Experimental breakthrough data of L-phenylalanine were obtained from cation-exchange columns over a wide range of feed concentrations and flow rates. Simulations based on three rate models were used to analyze the data. The models include three different intraparticle diffusion mechanisms : pore diffusion, surface diffusion, or parallel pore and surface diffusion, respectively. Simulated column breakthrough curves were compared to data from both packed beds and fluidized beds. Both the pore diffusion and surface diffusion models gave reasonable fits for breakthrough curves; however, these fits required that the apparent surface diffusivities obtained from the surface diffusion model increase with increasing concentration, and the apparent pore diffusivities obtained from the pore diffusion model decrease with increasing concentration and are larger than Brownian diffusivity. These results suggest that both pore diffusion and surface diffusion are important intraparticle transport mechanisms. A systematic method was developed to estimate the pore diffusivity and the surface diffusivity independently, which were used in the parallel diffusion model for data analysis. The parallel diffusion model can accurately simulate the data over a wide range of concentrations and flow rates using concentration-independent pore and surface diffusivities. The parallel diffusion model with diffusivities estimated from packed-bed data can successfully predict the data for fluidized beds with large (>2) length to diameter ratios. Correlations for mass-transfer parameters in the bulk phase of expanded or fluidized beds were successful for simulation of breakthroughs from columns with large length to diameter ratios.