Skip to content

Molar Mass Averages and Distributions

CSIR NET polymer chemistry: number-, weight-, z- and viscosity-average molar masses, dispersity, degree of polymerisation and molar mass distributions.

  • 18 min read
  • Updated
On this page · 14 sections
  1. Prerequisites
  2. The Hook: One Sample, Three “Correct” Molar Masses
  3. Polymers, Repeat Units and the Degree of Polymerisation
  4. The Averages
  5. Dispersity and What It Measures
  6. Worked Examples
  7. Molar Mass Distributions
  8. Memory Tricks
  9. Don’t Confuse
  10. Common Mistakes
  11. Practice
  12. Quick Revision Table
  13. Key Takeaways
  14. What’s Next

Prerequisites

The Hook: One Sample, Three “Correct” Molar Masses

Polymers, Repeat Units and the Degree of Polymerisation

A polymer (macromolecule) is built by linking many small molecules, the monomers, through covalent bonds. The repeat unit (constitutional repeating unit) is the smallest unit whose repetition describes the chain: \( \mathrm{-CH_2-CH(C_6H_5)-} \) for polystyrene, \( \mathrm{-NH(CH_2)_6NH-CO(CH_2)_4CO-} \) for nylon-6,6.

The degree of polymerisation \( X \) (also written DP) is the number of monomer-derived units in a chain. If end groups are negligible,

$$M = X\,M_0$$

where \( M_0 \) is the molar mass of the monomer-derived (structural) unit.

PolymerStructural unit\( M_0 \) / g mol⁻¹
Polyethylene\( \mathrm{-CH_2CH_2-} \)28.05
Poly(vinyl chloride)\( \mathrm{-CH_2CHCl-} \)62.50
Poly(methyl methacrylate)\( \mathrm{-CH_2C(CH_3)(CO_2CH_3)-} \)100.12
Polystyrene\( \mathrm{-CH_2CH(C_6H_5)-} \)104.15
Poly(ethylene oxide)\( \mathrm{-CH_2CH_2O-} \)44.05
Nylon-6,6 (repeat unit)\( \mathrm{-NH(CH_2)_6NHCO(CH_2)_4CO-} \)226.32

Polymer chains in a sample differ in \( X \) because chain growth and termination are random events. Proteins and nucleic acids made by templated biosynthesis are the exception: every molecule of a given protein has the same sequence and the same molar mass.

The Averages

Let a sample contain \( N_i \) moles (or molecules) of chains of molar mass \( M_i \). The mass of that fraction is \( m_i = N_iM_i \), its number fraction is \( n_i = N_i/\sum N_i \) and its weight (mass) fraction is \( w_i = m_i/\sum m_i \).

Number-average molar mass

Number average
$$\overline{M}_n = \frac{\sum_i N_iM_i}{\sum_i N_i} = \sum_i n_iM_i = \frac{\sum_i m_i}{\sum_i m_i/M_i} = \frac{1}{\sum_i w_i/M_i}$$

It is the total mass divided by the total number of molecules: every chain counts once, whatever its size.

Any method that counts molecules gives \( \overline{M}_n \): colligative properties (osmotic pressure, vapour-pressure lowering), end-group titration or NMR, and the peak areas of a MALDI mass spectrum.

Weight-average molar mass

Weight average
$$\overline{M}_w = \frac{\sum_i N_iM_i^2}{\sum_i N_iM_i} = \sum_i w_iM_i$$

Each chain is weighted by its mass, so the heavy chains count more.

Light scattering gives \( \overline{M}_w \): the intensity scattered by a molecule is proportional to the square of its mass, so the signal from a sample of mass concentration \( c \) is proportional to \( c\,\overline{M}_w \).

z-average and higher averages

$$\overline{M}_z = \frac{\sum_i N_iM_i^3}{\sum_i N_iM_i^2}, \qquad \overline{M}_{z+1} = \frac{\sum_i N_iM_i^4}{\sum_i N_iM_i^3}$$

Sedimentation equilibrium in the ultracentrifuge yields \( \overline{M}_w \) and \( \overline{M}_z \). The radius of gyration measured by light scattering is a z-average quantity.

Viscosity-average molar mass

Dilute-solution viscometry with the Mark–Houwink equation \( [\eta] = KM^a \) gives

$$\overline{M}_v = \left(\frac{\sum_i N_iM_i^{1+a}}{\sum_i N_iM_i}\right)^{1/a} = \left(\sum_i w_iM_i^{a}\right)^{1/a}$$

For \( a = 1 \), \( \overline{M}_v = \overline{M}_w \); for \( a = -1 \) the formula would give \( \overline{M}_n \). Flexible coils have \( 0.5 \le a \le 0.8 \), so \( \overline{M}_v \) lies between \( \overline{M}_n \) and \( \overline{M}_w \), usually much closer to \( \overline{M}_w \).

Moments: one formula for all of them

Define the \( k \)th moment of the number distribution \( \mu_k = \sum_i N_iM_i^k \). Then

$$\overline{M}_n = \frac{\mu_1}{\mu_0}, \qquad \overline{M}_w = \frac{\mu_2}{\mu_1}, \qquad \overline{M}_z = \frac{\mu_3}{\mu_2}$$

Each average is a ratio of successive moments, and each higher moment weights the large chains more strongly. That is why, for any polydisperse sample,

$$\boxed{\overline{M}_n < \overline{M}_v < \overline{M}_w < \overline{M}_z < \overline{M}_{z+1}}$$

with equality throughout only for a uniform (monodisperse) sample.

graph LR
    A["M_n
count molecules
osmometry, end groups"] --> B["M_v
viscometry
(0.5 ≤ a ≤ 0.8)"] B --> C["M_w
weight by mass
light scattering"] C --> D["M_z
weight by mass²
sedimentation equilibrium"]

Dispersity and What It Measures

Dispersity
$$Đ = \frac{\overline{M}_w}{\overline{M}_n} \ge 1$$

IUPAC now recommends the name dispersity (symbol Đ); older texts call the same ratio the polydispersity index (PDI). A uniform polymer has Đ = 1.

Dispersity is not just an empirical index. Write \( \sigma_n^2 \) for the variance of the number distribution:

$$\sigma_n^2 = \frac{\mu_2}{\mu_0} - \left(\frac{\mu_1}{\mu_0}\right)^2 = \overline{M}_w\overline{M}_n - \overline{M}_n^2$$

Dividing by \( \overline{M}_n^2 \):

$$\boxed{\frac{\overline{M}_w}{\overline{M}_n} - 1 = \frac{\sigma_n^2}{\overline{M}_n^2}}$$

So Đ − 1 is the squared relative standard deviation of the number distribution. A sample with \( \overline{M}_n = 50{,}000 \) g mol⁻¹ and Đ = 1.2 has \( \sigma_n = 50{,}000\sqrt{0.2} \approx 22{,}400 \) g mol⁻¹. Even a dispersity that “looks narrow” hides a spread of nearly half the mean.

Dispersities that come from mechanisms

OriginIdeal ĐWhere derived
Uniform sample (a pure protein)1—
Living anionic polymerisation (Poisson distribution)\( 1 + 1/X_n \) approximately, close to 1Chain-growth page
Radical polymerisation, termination by combination, low conversion1.5Chain-growth page
Radical polymerisation, termination by disproportionation or transfer, low conversion2Chain-growth page
Linear step-growth at extent of reaction \( p \)\( 1 + p \to 2 \)Step-growth page

Real samples are usually broader than these ideal values: the monomer and initiator concentrations change during a batch, the gel effect sets in at high conversion, and branching broadens the distribution further.

Worked Examples

Blending two polydisperse samples

Each sample \( j \) contributes mass \( m_j \), with its own \( \overline{M}_{n,j} \) and \( \overline{M}_{w,j} \). Because \( \overline{M}_n \) is total mass over total moles, and \( \overline{M}_w \) is a mass-weighted mean,

$$\overline{M}_{n,\text{blend}} = \frac{\sum_j m_j}{\sum_j m_j/\overline{M}_{n,j}}, \qquad \overline{M}_{w,\text{blend}} = \frac{\sum_j m_j\overline{M}_{w,j}}{\sum_j m_j}$$

Molar Mass Distributions

A distribution can be plotted as a number fraction \( n(M) \) or as a weight fraction \( w(M) \) against \( M \) (or \( X \)). The two curves of the same sample look different: the weight curve is shifted to higher \( M \) because \( w(M) \propto M\,n(M) \). Size-exclusion chromatography records something proportional to the weight distribution, because a concentration detector responds to mass.

The most probable (Flory) distribution

For linear step-growth polymerisation in which every functional group has the same reactivity regardless of chain length (Flory’s equal-reactivity principle), let \( p \) be the fraction of functional groups that have reacted. The probability that a chain has exactly \( x \) units is the probability of \( x - 1 \) reacted links followed by one unreacted end:

$$n_x = (1 - p)\,p^{x-1}, \qquad w_x = x(1 - p)^2p^{x-1}$$

Summing the moments gives

$$\boxed{X_n = \frac{1}{1 - p}, \qquad X_w = \frac{1 + p}{1 - p}, \qquad Đ = 1 + p}$$
\( p \)\( X_n \)\( X_w \)Đ
0.9010191.90
0.991001991.99
0.999100019991.999

Two features are tested repeatedly:

  1. By number, the monomer is always the most abundant species, at every \( p \): \( n_x \) falls monotonically with \( x \).
  2. By weight, the distribution has a maximum near \( x = -1/\ln p \approx X_n \). At \( p = 0.99 \) the maximum is at \( x \approx 99.5 \), and the weight fraction of the 100-mer is only \( 100(0.01)^2(0.99)^{99} = 0.0037 \).

The same exponential distribution arises in radical polymerisation when chains stop by disproportionation or chain transfer (each growing chain has a constant probability of stopping at each step), which is why Đ = 2 there too.

The Poisson distribution

In a living polymerisation all chains start at the same time and grow at the same rate, with no termination. If the mean number of monomers added per chain is \( \nu \), the number fraction with \( x = \nu' + 1 \) units follows a Poisson distribution in \( \nu' \), and

$$X_n = 1 + \nu, \qquad \frac{X_w}{X_n} = 1 + \frac{\nu}{(1 + \nu)^2} \approx 1 + \frac{1}{X_n}$$

For \( X_n = 100 \), Đ ≈ 1.01. The narrowness comes from all chains being born together, not from any special chemistry of the chain end.

Other shapes

Radical polymerisation with termination by combination joins two exponentially distributed chains, which narrows the distribution to Đ = 1.5. More generally, a two-parameter family (the Schulz–Zimm distribution) interpolates between these cases. Branched polymers and polymers made at high conversion can have Đ well above 2.

Memory Tricks

Don’t Confuse

ThisvsThat
\( \overline{M}_n = 1/\sum(w_i/M_i) \): harmonic mean over weight fractions\( \overline{M}_w = \sum w_iM_i \): arithmetic mean over weight fractions
Number fraction \( n_i \)Weight fraction \( w_i = n_iM_i/\overline{M}_n \)
Repeat unit of nylon-6,6 (226.32 g mol⁻¹)Structural unit counted in \( X_n \) (113.16 g mol⁻¹ on average)
Most probable distribution: monomer is most abundant by numberMost probable distribution: weight fraction peaks near \( X_n \)
Dispersity Đ = \( \overline{M}_w/\overline{M}_n \) (current IUPAC name)Polydispersity index (older name for the same ratio)
\( \overline{M}_v \) depends on the solvent through \( a \)\( \overline{M}_n \), \( \overline{M}_w \) are properties of the sample alone
Poisson (living): Đ ≈ 1 + 1/\( X_n \)Most probable (step-growth): Đ = 1 + p

Common Mistakes

Practice

All questions below are original practice questions written for this page. They are not from any CSIR NET paper.

Level 1: Foundation

  1. Which molar-mass average does membrane osmometry give?
  2. State the relation between Đ and the variance of the number distribution.
  3. For the most probable distribution at \( p = 0.95 \), what are \( X_n \) and Đ?
Answers
  1. \( \overline{M}_n \): osmotic pressure depends on the number of solute molecules.
  2. \( Đ - 1 = \sigma_n^2/\overline{M}_n^2 \).
  3. \( X_n = 1/(1 - 0.95) = 20 \); Đ = 1 + 0.95 = 1.95.

Level 2: Part B Style

Practice Q1. A sample contains 30% by mass of chains with \( M = 15{,}000 \) g mol⁻¹ and 70% by mass of chains with \( M = 60{,}000 \) g mol⁻¹. Its number-average molar mass (g mol⁻¹) is closest to:

(a) 46,500 (b) 37,500 (c) 32,000 (d) 26,700

Answer: (c) 32,000

\( \overline{M}_n = 1/(0.3/15{,}000 + 0.7/60{,}000) = 1/(2.0 \times 10^{-5} + 1.167 \times 10^{-5}) = 31{,}600 \approx 32{,}000 \) g mol⁻¹.

Option (a) is \( \sum w_iM_i = 46{,}500 \), the weight average. Option (b) is the plain mean of the two \( M \) values.

Practice Q2. For a polydisperse linear polymer of flexible chains in a good solvent (Mark–Houwink exponent 0.7), the correct order is:

(a) \( \overline{M}_w < \overline{M}_v < \overline{M}_n < \overline{M}_z \) (b) \( \overline{M}_n < \overline{M}_w < \overline{M}_v < \overline{M}_z \) (c) \( \overline{M}_n < \overline{M}_v < \overline{M}_w < \overline{M}_z \) (d) \( \overline{M}_v < \overline{M}_n < \overline{M}_w < \overline{M}_z \)

Answer: (c)

For \( 0 < a < 1 \), \( \overline{M}_v \) lies between \( \overline{M}_n \) and \( \overline{M}_w \); \( \overline{M}_z \) is a higher moment ratio and is the largest of the four.

Practice Q3. Equal numbers of moles of two uniform polymers with \( M = 20{,}000 \) and \( M = 80{,}000 \) g mol⁻¹ are mixed. The dispersity of the mixture is:

(a) 1.25 (b) 1.36 (c) 1.60 (d) 2.00

Answer: (b) 1.36

\( \overline{M}_n = 50{,}000 \). \( \overline{M}_w = (20{,}000^2 + 80{,}000^2)/(100{,}000) = (4 \times 10^8 + 6.4 \times 10^9)/10^5 = 68{,}000 \). Đ = 68,000/50,000 = 1.36.

Practice Q4. A linear polyester follows the most probable distribution with \( X_n = 50 \). The weight-average degree of polymerisation is:

(a) 50 (b) 75 (c) 99 (d) 100

Answer: (c) 99

\( X_n = 1/(1 - p) = 50 \) gives \( p = 0.98 \). \( X_w = (1 + p)/(1 - p) = 1.98/0.02 = 99 \). Option (d) assumes Đ = 2 exactly, which holds only as \( p \to 1 \).

Level 3: Part C Style

Practice Q5. A polymer sample has \( \overline{M}_n = 40{,}000 \) g mol⁻¹ and Đ = 1.5. Consider:

  1. The standard deviation of its number distribution is about 28,300 g mol⁻¹.
  2. Its weight-average molar mass is 60,000 g mol⁻¹.
  3. If 4 g of it is blended with 1 g of a uniform polymer of \( M = 10{,}000 \) g mol⁻¹, the blend has \( \overline{M}_n = 25{,}000 \) g mol⁻¹.
  4. In the blend of statement 3, \( \overline{M}_w = 50{,}000 \) g mol⁻¹.

Which are correct?

(a) 1 and 2 only (b) 1, 2 and 3 only (c) 1, 2, 3 and 4 (d) 2 and 4 only

Answer: (c) 1, 2, 3 and 4
  1. \( \sigma_n = \overline{M}_n\sqrt{Đ - 1} = 40{,}000\sqrt{0.5} = 28{,}300 \). Correct.
  2. \( \overline{M}_w = 1.5 \times 40{,}000 = 60{,}000 \). Correct.
  3. \( \overline{M}_n = 5/(4/40{,}000 + 1/10{,}000) = 5/(2 \times 10^{-4}) = 25{,}000 \). Correct.
  4. \( \overline{M}_w = (4 \times 60{,}000 + 1 \times 10{,}000)/5 = 50{,}000 \). Correct. (The blend has Đ = 2.0, broader than the parent.)

Practice Q6. A step-growth polymer is sampled at \( p = 0.99 \) and is assumed to follow the most probable distribution. Which statements are correct?

  1. Monomer molecules are 1% of all molecules but only 0.01% of the mass.
  2. The weight-fraction distribution has its maximum near \( x = 100 \).
  3. The 100-mer is the most abundant species by number.
  4. \( \overline{M}_w/\overline{M}_n = 1.99 \).

(a) 1, 2 and 4 only (b) 1 and 3 only (c) 2, 3 and 4 only (d) 1, 2, 3 and 4

Answer: (a) 1, 2 and 4 only
  1. \( n_1 = 1 - p = 0.01 \); \( w_1 = (1 - p)^2 = 10^{-4} \). Correct.
  2. The maximum of \( w_x = x(1 - p)^2p^{x-1} \) is at \( x = -1/\ln p = 99.5 \). Correct.
  3. By number the monomer is always the most abundant species. Wrong.
  4. Đ = 1 + p = 1.99. Correct.

Practice Q7. Three fractions of a polymer are isolated: 1 mol at 20,000, 2 mol at 50,000 and 1 mol at 100,000 g mol⁻¹. A viscometric measurement in a solvent with Mark–Houwink exponent \( a = 0.5 \) is made on the whole sample. Which set of averages (g mol⁻¹) is correct?

(a) \( \overline{M}_n = 55{,}000 \), \( \overline{M}_v = 66{,}700 \), \( \overline{M}_w = 70{,}000 \) (b) \( \overline{M}_n = 55{,}000 \), \( \overline{M}_v = 72{,}000 \), \( \overline{M}_w = 70{,}000 \) (c) \( \overline{M}_n = 70{,}000 \), \( \overline{M}_v = 66{,}700 \), \( \overline{M}_w = 81{,}700 \) (d) \( \overline{M}_n = 57{,}000 \), \( \overline{M}_v = 61{,}000 \), \( \overline{M}_w = 70{,}000 \)

Answer: (a)

\( \overline{M}_n = 2.2 \times 10^5/4 = 55{,}000 \); \( \overline{M}_w = 1.54 \times 10^{10}/2.2 \times 10^5 = 70{,}000 \) (Example 3).

Weight fractions: 0.0909, 0.4545, 0.4545. \( \sum w_iM_i^{0.5} = 0.0909(141.4) + 0.4545(223.6) + 0.4545(316.2) = 258.2 \); squaring gives \( \overline{M}_v = 66{,}700 \) g mol⁻¹, between \( \overline{M}_n \) and \( \overline{M}_w \). Option (b) puts \( \overline{M}_v \) above \( \overline{M}_w \), impossible for \( a < 1 \). Option (c) lists \( \overline{M}_w \) and \( \overline{M}_z \) in the n and w slots.

Quick Revision Table

QuantityFormulaMeasured by
\( \overline{M}_n \)\( \sum N_iM_i/\sum N_i = 1/\sum(w_i/M_i) \)osmometry, end groups, VPO, MALDI
\( \overline{M}_w \)\( \sum N_iM_i^2/\sum N_iM_i = \sum w_iM_i \)light scattering, sedimentation equilibrium
\( \overline{M}_z \)\( \sum N_iM_i^3/\sum N_iM_i^2 \)sedimentation equilibrium
\( \overline{M}_v \)\( (\sum w_iM_i^a)^{1/a} \)dilute-solution viscometry
Order\( \overline{M}_n < \overline{M}_v < \overline{M}_w < \overline{M}_z \)equality only if uniform
DispersityĐ = \( \overline{M}_w/\overline{M}_n \); Đ − 1 = \( \sigma_n^2/\overline{M}_n^2 \)—
Most probable\( n_x = (1 - p)p^{x-1} \); \( X_n = 1/(1 - p) \); Đ = 1 + pstep growth; radical (disproportionation)
PoissonĐ ≈ 1 + 1/\( X_n \)living polymerisation
Radical, combinationĐ = 1.5low conversion
Blend\( \overline{M}_n \): harmonic by mass; \( \overline{M}_w \): arithmetic by mass—

Key Takeaways

What’s Next