Dissolution · Release kinetics
Dissolution kinetic models: equations, interpretation and how to choose
Zero-order, first-order, Higuchi, Korsmeyer–Peppas, Weibull and Hixson–Crowell are the six models most often fitted to drug-release data. This guide gives each equation in the form used on the original percentage scale, what its parameters mean, and how to compare fits without over-reading R².
Fit all six models to your data Read the fitting conventions
The six models
Q is the cumulative percentage released at time t; k is the model's rate constant.
| Model | Equation | Typical use |
|---|---|---|
| Zero-order | Q = k0 · t | Constant-rate release, e.g. some osmotic or reservoir systems |
| First-order | Q = 100 · (1 − e^(−k1·t)) | Release proportional to the amount remaining; many immediate-release and porous matrices |
| Higuchi | Q = kH · √t | Diffusion from an insoluble matrix (Higuchi, 1963) |
| Korsmeyer–Peppas | Q = kKP · t^n | Distinguishing release mechanisms from the exponent n; fitted to the early part of the curve (typically Q ≤ 60 %) |
| Weibull | Q = 100 · (1 − exp(−(t/td)^β)) | Flexible empirical description of the whole curve (Langenbucher, 1972) |
| Hixson–Crowell | 1 − (1 − Q/100)^(1/3) = kHC · t | Release limited by a shrinking particle surface (cube-root law) |
Interpreting the Korsmeyer–Peppas exponent n
For non-swelling systems, Ritger and Peppas (1987) gave geometry-dependent limits:
| Geometry | Fickian diffusion | Anomalous (non-Fickian) | Case II transport |
|---|---|---|---|
| Thin film (slab) | n = 0.50 | 0.50 < n < 1.00 | n = 1.00 |
| Cylinder | n = 0.45 | 0.45 < n < 0.89 | n = 0.89 |
| Sphere | n = 0.43 | 0.43 < n < 0.85 | n = 0.85 |
The exponent is only meaningful within the fitted window and for the geometry assumed; it is evidence about mechanism, not proof of it.
Comparing fits: why R² is not enough
- R² always rises when a model gains parameters, so a two-parameter Weibull fit will often beat a one-parameter model on R² alone.
- Adjusted R² penalises extra parameters and is safer for ranking models fitted to the same data.
- AIC (here N·ln(SSE/N) + 2p, constants omitted) and the model selection criterion (MSC) balance fit against complexity. Compare AIC only between models fitted to identical observations — Korsmeyer–Peppas uses a restricted window, so its AIC is not directly comparable with full-profile models.
- Always inspect residuals: a systematic pattern means the model shape is wrong, whatever the R².
A practical workflow
- Fit the mean profile and, separately, each replicate — the mean of replicate parameters is not the same as the parameters of the mean curve.
- Rank models by adjusted R² and AIC on the same observations; check residual plots.
- Fit Korsmeyer–Peppas to Q ≤ 60 % and interpret n against the dosage-form geometry.
- Report the equation, fitting window, weighting (SCIOVANTA fits unweighted on the original scale), parameters with units, and the criteria used to choose.
Model fitting describes the data; it does not by itself establish a release mechanism, clinical performance or regulatory acceptability.
Worked numbers on this page are reproduced exactly by the SCIOVANTA calculators linked above. Acceptance criteria differ between regulators and products; check the current guidance that applies to your submission.
References
- Costa P, Sousa Lobo JM (2001). Modeling and comparison of dissolution profiles. European Journal of Pharmaceutical Sciences 13(2):123–133.
- Higuchi T (1963). Mechanism of sustained-action medication: theoretical analysis of rate of release of solid drugs dispersed in solid matrices. Journal of Pharmaceutical Sciences 52:1145–1149.
- Korsmeyer RW, Gurny R, Doelker E, Buri P, Peppas NA (1983). Mechanisms of solute release from porous hydrophilic polymers. International Journal of Pharmaceutics 15:25–35.
- Ritger PL, Peppas NA (1987). A simple equation for description of solute release I. Fickian and non-Fickian release from non-swellable devices in the form of slabs, spheres, cylinders or discs. Journal of Controlled Release 5:23–36.
- Langenbucher F (1972). Linearization of dissolution rate curves by the Weibull distribution. Journal of Pharmacy and Pharmacology 24:979–981.
- Hixson AW, Crowell JH (1931). Dependence of reaction velocity upon surface and agitation. Industrial & Engineering Chemistry 23:923–931.
- Zhang Y, Huo M, Zhou J, et al. (2010). DDSolver: an add-in program for modeling and comparison of drug dissolution profiles. The AAPS Journal 12(3):263–271.
Last updated 2026-11-17.