Kilean’s JMS Paper Figures

ByKilean

Hi All,

This post is more of a dump of figures that I have put together for Dan’s and my upcoming Journal of Membrane Science manuscript. I have included figures and captions, so any feedback would be greatly appreciated.

Figure 1: Development of a COMSOL Model representing a microfluidic tangential flow for analyte capture (TFAC) system. (A) The geometry is modeled to scale for all channels. However, the membrane (red line, inset) is modeled as a 10 μm thick layer for ease of computation. (B) The model allowed for testing various geometry elements, such as channel height, to find the best system geometry for capture. (C) By analyzing the change in velocity profile for increasing supply flow rates, we were able to observe the effect of the porous membrane on flow and thus particle capture. (D) When analyzing the capture of particles on the membrane, the model predicts that it is not the flow rate that is important for capture, but rather flow rate ratio between the supply rate and the ultrafiltration rate.
Figure 2: Particle tracing modeling results in a to scale TFAC system. (A) By counting the number of particles retained on the membrane at the end of a simulation, we can determine the percentage of total particles that have been captured as a function of the flow rate ratio. As expected, we see an increased percentage capture as the flow rate ratio approaches unity. This effect can be observed visually in the simulation, as shown in (B) for a flow rate combination of QS = 100 μL/min and QU = 0 μL/min and (C) for a flow rate combination of QS = 70 μL/min and QU = 60 μL/min.
Figure 3: Experimental validation of computationally predicted capture with 60 nm gold nanoparticles. Five flow ratios were randomly chosen to test the conditions and results predicted by the computational model. These ratios were 0% ultrafiltration (A), 20% ultrafiltration (B), 50% ultrafiltration (C), 65% ultrafiltration (not shown), and 85% ultrafiltration (D). Each condition was run in triplicate and the results (E, orange triangles) were plotted against the computational results (E, blue circles). The experimental results showed poor fit when compared to the simulated results.
Figure 4: Low magnification images of gold nanoparticle aggregation. Gold nanoparticle aggregation was observed on all samples, with size and density increasing with flow rate ratio. This is demonstrated for 0% ultrafiltration (A), 20% ultrafiltration (B), 50% ultrafiltration (C), and 85% ultrafiltration (D).
Figure 5: Experimental validation of computationally predicted capture with 60 nm gold nanoparticles and 1 mg/mL BSA. To prevent nanoparticle aggregation, solutions of 60 nm nanoparticles were made with 1 mg/mL BSA. SEM inspection of the surface showed capture closer to experimental conditions at 0% ultrafiltration (A), 20% ultrafiltration (B), 50% ultrafiltration (C), 65% ultrafiltration (not shown), and 85% ultrafiltration (D). Comparting the experimental results (E, orange triangles) to the computational results (E, blue circles) showed a better fit. This indicates that the discrepancy observed in Figure 3 was likely due to the observed aggregation of nanoparticles.
Figure 6: Low magnification images of aggregate prevention from a 1 mg/mL BSA solution. Gold nanoparticle aggregation was prevented on all samples with the addition of 1 mg/mL to the solution. This is demonstrated for 0% ultrafiltration (A), 20% ultrafiltration (B), 50% ultrafiltration (C), and 85% ultrafiltration (D).

Similar Posts