9)

where
1
Φ= (2.10)
1+α
Therefore we obtain:
R•n = R•n−1 Φ = R•n−2 Φ2 = ... = R•1 Φn−1 (2.11)

and using equation (2.5):

1 • α α
R•n = R = R •
(2.12)
(1 + α)n−1 1 + α (1 + α)n

Thus, we can consolidate the two cases (n = 1 and n ≥ 2) and write the general expression for
n ≥ 1:
α
R•n = R• (2.13)
(1 + α)n

20
CHAPTER 2. CHAIN LENGTH DISTRIBUTION OF POLYMERS

Note

• equation (2.13) can be simplified if α  1:

1
≈ exp (−αn) for α  1 (2.14)
(1 + α)n

R•n
⇒ ≈ α exp (−αn) (2.15)
R•

Rn
R

a increases

   
∞
R•n ∞
 exp (−αn) ∞
• dn = α exp (−αn) dn = α −  = 1
0 R• 0 α 0

To summarize, we can state that the distribution of active chains is normalized and, moreover,
it is characterized by only one kinetic parameter α defined by equation (2.6).

2.1.2 Dead Chains

1  • •
n−1
dPn
= (kfm M + kfs S + ktd R• ) R•n + ktc RR (2.16)
dt 2 j=1 j n− j

Here we can write:

⎧ 
⎪
⎪
⎪
1
R•1 R•2 + R•2 R•1 = R•1 R•2 (n = 3)
1 
n−1 ⎪
⎪
⎨
2 
ktc R•j R•n− j =⎪
⎪
1
R• R• + R•2 • • • • 1 •2
2 + R3 R1 = R1 R3 + 2 R2 (n = 4) (2.17)
2 ⎪
⎪
⎪
2 1 3
j=1 ⎩ 1 • • • • • • • • • • • •
R1 R4 + R2 R3 + R3 R2 + R4 R1 = R1 R4 + R2 R3 (n = 5)
2

21
CHAPTER 2. CHAIN LENGTH DISTRIBUTION OF POLYMERS

If we use equation (2.13) to express R•n , we obtain for the sum term:

n−1 n−1   
• • • α • α
ktc R j Rn− j = ktc R R
j=1 j=1
(1 + α) j (1 + α)n− j
R•2 α2 
n−1
= ktc (1)
(1 + α)n j=1
α2 (n − 1) •2
= ktc R (2.18)
(1 + α)n
Therefore we can rewrite equation (2.16) as follows:
dPn α 1 α2 (n − 1) •2
= (kfm M + kfs S + ktd R• ) R •
+ ktc R (2.19)
dt (1 + α)n 2 (1 + α)n
where we see that α alone is not anymore sufficient to fully describe the distribution, and we
have to introduce further parameters:
dPn   kfm kfs S ktd R•  α
 
ktc R• 1 α2 (n − 1)

•
= kp MR + + + (2.20)
dt kp kp M kp M (1 + α)n kp M 2 (1 + α)n
The introduced parameters are defined as follows:
kfm kfs S ktd R• τp τp τp
γ= + + = + + (2.21)
kp kp M kp M τ f m τ f s τtd

ktc R• τp
β= = (2.22)
kp M τtc
where
α=β+γ (2.23)
Finally, we obtain for n ≥ 1:
 
dPn α 1
= Rp γ + (n − 1) αβ (2.24)
dt (1 + α)n 2
Comparing equations (2.13) and (2.24) it is easy to see that in the case of active chain distribu-
tion only the ratio between the characteristic times of propagation and termination is relevant.
On the other hand, dealing with the dead chain distribution we have to distinguish between
two different types of chain reactions: chain transfers and disproportionation on one side, and
combination on the other side.
The question is: what is the conceptual difference between these reactions which makes them
to play a different role in determining the chain length distribution?

22
CHAPTER 2. CHAIN LENGTH DISTRIBUTION OF POLYMERS

2.2 Instantaneous Chain Length Distribution

In the following section we consider the instantaneous CLD of polymer chains. Different
distributions can be introduced.

2.2.1 Number Distribution and Number Average

The instantaneous CLD or number distribution is defined as follows:
dPn
polymer chains of length n
fN (n) = = ∞ dt (2.25)
total number of polymer chains  dPn
n=1
dt

Note that this is an instantaneous property, that is it represents the CLD of the infinitesimal
amount of polymer produced at a given time. Using equation (2.24) we obtain:

γ + 12 (n − 1) β (β + γ)
(1 + α)n
fN (n) = ∞ ⎡
 ⎢⎢ γ + 1 (n − 1) β (β + γ) ⎤⎥⎥
(2.26)
⎢⎢⎣ 2 ⎥⎥⎦
n=1
(1 + α) n

Since the following equality holds for geometric series


∞
aq bq
(a + nb) qn = + for |q| < 1 (2.27)
n=1
(1 − q) (1 − q)2

we can rewrite the denominator in equation (2.26) as follows:

∞   n  ∞   n
1 1 1 1
γ − β (β + γ) + n β (β + γ)
n=1
2 1 + α n=1
2 1+α
    
1 1 1 1+α
= γ − β (β + γ) + β (β + γ) (2.28)
2 α 2 α2

and since α = β + γ      
1 1 1 1+α 1 1
=γ − β+ β = γ+ β (2.29)
α 2 2 α α 2
Finally, we obtain for equation (2.26):
⎡ ⎤
α ⎢⎢⎢ γ + 12 (n − 1) β (β + γ) ⎥⎥⎥
fN (n) = ⎢⎣ ⎥⎦ (2.30)
(1 + α)n γ + 12 β

23
CHAPTER 2. CHAIN LENGTH DISTRIBUTION OF POLYMERS

By definition holds:

∞
fN (n) = 1 (2.31)
n=1

Let us focus now on the moments of this distribution. In particular, the j-th order moment of
fN (n) is defined as follows:

∞
μj = n j fN (n) (2.32)
n=1

From the first two moments we can calculate the number average of the CLD:
μ1 1
nN = ≈ for α  1 (2.33)
μ0 γ + 12 β

which, when multiplied by the molecular weight of the monomer, MM , leads to the so-called
number average molecular weight:
μ1
MN = MM (2.34)
μ0

2.2.2 Weight Distribution and Weight Average

Another way to characterize the distribution of the polymer chains is the so-called instanta-
neous weight distribution:
n · dPn
number of monomer units in the chains of length n
fW (n) = = dt (2.35)
total amount of consumed monomer Rp
or expressed in terms of fN :
n fN (n) n fN (n)
fW (n) = ∞ = (2.36)
1 n fN (n) μ1

If we compute the mean of this distribution we obtain the weight average of the CLD:

1  2
∞
μ2
nW = n fN (n) = (2.37)
μ1 1 μ1

Note that the moments used in the last two expressions refer to fN . Using equations (2.30) and
(2.36) we obtain: 
2 γ + 32 β
nW ≈ for α  1 (2.38)
(γ + β)2

24
CHAPTER 2. CHAIN LENGTH DISTRIBUTION OF POLYMERS

n N nW

fW

fN(n) , fW(n)
fN

which, when multiplied by the molecular weight of the monomer, MM , leads to the so-called
weight average molecular weight:
μ2
MW =
MM (2.39)
μ1
To summarize, we defined two different means nN and nW of only one distribution fN . The
reason for this is given in the following subsection.

2.2.3 Polydispersity
The polydispersity index (PDI) is a widely used quantity to characterize distributions in gen-
eral and the CLD in particular. This is defined as follows:
nW
σ= ≥1 (2.40)
nN

μ2 /μ1 μ2 μ0
⇒σ= = 2 (2.41)
μ1 /μ0 μ1
Thus, σ = 1 means that all chains have exactly the same length. The polydispersity index
describes the width of a distribution and is therefore an alternative way to represent its variance
v:
∞
v = (n − nN )2 fN (n) = μ2 − 2μ1 nN + n2N μ0
n=1
μ21 μ21 μ2
= μ2 − 2 + 2 μ0 = μ2 − 1
μ0 μ0 μ0
μ21
= (σ − 1) (2.42)
μ0

25
CHAPTER 2. CHAIN LENGTH DISTRIBUTION OF POLYMERS

Note that v = 0 corresponds to σ = 1. The usage of the polydispersity to characterize the width
of CLD of polymers is justified by the fact that equation (2.40) is particularly convenient in
practice. We have in fact various techniques available that can measure either the number
average or the weight average of polymer chains.

2.2.4 Summary
For the reaction scheme of the free-radical polymerization system described in section 1.4.1
we can write the following relationships for the instantaneous properties of the CLD (if α 
1):
1
nN = (2.43)
γ + β2

2 γ + 32 β
nW = (2.44)
(γ + β)2
 
2 γ + 32 β γ + 12 β
σ= (2.45)
(γ + β)2

26