Supporting Information

An automatic microfluidic sample transfer and introduction system

Kan Liu1,*, Nan-Gang Zhang1, Sheng-Xiang Wang1, Yuliang Deng2,*

1College of Electronic and Electrical Engineering, Wuhan Textile University, Wuhan 430073, P.R.China

2Key Laboratory of Systems Biomedicine(Ministry of Education), Shanghai Center for Systems Biomedicine, Shanghai Jiao Tong University, Shanghai, 200240, P.R.China.

* (Kan Liu); (Yuliang Deng)

Dynamic pressure control for liquid transfer

Because the whole system used pressure-driven module to drive the tiny volume of the sample, to control dynamic pressure for liquid transfer is very important for avoiding loss of liquid (leaving behind residue) during transfer. This is expected to occur as liquid is transferred from the vial to the tubing, and as the liquid slug travels through the tubing. Two experiments have been done in order to find out how much liquid will lose during the transfer and priming process.

One experiment was designed to estimate the loss during the transfer from vial to tubing by measuring of the radioactivity of a dilute F-18 solution (in 1 M K2CO3 in [18O]H2O). The outlet of the liquid transfer tubing was uncoupled from the microfliuidic chip. A known liquid volume of F-18 solution was pipetted into a 100 µL vial insert (Agilent) inside a 1.5 mL vial (Agilent，USA) with septum cap and the radioactivity was measured (CRC-5R，Capintec radioisotope calibrator,USA). Next, the preassembled cap and tubing were installed onto the vial. The septum was pierced with two tubes. One was connected with an electronic pressure regulator, and the other tubing was small-diameter (ID 0.012”OD 0.030”,Cole-Parmer，USA) transfer tubing which was 1 meter length with an angle cut at the end. The pressure was gradually increased from 0 to 3.5 kPa (0.5 kPa/sec) to ensure the liquid transfer from vial into tubing without splitting and leaving too much liquid in the vial. It also ensured that the slug did not break inside the tube later. Then the slug entered and traveled through 1 meter of tube into a waste container. The radioactivities of vial+cap and tube were separatelymeasured. The percentage volume loss from a vial to tubing was computed from the decay-corrected radioactivity measurements.

In the above experiments, we used 0.022"ID x 0.042"OD PTFE microbore tubing for pressure supply and 0.012"ID x 0.030"OD PTFE microbore tubing for transfer liquid slug. (Cole-Parmer)

The other experiment was designed to estimate the amount of deionized water left behind in the tubing. We measured the length of a slug of deionized water before and after traveling through 1 meter of tubing. The outlet of transfer tubing was connected to microfluidic chip. During slug transfer process, Vlatch was closed, Vinlet and Vbleed were opened. At the beginning, the liquid was loaded by the way of mentioned above. Once the liquid was fully loaded into tubing, valve Vinlet at the other end of the tubing was closed and shut down the pressure to stop the slug. Then we measured the length of the slug of deionized water inside the tubing and got a length Lbegin. Next, the valve was opened and the pressure was gradually increased to drive the liquid. When the liquid slug passed through 1 meter distance of tubing, at that point, the slug was stopped again and the length Lend was measured. Finally, the percentage volume loss was calculated from these lengths.

Our main concern was to avoid loss of liquid (leaving behind residue) during transfer from a conventional container/vial. This is expected to occur as liquid is transferred from the vial to the tubing, and as the liquid plug travel through the tubing. These happen normally due to suddenly increase the transfer pressure to break the liquid in the container/vial or inside the tubing and some liquid leave on the surfaces of them. In order to avoid breaking the liquid during the transfer and priming process, an electrical pressure regulator was used to gradually increase the driven pressure. In particular, pressure was dynamically controlled (0.5 kPa/sec,driven pressures in the range of 3-40kPa) were used, depending on type of liquid and volume of slug.) to avoid suddenly pressure changing to avoid fragmenting very small liquid plugs (volumes less than 1 µL); otherwise, substantial liquid losses would occur along tubing walls.From the experiment to find out the loss during the transfer from vial to tubing, we know 0.5 µL and 1 µL F-18 solution will leave in a vial less than 14% and 12%. (Table 1) Small diameter tubing was used (#30 AWG, 0.012" ID PTFE microbore tubing, Cole Parmer，USA) as we found slugs less likely to break than in larger diameter tubing. Using this tubing, the data of how much percent of liquid lose on the surface of 1 meter tubing is shown in Table. For example, 1 µL liquid loss on the surface of 1 meter tubing is less than 2.5% after passing through 1 meter tubing by using dynamical pressure control module. For larger samples, the effect is less severe and this component may not be needed. Ramping of pressure is useful in at least two places: (i) Initial pressurization of the reagent/sample vessel from zero to Ptransfer; and (ii) Increasing the driving pressure from Ptransfer to Ppurge once starting the valve actuation process.

In addition to providing means for rapidly purging air without losing sample through the bleed port, the system incorporates means for minimizing liquid loss in the reagent source and capillary tubing.

Table 1: Percent of 0.5 and 1 µL radioactive reagent loss in a vial

Volume (µL) / Loading Pressure (kPa) / Loss Percent
0.5 / 3.45 / 10±3.28%
1 / 3.45 / 8.1±2.92%

Table 2: Percent of liquid loss after going through 1 meter tubing

Volume (µL) / Loading Pressure (kPa) / Lbegin (mm) / Lend (mm) / Loss Percent
0.5 / 3.45 / 6.37 / 5.98 / 6.12±1.99%
1 / 3.45 / 10.93 / 10.81 / 1.1±1.19%
3 / 6.89 / 30.51 / 30.04 / 1.54±0.9%
5 / 6.89 / 51 / 50.63 / 0.73±0.82%
10 / 6.89 / 108.14 / 107 / 1.05±0.51%

Liquid-on-chip handling module design and fabrication

The liquid-on-chip handling module used in this experiment was integrated into a two-layer polydimethylsiloxane (PDMS) droplet generation microfluidic device. Its upper layer included a small sample introductionchamber, an inlet channel, a vent channel and two outlet channels. The sample introduction chamber was about 1400µmlong, 300µm wide, and 45µm high.All the Inlet, outlet and vent channels were 200µm wide, and 45µm high. The lowerlayer contained micro-valve control channels with 250 µm wide, 50 µm high. The microfluidic device was fabricated according to common fabrication practices(Quake and Scherer 2000; Thorsen et al. 2002).

Reference:

Quake SR, Scherer A (2000) From micro- to nanofabrication with soft materials. Science 290:1536-1540.

Thorsen T, Maerkl S, Quake S (2002) Microfluidic large-scale integration. Science 298:580-584.