Figure 1.
Effects of flow velocity and septum width on the transverse profile of the concentration.
The traces show the actual current measured at −100 mV on a recording pipette with an open tip moving at a constant velocity of 0.4 mm/s. (A–B) Concentration profile for an untreated application pipette with a septum width of 10 m at 100 cm/s and 400 cm/s respectively. In the later case, a spike at the boundary between channels indicates that solution from the bath is being dragged into the interface between streams. (C) Concentration profile for an application pipette with the septum thinned to 3 µm with the measures at 400 cm/s flow velocity. The concentration profile is comparable to that in A.
Figure 2.
The measures were made with a recording pipette still while moving the application pipette with the piezoelectric actuator. (A) Triangle shaped driving voltage causing the recording pipette alternate between the two solution streams. (B) Recorded current profiles at −100 mV after crossing between external solution and external solution diluted 1∶10. (C) Higher resolution current profile for the coordinates used in B. The width of the interface was estimated by assuming a linear movement of the piezo actuator of 0.1 µm/V. The broken lines indicate a region were the changes in current are approximately proportional to the changes in position. The solid line indicates a fit to the plotted equation.
Figure 3.
Response of the solution exchanger to a short pulse.
(A) A 20V 20 µs square pulse applied to the piezoelectric actuator at 1 Hz of repetition rate. The recording pipette is positioned at 0.85 µm from the interface. (B) Changes in current generated by the movement in response to the pulse. Dashed line indicates the current expected for a complete solution exchange. Mechanical oscillations of the piezoelectric translated into repeated partial solution exchanges. (C–D) Expanded scale of A and B. (D) Show the onset of the motion. The dotted line marks the position of square voltage command.
Figure 4.
Movements of the piezo and the interface after a short pulse command.
The measures were made optically for the piezo motion and with the recording pipette positioned at the center of the interface. (A) Big amplitude (50V) or a small amplitude (200 mV) 20 µs command square voltages were applied to the piezo. Big amplitude pulses generate movements big enough to be properly measured. Small amplitude pulses confine the movement of the interface to the linear displacement region. (B) Piezo motion in response to a big pulse. (C) Interface movement in response to voltage command. Calibration of the movement was done after data of figure 2C (D) Expanded scale of A. (E–F) Higher resolution views of the movement pattern generated by the pulsed activation of the piezo. Dotted line indicates the time of the pulse command. 100 and 148 traces were averaged for piezo and interface movements respectively.
Figure 5.
Power spectrum and the transfer function of the movement of the piezo alone and of the interface.
Power spectrum (Ys) and the corresponding transfer function of movement of the piezo alone (gp) and of the interface (gs) calculated from the data in Figure 4.
Figure 6.
Response to a voltage command optimized for a pulsed movement.
(A) Shape of the optimal command, calculated on the basis of the response depicted in Figure 4B to obtain a 20 µs pulse. (B) Expanded scale of A. (C) Measurement of the solution exchange, expressed as percentage of a complete exchange, generated by the application of optimized command scaled to have a maximum excursion of 20 V. A clean but incomplete exchange is clear. (D) Expanded scale of C. (E) Measurement of the movement of the interface after the application of the optimal command, but scaled to have a maximum excursion of 250 mV. (F) Expanded scale of E. The measurements were the average of 100–120 traces.
Figure 7.
Power spectrum and the transfer function of movement.
Power spectrum (Ys) and the corresponding transfer function (gs) of movement of the interface calculated from the data in Figure 6.
Figure 8.
Response to a further optimized command.
(A) Shape of the optimal command calculated on the basis on the transfer function depicted in Figure 7. (B) Expanded scale of A. (C) Measure of the solution exchange after the application of the optimal command, expressed as percentage of a complete exchange. A single pulse that crosses almost the entire interface was obtained by applying the optimal command with a maximum excursion of 10 V (D) Expanded scale of C. The width of the pulse is about 25 µs. Average of 150 traces.
Figure 9.
Response to the optimized command for pulses of 100 µs.
(A) Optimal command calculated after two cycles. (B) The same graph with a larger scale. (C) Solution exchange shows a single pulse after amplitude of 10 V. A larger scale is shown in D. The mean duration of the pulse is approximately 100 microseconds. Average of 130 traces.