Microfluidic Mixing and Reaction  Home
Theory & Design

Results Manufacture
Non Reactive Mixing
Electrode Integration
Reactive Mixing

The saponification of ethyl acetate is carried out in the serpentine structure. As the reaction progresses, sodium hydroxide is consumed and sodium acetate is produced in a 1:1 stoichiometric ratio. Because ethyl acetate and ethanol are not electrically conductive, conductivity measurements can be used to measure the extent of this reaction. To construct a curve that could be used to relate conversion of sodium hydroxide to the measured conductance, the dependence of conductivity on concentrations of both sodium hydroxide and sodium acetate had to be determined. Aqueous solutions of sodium hydroxide from 0.00 M (deionized water) to 0.05 M were prepared. The conductance of each solution was measured. It was determined that the conductance of the sodium hydroxide solutions varied linearly with concentration of sodium hydroxide. The data points were plotted, and an equation relating to concentration of sodium hydroxide was determined. Conductance is used for ease of measurement in the experiments. The cell constant of the probe was measured to be 0.137 cm-1 and can be used to convert experimental valued to terms of conductivity if necessary. To prepare solutions of sodium acetate, equal amounts of sodium hydroxide and ethyl acetate were mixed and allowed to stand overnight for complete reaction to occur. Once it was confirmed that the conductivity was consistent over time, the conductance of these solutions were measured. Conductance and concentration of sodium acetate were plotted against each other, and a linear fit was determined. The plots of the sodium hydroxide solutions and sodium acetate solutions are reproduced below.

Calibration Curve
Conductance as a function of ionic species concentration in the saponification of ethyl acetate.

Using these results, a relationship between conversion of sodium hydroxide at the end of the channel and the conductance measured at the end of the channel could be related. By expressing sodium acetate concentration in terms of sodium hydroxide concentration, an equation is derived:

Conversion in terms of Conductance and Initial Concentation
Conversion in terms of Conductivity and Initial Concentation

The initial concentration term in the denominator on the right hand side must account for the effective dilution of the concentration upon mixing the sodium hydroxide stream with the ethyl acetate stream.

After seeing the performance of the chips during nonreactive mixing experiments, the serpentine structure was determined to be the best performing structure. Noting that the mixing performance of the serpentine chip increased with increasing Reynolds number, a similar trend with conversion was expected for reactive mixing. As more mixing occurred, the conversion should increase.However, the figures below indicate a different trend.

Results from reactive mixing in terms of conductance.  Note the scale of the vertical axis, starting at 0.004 S.

Results in terms of conversion of sodium hydroxide.  Note the scale and bounds of the vertical axis.

The data indicate that conversion of sodium hydroxide decreases with increasing Reynolds number. One possible explanation for the apparent discrepancy is that the mixing performance of the serpentine chip may be so effective that the limiting factor in conversion becomes reaction rate. This would explain why the conversion decreases with increasing Reynolds number, since residence time is also decreasing. Additionally, the figure above seems to indicate a nonlinear dependence of conversion on Reynolds number.

A maximum may be present at some Reynolds number lower than the Reynolds number for the left most data point (at Re = 7). With these particular reaction conditions, there may be an optimum point which represents a balance between mixing performance of this chip and the reaction kinetics. To verify this alternative hypothesis relating mixing and reaction, a more detailed analysis in the low Reynolds number range (Re<10) would be required.
Attempts to use the pH indicator, indigo carmine, to characterize both mixing and relative extent of reaction in a series of tests proved impossible. As mixing occurs in the channel the indigo carmine indicator becomes too faint to use the Vision software. Additionally problems with decay of the indicator were apparent. Over time the indicator would lose its pH dependent color properties thus making measurements with it difficult.