Chemical engineering transport & equilibrium foundations of MSMPR Continuous Crystallization.
Industrial crystallization from solution is a primary separation and purification unit operation in chemical, fertilizer, pharmaceutical, and specialty materials manufacturing. In continuous crystallization, product crystal size distribution (CSD) directly determines downstream filtration cake porosity, washability, drying kinetics, pneumatic conveying flowability, and bulk dissolution rates. The benchmark theoretical framework for continuous industrial crystallization is the Mixed-Suspension Mixed-Product-Removal (MSMPR) crystallizer formalized by Alan Randolph and Maurice Larson in 1971. In an ideal MSMPR crystallizer, steady-state continuous feed enters an intensely agitated vessel (mean residence time τ = V / v_0 = 40.0 min), supersaturation is generated continuously (via cooling or solvent evaporation), crystals nucleate and grow throughout the suspension, and the exit slurry has an identical composition and particle size distribution to the magma inside the vessel. Applying the macroscopic one-dimensional Population Balance Equation (PBE) under steady-state conditions with size-independent linear crystal growth (McCabe's ΔL law, dG/dL = 0) and negligible crystal breakage or agglomeration yields: d(n · G)/dL + n / τ = 0, where n(L) is the crystal population density (number of crystals per unit size per unit slurry volume, no./(μm·mL)), G is the linear crystal growth rate (dL/dt, μm/min), and τ is the crystallizer residence time (min). Integrating this first-order ODE subject to boundary condition n(0) = n_0 = B_0 / G (where n_0 is nuclei population density and B_0 is primary nucleation rate, no./(min·mL)) gives the exponential population density distribution: n(L) = n_0 · exp[-L / (G · τ)]. Taking natural logarithms linearizes the MSMPR population model: ln[n(L)] = -S · L + c_0, where the population decay slope is S = 1 / (G · τ) = 0.0125 μm^-1 and the intercept is c_0 = ln(n_0) = 15.2000. With known residence time τ = 40.0 min, the linear crystal growth rate is evaluated directly as G = 1 / (S · τ) = 2.0000 μm/min. In this laboratory experiment, a 4.0-liter continuous cooling MSMPR crystallizer precipitating potassium sulfate (K_2SO_4) from aqueous solution at 25.0 °C is operated at steady state to sample crystal slurries across size fractions from 50.0 to 300.0 μm, verify the log-linear population decay, and extract G = 2.0000 μm/min and B_0 = 7.97 × 10^6 no./(min·mL).
Population Balance Equation & The MSMPR Crystallizer
In particulate unit operations, the number distribution of particles is governed by the population balance framework formulated by Randolph and Larson (1971). For a mixed-suspension mixed-product-removal (MSMPR) crystallizer operating at steady state with slurry volume $V$ and volumetric withdrawal flow rate $v_0$ (mean residence time $\tau = V / v_0$), the one-dimensional population balance equation along characteristic size coordinate $L$ is:
\[ \frac{\partial n}{\partial t} + \frac{\partial(G \, n)}{\partial L} + D(L) - B(L) = \frac{n_{\text{in}} - n}{\tau} \]where $n(L)$ is the crystal population density ($ ext{no.}/(\mu\text{m}\cdot\text{mL})$), $G(L) = dL / dt$ is the linear crystal growth rate, and $B(L)$ and $D(L)$ represent particle birth and death due to breakage and agglomeration. Under standard ideal MSMPR assumptions:
- Steady-state operation: $\partial n / \partial t = 0$.
- Clear liquid feed: $n_{\text{in}} = 0$.
- McCabe's $\Delta L$ law (size-independent growth): $G(L) = G = \text{const}$.
- Negligible breakage and agglomeration: $B(L) = D(L) = 0$.
The population balance simplifies to the first-order differential equation:
\[ G \, \frac{d n}{d L} + \frac{n}{\tau} = 0 \implies \frac{d n}{d L} = -\frac{n}{G \, \tau} \]Exponential Distribution & Semi-Log Linearization
Integrating from $L = 0$ where population density is the nuclei density $n(0) = n_0 = B_0 / G$ (with $B_0$ being the primary nucleation rate in $ ext{no.}/(\text{min}\cdot\text{mL})$):
\[ \int_{n_0}^n \frac{d n}{n} = -\int_0^L \frac{d L}{G \tau} \implies n(L) = n_0 \, \exp\left( -\frac{L}{G \, \tau} \right) \]Taking the natural logarithm yields the linear diagnostic equation:
\[ \ln[n(L)] = \ln(n_0) - \left( \frac{1}{G \, \tau} \right) \cdot L = c_0 - S \cdot L \]where the slope magnitude is $S = \frac{1}{G \, \tau} = 0.0125\ \mu\text{m}^{-1}$ and the intercept is $c_0 = \ln(n_0) = 15.2000$.
Kinetic Parameter Extraction
From the linear regression parameters, the fundamental crystallization kinetics are evaluated directly:
\[ G = \frac{1}{S \cdot \tau} = \frac{1}{0.0125 \cdot 40.0} = 2.0000\ \mu\text{m/min} \] \[ n_0 = e^{c_0} = e^{15.2000} = 3.9928 \times 10^6\text{ no./(}\mu\text{m}\cdot\text{mL)} \] \[ B_0 = n_0 \cdot G = (3.9928 \times 10^6) \times 2.0000 = 7.9856 \times 10^6\text{ no./(min}\cdot\text{mL)} \]Viva voce questions & analytical theory
What is McCabe's ΔL law and when is it valid in crystallization?
McCabe's ΔL law states that all geometrically similar crystals growing in the same supersaturated solution grow at the same linear rate G = dL/dt, independent of crystal size L; it is valid when surface reaction kinetics or uniform boundary layer mass transfer dominates, without size-dependent growth dispersion.
What is the physical meaning of population density n(L)?
Population density n(L) = dN/dL is the differential number of crystals dN per unit size increment dL per unit slurry volume; integrating n(L) dL gives total crystal number density, while multiplying by shape factor k_v L^3 gives crystal volume.
How are nucleation rate B_0 and growth rate G extracted from an MSMPR population plot?
In a semi-log plot of ln[n] versus L, the slope is -1 / (G · τ), from which growth rate G is calculated as G = -1 / (slope · τ); the y-intercept at L = 0 is ln(n_0), from which nuclei density n_0 = exp(intercept) and nucleation rate B_0 = n_0 · G are evaluated.
Why must an MSMPR crystallizer run for 10 to 12 residence times before sampling?
Unlike liquid-phase reactors that reach steady state in 3 to 4 residence times, crystallization involves complex feedback between supersaturation, nucleation, and surface area growth; cycling instabilities and damped CSD waves require 10 to 12 residence times to achieve steady CSD.
What causes non-linear downward or upward curvature in an MSMPR semi-log population plot?
Downward curvature at small sizes indicates size-dependent growth (smaller crystals growing slower) or fines dissolution; upward curvature at large sizes indicates crystal agglomeration or internal classification (non-representative product removal).
What is the difference between primary nucleation and secondary nucleation?
Primary nucleation occurs spontaneously in clear supersaturated solutions in the absence of crystalline seeds (homogeneous or heterogeneous); secondary nucleation is induced by the presence of existing parent crystals through fluid shear or crystal-impeller collisions, dominating industrial MSMPR crystallizers.
Experimental precautions and process safety
- Ensure isokinetic sampling velocity matches draft-tube slurry velocity to prevent selective inertial bias toward fine or coarse crystals.
- Wash sampled crystals immediately with ice-cold saturated ethanol to quench residual liquor and prevent crystallization on the filter paper.
- Verify that draft-tube impeller speed (450 RPM) is high enough to suspend all crystals off the dished bottom without causing mechanical attrition.
- Maintain jacket cooling temperature strictly constant within ±0.05 °C to prevent cyclic oscillations in supersaturation.
Theoretical assumptions & boundary condition limits
- The transport and thermodynamic models apply within the specified calibrated operating window (50.0 to 300.0 μm).
- Assumes ideal continuum flow and homogeneous phase mixing without unmodeled parasitic wall losses.
- Continuous sensor readings reflect ordinary least-squares regression across successive observations.
- High-temperature or extreme concentration excursions require non-ideal equation-of-state corrections.