Light Scattering & Dilute Solution Viscometry (Zimm Plots, Mark-Houwink)

When an electromagnetic wave passes through a transparent dielectric medium, its oscillating electric field induces periodic polarization in the electron clouds of the constituent molecules. These oscillating electric dipoles act as secondary antennas, reradiating electromagnetic energy in all directions at the identical frequency—a phenomenon known as elastic light scattering.

Rayleigh Scattering by Small Particles ($d \ll \lambda / 20$)

Lord Rayleigh derived the intensity of light scattered by an isolated isotropic particle whose dimensions are much smaller than the wavelength of incident radiation ($d < \lambda / 20 \approx 20 - 30\text{ nm}$). For unpolarized incident light of intensity $I_0$ and vacuum wavelength $\lambda_0$, the scattered intensity $I_\theta$ observed at distance $r$ and angle $\theta$ is:

where $\alpha$ is the molecular polarizability and $\theta$ is the scattering angle relative to the incident beam direction. The characteristic Rayleigh ratio $R_\theta$ normalizes the scattered intensity for geometric distance and incident beam intensity:

For pure liquids and solutions, perfect destructive interference would completely cancel the scattered light in all non-forward directions if the molecules were arranged in a perfectly homogeneous or periodic lattice. Scattering in homogeneous liquids arises exclusively from microscopic, spontaneous thermal fluctuations.

Concentration Fluctuations in Polymer Solutions

Albert Einstein (1910) and Peter Debye (1944) showed that excess scattering from a polymer solution over pure solvent ($\Delta R_\theta = R_{\theta, \text{solution}} - R_{\theta, \text{solvent}}$) originates from local concentration fluctuations $\langle (\delta c)^2 \rangle$ within microscopic volume elements $\delta V$. The mean-square concentration fluctuation is governed by the second derivative of the Gibbs free energy of mixing, which is directly linked to the osmotic pressure gradient:

Because fluctuations in concentration produce proportional fluctuations in dielectric permittivity $\delta \epsilon = 2 n (dn/dc) \delta c$, integrating over volume yields the excess Rayleigh ratio:

Defining the universal optical contrast constant $K$:

where:

• $n_0$ is the refractive index of the pure solvent.
• $dn/dc$ is the specific refractive index increment of the polymer in that solvent (typically $0.05 - 0.20\text{ mL/g}$).
• $\lambda_0$ is the vacuum wavelength of the laser source (e.g., $632.8\text{ nm}$ for He-Ne).
• $N_A$ is Avogadro's number ($6.022 \times 10^{23}\text{ mol}^{-1}$).

Using the osmotic virial equation $\frac{\partial \Pi}{\partial c} = R T \left( \frac{1}{M} + 2 A_2 c + \dots \right)$, substitution yields:

This landmark equation demonstrates that measuring excess scattered light intensity as a function of polymer concentration provides an absolute, calibration-free determination of molecular weight and the second virial coefficient.

Intraparticle Interference

When the physical dimensions of a macromolecule exceed approximately $\lambda / 20$ (typically $> 25\text{ nm}$ for visible light), different segments within the same macromolecule no longer scatter light in phase. Consider two scattering segments $i$ and $j$ separated by vector $\mathbf{r}_{ij}$ within a single polymer chain:

• Rays scattered in the exact forward direction ($\theta = 0^\circ$) traverse identical optical path lengths; hence their electric fields interfere constructively without any phase difference.
• At any non-zero scattering angle ($\theta > 0^\circ$), the optical path difference $\Delta s$ between rays scattered from segment $i$ and segment $j$ produces a phase difference $\phi_{ij} = \mathbf{q} \cdot \mathbf{r}_{ij}$, where $\mathbf{q}$ is the scattering vector:

Destructive interference between rays scattered from different parts of the coil attenuates the scattered intensity at higher angles.

The Particle Scattering Factor $P(\theta)$

To account for this intraparticle phase cancellation, Debye introduced the dimensionless particle scattering factor (or form factor) $P(\theta)$:

For a macromolecule composed of $N$ identical scattering segments:

Averaging over all random spatial orientations in an isotropic solution:

yielding the general Debye scattering formula:

At $\theta = 0^\circ$, $q = 0$, so $\sin(q r_{ij}) / (q r_{ij}) \to 1$ and $P(0) = 1.0$ identically for all particles regardless of shape or mass.

Guinier Expansion at Small Scattering Vectors ($q R_g < 1$)

At small scattering angles such that $q r_{ij} \ll 1$, we expand the cardinal sine function in a Taylor series:

Substituting this expansion into the Debye double sum:

By definition of the radius of gyration $R_g$ for an assembly of $N$ identical mass elements:

Therefore:

Substituting into the expansion yields the fundamental Guinier approximation:

Remarkably, this low-angle limiting expansion is completely model-independent—it holds identically for spheres, rods, random coils, and branched dendrimers!

The Reciprocal Form

Because light scattering data is analyzed reciprocally, we invert $P(\theta)$:

Debye Scattering Function for Gaussian Random Coils

For an unperturbed Gaussian flexible polymer coil obeying random walk statistics, Debye integrated the intersegment distance distribution to obtain the closed-form analytical expression:

where $u = q^2 R_g^2 = \frac{16 \pi^2 n^2 R_g^2}{\lambda_0^2} \sin^2\left(\frac{\theta}{2}\right)$.

• When $u \ll 1$: $P(u) \approx 1 - u/3$, reproducing the universal Guinier expansion.
• When $u \gg 1$ (high angle / high $q$): $e^{-u} \to 0$, so $P(u) \to 2/u = 2 / (q^2 R_g^2)$. A plot of $q^2 I(q)$ vs $q$ (Kratky plot) plateaus to a constant value, providing an experimental fingerprint of Gaussian coil topology.

Combining the interparticle osmotic virial expansion with the intraparticle form factor $P(\theta)$ yields the master equation of static light scattering:

Substituting the low-angle expansion $\frac{1}{P(\theta)} = 1 + \frac{16 \pi^2 n^2 R_g^2}{3 \lambda_0^2} \sin^2\left( \frac{\theta}{2} \right)$ yields:

This equation contains three fundamental macromolecular parameters:

1. $M_w$: The weight-average molecular weight.

2. $R_g$: The z-average radius of gyration.

3. $A_2$: The second virial coefficient.

The Zimm Double Extrapolation Construction

In 1948, Bruno Zimm devised an elegant graphical methodology to resolve these three parameters simultaneously from experimental data collected at multiple angles $\theta$ and multiple concentrations $c$. To separate overlapping curves, Zimm plotted $\frac{K c}{\Delta R_\theta}$ on the y-axis against a composite x-axis:

where $k'$ is an arbitrary plotting scale factor (typically $k' = 100\text{ to }1000\text{ cm}^3/\text{g}$) chosen to space out the concentration lines.

The Dual Extrapolations:

1. Extrapolation to Zero Angle ($\theta \to 0$):

• For each fixed concentration $c$, the angular data is extrapolated to $\theta = 0$ (where $\sin^2(\theta/2) = 0$).
• Along this zero-angle envelope line:

• The slope of this line with respect to $k' c$ is:

2. Extrapolation to Zero Concentration ($c \to 0$):

• For each fixed angle $\theta$, the concentration data is extrapolated to $c = 0$.
• Along this zero-concentration envelope line:

• The initial slope with respect to $\sin^2(\theta/2)$ is:

3. Shared Common Intercept:

Both envelope lines converge to the exact same y-intercept at $\theta = 0, c = 0$:

Light scattering yields strictly the weight-average molecular weight ($M_w$) because the scattering intensity of an isolated coil is proportional to its mass squared ($I \propto M^2$), and normalizing by concentration ($c \propto M$) gives $\langle M^2 \rangle / \langle M \rangle = M_w$.

Whereas Static Light Scattering (SLS) measures time-averaged scattering intensity to yield $M_w, R_g$, and $A_2$, Dynamic Light Scattering (DLS)—also known as Photon Correlation Spectroscopy (PCS) or Quasi-Elastic Light Scattering (QELS)—monitors fast microsecond fluctuations in scattered light intensity caused by Brownian diffusion.

Intensity Fluctuations and the Autocorrelation Function

Macromolecules in solution undergo continuous Brownian motion. As their spatial positions change, the relative phase differences between scattered light rays fluctuate, creating a flickering speckle pattern at the photodetector. The temporal decay of these fluctuations is quantified by the normalized second-order intensity autocorrelation function $g^{(2)}(\tau)$:

where $\tau$ is the delay time. According to the Siegert relation, for Gaussian optical fields:

where $\beta \le 1$ is an instrument coherence factor and $g^{(1)}(\tau)$ is the normalized first-order electric field correlation function.

The Diffusion Decay Rate $\Gamma$

For a monodisperse suspension of diffusing particles:

The decay rate $\Gamma$ is proportional to the translational diffusion coefficient $D$:

where $q = \frac{4 \pi n}{\lambda_0} \sin(\theta/2)$. Measuring $\Gamma$ at multiple scattering angles confirms Brownian diffusion when a plot of $\Gamma$ vs $q^2$ passes through the origin with slope $D$.

The Stokes-Einstein Equation and Hydrodynamic Radius ($R_h$)

From the translational diffusion coefficient at infinite dilution ($D_0$), the hydrodynamic radius $R_h$ is calculated using the Stokes-Einstein equation:

where:

• $k_B$ is Boltzmann's constant ($1.38065 \times 10^{-23}\text{ J/K}$).
• $\eta_0$ is the dynamic shear viscosity of pure solvent.
• $R_h$ is the radius of an equivalent hard sphere undergoing the same frictional translation as the solvated macromolecule.

Geometric Conformation Ratio: $R_g / R_h$

The dimensionless ratio $\rho = R_g / R_h$ provides profound diagnostic insight into macromolecular topology:

• Uniform Hard Sphere: $R_g = \sqrt{3/5} R_h \approx 0.775 R_h \implies R_g / R_h = 0.775$
• Gaussian Random Coil in $\Theta$-Solvent: $R_g / R_h = \frac{3 \sqrt{\pi}}{8} \approx 1.504$
• Flexible Coil in Good Solvent: $R_g / R_h \approx 1.78$
• Rigid Rod (needle-like): $R_g / R_h > 2.0$ (diverges with aspect ratio as $\ln(L/d)$).

Dilute solution viscometry is the most widely practiced laboratory technique for polymer molecular weight characterization due to its high precision, simplicity, and low equipment cost.

Capillary Flow and the Hagen-Poiseuille Law

In a capillary viscometer (e.g., Ostwald or Ubbelohde viscometer), liquid drains through a precision glass capillary of radius $R$ and length $L$ under its own hydrostatic head. According to the Hagen-Poiseuille equation for laminar flow:

where $t$ is the efflux time, $\rho$ is liquid density, and $V$ is the bulb volume. For dilute solutions ($c < 10\text{ g/L}$), the solution density $\rho$ is essentially indistinguishable from the solvent density $\rho_0$ ($\|\rho - \rho_0\| / \rho_0 < 0.2\%$). Thus, the ratio of viscosities equals the ratio of efflux times:

Standard Viscometric Definitions

1. Relative Viscosity (Viscosity Ratio):

2. Specific Viscosity: The fractional increase in viscosity attributable to the dissolved macromolecular solute:

3. Reduced Viscosity (Viscosity Number): The specific viscosity normalized per unit solute concentration:

(Standard units: $\text{dL/g}$ or $\text{cm}^3/\text{g}$ or $\text{mL/g}$).

4. Inherent Viscosity (Logarithmic Viscosity Number):

Ubbelohde vs Ostwald Viscometers

The Ubbelohde suspended-level viscometer is universally preferred over Ostwald designs because it features an open venting side-arm that isolates the capillary pressure head from total liquid volume in the reservoir. Consequently, serial dilutions can be performed directly inside the viscometer cell without emptying, cleaning, or recalibrating between runs.

Intrinsic Viscosity $[\eta]$

As concentration approaches zero, interchain hydrodynamic and thermodynamic interactions vanish. The intrinsic viscosity $[\eta]$ (also called the Limiting Viscosity Number, LVN) represents the isolated hydrodynamic volume increment imparted by a single macromolecule per unit mass:

Note that despite being named 'viscosity', $[\eta]$ has dimensions of reciprocal density or specific volume ($[\eta] \sim \text{Volume} / \text{Mass}$, typically expressed in $\text{dL/g}$ or $\text{cm}^3/\text{g}$).

The Huggins Equation

Expanding the reduced viscosity $\eta_{\text{sp}}/c$ as a Taylor power series in concentration $c$:

where $k_H$ is the dimensionless Huggins constant.

• In thermodynamically good solvents: polymer coils are well-solvated, and interchain segment collisions are minimal, yielding $k_H \approx 0.30 - 0.40$.
• In thermodynamically poor solvents (near theta conditions): polymer-polymer attraction increases segment clustering, driving $k_H \approx 0.50 - 0.80$.
• Values of $k_H > 1.0$ indicate extensive multimolecular aggregation or microgel formation.

The Kraemer Equation

Expanding the inherent viscosity $\ln(\eta_{\text{rel}})/c$ using the series $\ln(1 + x) = x - x^2/2 + \dots$:

Dividing by $c$ and substituting $\eta_{\text{sp}} = [\eta]c + k_H [\eta]^2 c^2$:

where $k_K$ is the dimensionless Kraemer constant. Comparing terms reveals the mathematical identity:

In practice, experimental data are analyzed by simultaneously plotting both $\eta_{\text{sp}}/c$ (upward slope) and $(\ln \eta_{\text{rel}})/c$ (downward slope) on the same graph against concentration $c$. Both lines must extrapolate to the identical y-intercept at $c = 0$, guaranteeing experimental rigor.

The Mark-Houwink-Sakurada (MHS) Empirical Relation

In 1938–1940, Herman Mark, Roelof Houwink, and Ichiro Sakurada established the empirical power-law relationship between intrinsic viscosity $[\eta]$ and molecular weight:

where:

• $K$ is the Mark-Houwink pre-exponential constant (typically $10^{-4} - 10^{-2}\text{ dL/g}$).
• $a$ is the Mark-Houwink conformational exponent (dimensionless).
• $M_v$ is the viscosity-average molecular weight.

Both $K$ and $a$ are specific to a given polymer-solvent-temperature triplet and are tabulated in chemical handbooks. Taking the natural logarithm yields a linear calibration equation:

Physical Origin: The Flory-Fox Hydrodynamic Equation

Paul Flory and Thomas Fox demonstrated that the intrinsic viscosity of a flexible polymer coil is proportional to its hydrodynamic volume per unit mass:

where $\Phi_0$ is the universal Flory hydrodynamic constant ($\Phi_0 \approx 2.5 \times 10^{23}\text{ mol}^{-1}$ when $[\eta]$ is in $\text{cm}^3/\text{g}$), and $\langle R^2 \rangle^{1/2}$ is the root-mean-square end-to-end distance. Expressing the end-to-end distance as $\langle R^2 \rangle = \alpha^2 \langle R_0^2 \rangle$:

Physical Meaning of the Mark-Houwink Exponent $a$

The value of the exponent $a$ reveals the three-dimensional hydrodynamic conformation of the macromolecule in that solvent:

1. $a = 0$ (Hard Solid Spheres):

• Einstein's viscosity law: $\eta_{\text{sp}} = 2.5 \phi = 2.5 c \bar{v} \implies [\eta] = 2.5 \bar{v} = \text{constant}$ (independent of $M$).
• Examples: Globular proteins (myoglobin, hemoglobin), hyperbranched dendrimers, compact latex nanospheres.

2. $a = 0.50$ (Random Coil in $\Theta$-Solvent):

• The coil is unperturbed ($lpha = 1$). $[\eta] = K_\theta M^{1/2}$.
• Flory theta conditions (e.g., Polystyrene in cyclohexane at $34.5^\circ\text{C}$).

3. $a = 0.65 - 0.80$ (Flexible Random Coil in Good Solvent):

• The coil is thermodynamically swollen by excluded volume ($lpha \propto M^{0.1}$).
• Examples: Polystyrene in toluene ($a = 0.72$), PMMA in chloroform ($a = 0.76$).

4. $a = 1.0 - 1.2$ (Semi-Rigid Wormlike Chain / Extended Helix):

• Poly(gamma-benzyl-L-glutamate) in helicogenic solvents, sodium hyaluronate.

5. $a = 1.7 - 2.0$ (Rigid Inflexible Rod):

• Stiff cylinders rotating in shear flow.
• Examples: Native double-stranded DNA, poly(p-phenylene terephthalamide) (Kevlar) in concentrated $\text{H}_2\text{SO}_4$, tobacco mosaic virus.