(638d) Dynamics of Diffusivity and Pressure Drop In Flow-Through and Parallel Flow Bioreactors During Tissue Regeneration

Utilization of bioreactors to regenerate tissues outside the body has been intensely investigated in functional tissue engineering.  Different configurations of bioreactors have been explored to regenerate tissues including flow-through (also referred as transverse flow), parallel-flow, rotating wall, and microgravity.  For regenerating large tissues, flow-through and parallel-flow configurations are preferred due to uniform support offered to the scaffold.  However, transport characteristics in high aspect ratio scaffolds (for example, 100 mm diameter and 2 mm thick) during tissue regeneration are not well understood. 

In this study, two designs i) flow-through (Design 6) and ii) parallel flow (Design 8) reactors were utilized to analyze the effect of tissue development.  First both designs were simulated using COMSOL 3.5 Multiphysics (COMSOL, Inc., Burlington, MA) with properties of 2%-2% chitosan-gelatin porous structures, similar to previous publication [1].  The fluid flow was defined by Brinkman equation on the porous regions using pore characteristics of Chitosan-Gelatin scaffold formed at -80°C (55 µm and 318 pores/mm²).  These results were then compared with experiments performed using similar dimension reactors that could hold 100 mm diameter and 2 mm porous structures.  Pressure drop values across the reactor were calculated at flow rates 5 mL/min to 25 mL/min with 5 mL/min increment.  At low flow rates experimental results were in agreement with the simulation results for both reactor configurations.  However, increased flow rate in Design 6, showed channeling at circumference of the reactor leading to lower pressure drop relative to simulation. 

Since effective diffusivity data for oxygen and glucose through different chitosan gelatin porous structures was unavailable in literature, effective diffusivity of glucose was experimentally determined using custom-built apparatus.  The pore size of scaffolds with different composition was also obtained from experiments using scanning electron micrographs.  The experimental results were then used for computer simulation, to study nutrient distribution and consumption by smooth muscle cells across the scaffold for both reactors.  To understand whether nutrient transfer is limited by diffusion or convection, Peclet number inside the scaffold were determined.  Velocities obtained from simulations along with diffusivity values from experiment were used to calculate Peclet number for both reactors.  These results showed that Design 6 is convective flow dominant whereas Design 8 is diffusion-dependent.  In Design 6, highest flow rate that can be used depend on mechanical strength of the porous scaffold.  While in Design 8, reduced porosity (<20%) significantly restricted nutrient distribution.  In order to enhance nutrient distribution, additional inlets and outlets were included into the reactor.  These new designs improved nutrient distribution for 2 mm scaffolds with variable cell densities.  However, it was found that parallel-flow configuration is not suitable for porous structures greater than 4 mm thick scaffolds, especially at very low porosities. 

References

[1] Devarapalli M, Lawrence BJ, Madihally SV.  Modeling Nutrient Consumptions in Large Flow-Through Bioreactors for Tissue Engineering.  Biotechnology/ Bioengineering. 103(5):1003-1015, 2009.