Device for manipulation of packets in micro-containers, in particular in microchannels

Abstract

A microfluidic device for performing physical, chemical or biological treatment to at least one packet without contamination.

Claims

1. A device for performing a PCR, comprising: a microchannel comprising a coil comprising a capillary tube defining an internal space of the microchannel, wherein the capillary tube comprises a bulk fluorinated material that is non-internally coated, wherein the capillary tube is at least partly filled with: a carrier fluid comprising a fluorosolvent containing a surfactant, and aqueous droplets surrounded by the carrier fluid, the coil comprising a denaturing region and an annealing region, both regions being at different temperatures, wherein the difference between an interfacial tension between a droplet and the capillary tube and an interfacial tension between the droplet and the carrier fluid is at least 26 mN/m.

2. The device according to claim 1, wherein the bulk fluorinated material is a fluoropolymer.

3. The device according to claim 1, comprising a cylinder comprising three regions corresponding to the denaturing, annealing and elongation regions.

4. The device according to claim 3, the capillary tube being wound around the cylinder.

5. The device according to claim 3, the denaturing, annealing and elongation regions being isolated one from another by sheets, which are affixed between the pieces of the cylinder.

6. The device according to claim 3, the size of the elongation region being twice the size of the denaturing and annealing regions.

7. The device according to claim 1, the device having a cylinder shape and comprising three ventilation holes per quarter cylinder, said holes being drilled through the entire device to provide vents for air cooling.

8. The device according to claim 7, further comprising a turbine configured to blow ambient air through the ventilation holes providing temperature control and uniformity.

9. The device according to claim 1, the device having a cylinder shape, two holes for receiving thermocouples and one central hole for receiving a heater being drilled partially through the cylinder in each temperature region, the central hole being larger than the holes for receiving the thermocouples.

10. The device according to claim 1, each region comprising a resistance heater and two thermocouples.

11. The device according to claim 10, the resistance heaters being located in the center of their respective region, and the thermocouples being located near the interface between regions.

12. The device according to claim 1, the capillary tube comprising a DNA fragment.

13. The device according to claim 1, wherein the coil comprises an elongation region being at a temperature different from the temperature of the denaturing and the annealing regions.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) The invention may be better understood on reading the following detailed description of non-limiting embodiments, and on examining the accompanying drawings, in which:

(2) FIG. 1 is a diagrammatic partial view of a microfluidic device according to the invention,

(3) FIG. 2 is a diagrammatic view in cross section of the device of FIG. 1,

(4) FIG. 3 is a diagrammatic perspective view of a support member of the device of FIG. 2,

(5) FIGS. 4A-4C and 5A-5C illustrate diagrammatically respectively three steps of two coalescence operations according to the invention,

(6) FIGS. 6A-6C illustrate diagrammatically and partially three steps of the aliquoting of a droplet according to the invention,

(7) FIG. 7 is a diagrammatic view of an electric field distribution in a portion of the microchannel,

(8) FIG. 8 is a diagrammatic partial view of a device according to a variant of the invention, and

(9) FIGS. 9 to 13 illustrate diagrammatically and partially other variants of the invention.

(10) FIG. 14 plots the interfacial tension between fluorosurfactant 1H,1H,2H,2H perfluorodecan-1-ol and fluorinated oil FC-40, as measured in example 11,

(11) FIG. 15 illustrates diagrammatically and partially a cycling PCR system used in examples 12 and 13 to amplify DNA in segmented flow according to the invention,

(12) FIG. 16 is a gel electrophoresis of the product of amplification in a train of droplets, as described in example 12,

(13) FIG. 17 is a verification of the absence of contamination between neighboring droplets, as described in example 13,

(14) FIGS. 18A-18E show droplet shapes in an untreated, unwashed silicone capillary, with FC40 as a carrier fluid (a) Initial droplet, (b) Later droplet (>60.sup.th droplet in the train); with fluorosurfactant solutions at different concentrations (c) 0.15 wt % fluorosurfactant, (d) 0.5 wt % fluorosurfactant, (e) 3.0 wt % fluorosurfactant (examples 14A and 14 B),

(15) FIGS. 19A-19E show droplet shapes in silicon tubes silinazed in an Argon atmosphere, (a) 7.5 vol % silane, (b) 5 vol % silane, initial droplets, (c) 5 vol % silane, later droplets (>100.sup.th droplet), (d) 1 vol % silane, (e) 0.25 vol % silane (example 14C),

(16) FIGS. 20A-20E show droplet shapes in tubes silanized in air, (a) 10 vol % silane, 5 minute reaction, (b) 7.5 vol % silane, 5 minute reaction, later droplet shape (>110th droplet in train), (c) 5 vol % silane, 30 minute reaction, HCl activation, (d) 5 vol % silane, 30 minute reaction, plasma activation (e) 2.5 vol % silane, 30 minute reaction (Example 14 D),

(17) FIGS. 21A-21B show droplet shapes in silanized silicon tubes with 0.5 wt % fluorosurfactant (a) 2.5 vol % silane, 30 minute reaction, 200th drop, (b) 5 vol % silane, 30 minute reaction, 200th drop (Example 14E), and

(18) FIGS. 22 to 24 represent diagrammatically and partially a contamination free connector according to the invention.

1: FIRST EXEMPLARY EMBODIMENT OF THE INVENTION

(19) FIG. 1 shows a microfluidic device 1 according to the invention, said device comprising a microchannel 2 and an electrode assembly 3. The microchannel 2 has a longitudinal axis X and an internal cross section that is circular.

(20) The electrode assembly 3 comprises a pair of electrodes 4, each electrode 4 comprising a metal cylinder, for example aluminium. The length of each electrode 4 is for example 4 mm and the inner diameter 1.5 mm and the outer diameter 1.9 mm.

(21) The electrodes 4 are placed around the microchannel 2 and are spaced along the axis X by a gap 6. The electrodes 4 are connected to a generator unit 9 via connection elements 8 comprising electrical wires.

(22) The electrodes 4 are housed in a support 10 comprising two support members 11, each being a substantially rectangular parallelepiped made of Plexiglas®, for example with a width of 24 mm, a height of 20 mm and a depth of 20 mm.

(23) A first cylindrical hole 12, for example of diameter 1.9 mm, is drilled in each support member 11 along the axis X at the center of the support member 11 for holding the electrode 4. The first hole 12 extends from a front face 13 towards a rear face 14 of the support member 11, opposite to the front face 13.

(24) A second hole 17 is drilled perpendicular to the first hole 12 for receiving a connection element 8 connecting the corresponding electrode 4 to the generator unit 9.

(25) Each support member 11 further comprises two holes 20 and 21 whose axes are parallel to the axis of hole 12 and configured for receiving respectively a Teflon® screw 22 and a metal rod 23 in order to maintain the support members 11 assembled with the holes 12 being collinear.

(26) Each electrode 4 is mounted in the corresponding support member 11 such that the electrode 4 is flush with the front face 13, as illustrated in FIG. 2.

(27) The front faces 13 of the support members 11 are spaced for example by a length of 2 mm defining a 2 mm gap 6 between the electrodes 4.

(28) The generator unit 9 comprises for example a function generator connected to an amplifier such as to deliver sinusoidal voltages up to 2 kV with frequencies up to 1 kHz. The generator unit 9 may also comprise a central processing unit such as a computer to programmably control the voltage delivered to the electrodes 4.

(29) FIG. 3 shows diagrammatically the orientation of electric field given by the arrows together with the equipotential lines.

(30) As one can see, the electric field is substantially collinear to the axis X of the microchannel 2 and thus favors electrocoalescence and minimizes any effect of dielectrophoresis.

(31) The device 1 may be mounted on an observation stage of a binocular microscope 30 connected to a CCD camera 31 and a video recorder 32, as illustrated on FIG. 1, thus enabling the monitoring of the collapsing of two packets or the splitting of a packet in the microchannel.

(32) The device may be configured such that after the collapsing or splitting, the packet(s) are drained off.

2: SECOND EXEMPLARY EMBODIMENT OF THE INVENTION

(33) FIG. 8 shows an electrode assembly 3′ comprising two composite electrodes 35 each having a pair of substantially parallel equipotential plates 36 facing each other and sandwiching a microchannel 2′ which has a rectangular cross-section. Each pair of plates 36 is connected to a respective pole of the generator unit 9.

3: EXAMPLE OF AN OSCILLATION METHOD

(34) In an embodiment, the droplet fluid is TBE 5× buffer (0.45 M Trisbase®, 0.45 M boric acid and 0.01 M EDTA; Sigma®) dyed with 0.25 wt % bromophenol blue for observation in a carrier fluid of fluorinated oil (FC-40, 3M) with 0.5 wt % 1H,1H,2H,2H perfluorodecan-1-ol (Fluorochem®) added to prevent interactions with the wall of the microchannel 2. The droplet conductivity is 3 mS/cm and the carrier fluid conductivity is 2.5.10.sup.−13 mS/cm. For droplet formation, the two fluids are layered in a 1.5 ml Eppendorf® tube so that the bottom layer consists of approximately 0.6 ml of the carrier fluid (FC-40/1H,1H,2H,2H perfluorodecan-1-ol) and the upper layer consists of approximately 0.6 ml of the droplet fluid (TBE 5×/bromophenol blue).

(35) The microchannel 2 is filled from a syringe pump (commercialised by KD Scientific) using for example a Hamilton® Gas-Tight 250 μl syringe filled with the carrier fluid. The excess fluid pumped into the microchannel may be collected in a waste reservoir. After completely filling the capillary, the microchannel 2 is placed into the carrier fluid phase of the layered Eppendorf® tube. The pump is then aspirated at a rate of 1 ml/hr. Droplets are formed by oscillating the microchannel between the carrier phase and the droplet phase, either manually or by attaching the microchannel to a mechanical oscillator, at for example approximately 2 Hz. Droplets formed by this method have approximately the same diameter as the channel.

4: EXEMPLARY METHOD FOR DISPLACING A DROPLET IN A MICROCHANNEL

(36) The method may be carried out with any of the devices defined above.

(37) A single droplet is formed by the oscillation method described above. Using the syringe pump, the droplet is aspirated into the gap 6 between the electrodes 4. When the droplet has reached the section of the gap just before the upstream electrode, the flow is stopped and the system allowed to settle to equilibrium. A continuous voltage is then applied, with the positive voltage applied to the electrode 4 closest to the droplet and with the farthest electrode grounded. The motion of the droplet towards the grounded electrode can be recorded on video and the time for a given displacement measured. The droplet only moves when it is between the electrodes 4 and stops when it is under the grounded electrode.

5: EXAMPLE OF STATIC COALESCENCE

(38) The static coalescence may be performed by any of the devices defined above.

(39) Droplets are formed by the oscillation method described above such that the spacing between the two droplets is larger than the gap 6 between the electrodes 4. A first droplet is initially brought into the gap 6 between the electrodes 4 using the syringe pump. When the droplet arrives at the upstream electrode, the flow is stopped. The droplet is moved against the direction of the previously applied flow using the electric field actuation described above until it reaches the downstream electrode. The flow is restarted until the first droplet returns to the upstream electrode. This procedure is repeated until a second droplet appears between the electrodes 4. The microchannel position in the electrodes is then adjusted so that the second droplet is outside of the gap 6 between the electrodes. The first droplet is moved against the direction of the previously applied flow until the gap between the two droplets is for example 0.5 mm. The microchannel is then repositioned such that the midpoint between the two closest edges of the droplets is centered between the two electrodes and the system is allowed to settle to equilibrium.

(40) In the example depicted in FIGS. 4A to 4C, the first droplet 40 has a diameter of about 540 μm and the second droplet 41 has a diameter of about 560 μm. Upon applying a 2 kV, 1 kHz sinusoidal voltage to the electrodes 4, initial droplet motion is steady, with the smaller droplet 40 moving at a slightly higher velocity (FIGS. 4A and 4B). When the droplets 40 and 41 come into close contact they rapidly accelerate and drain the intervening film (FIG. 4C), which indicates that the device 1 should provide essentially instantaneous coalescence of droplets which are initially close together, such as occurs after two droplets arrive simultaneously at a T-junction.

6: EXAMPLE OF IN-FLIGHT COALESCENCE

(41) The in-flight coalescence may be performed by any of the devices defined above.

(42) The first droplet 43 may have a diameter of about 580 μm and the second droplet may have a diameter of about 560 μm.

(43) After positioning the droplets, the microchannel is displaced such that both droplets are outside the gap 6 between the electrodes 4. A 2 kV, 1 kHz sinusoidal voltage is then applied to the 2 mm-spaced electrodes and kept on throughout the duration of the experiment. After applying the electric field, the flow is started by aspirating with the syringe pump at 50 μL/hr.

(44) The droplet 42 enters the gap 6 between the electrodes 4 and moves at a constant velocity in the absence of the droplet 43 (FIG. 5A). The leading interface of the droplet 43 appears after 13 sec, but has no effect on the droplet 42. Only when the trailing droplet is well inside the gap 6 between the electrodes does the dipolar force manifest itself. Thereafter, the coalescence time is essentially the same as in the static case (approximately 8 sec) for these widely separated droplets, but the dynamics are slightly different due to the flow. The dipolar force is sufficiently strong to stop the droplet 42 (FIG. 5B), whereupon the droplet 43 moves toward it at a constant rate, closing the distance at essentially the same rate as in the static case. Once the droplets are close together, the strong dipolar force rapidly drains the intervening fluid and coalescence is achieved (FIG. 5C).

7: EXAMPLE OF DROPLET SPLITTING

(45) A single large droplet 46 is formed by oscillating the interface as described above but at a lower frequency. The droplet 46 is brought into the gap 6 between the electrodes, by aspirating with the syringe pump (FIG. 6A). The drop depicted in the present embodiment is an ellipsoid of revolution with a 2.5 mm long axis. Upon applying a 2 kV, 0.1 Hz square tension, the drop 46 splits into two smaller, stable drops 47 (FIGS. 6B and 6C) that are ejected from the gap 6.

(46) Typical operating conditions for achieving a clean droplet splitting, that is to say one big droplet splitting into two smaller and stable ones without formation of any satellite drops may consist in a square voltage with a frequency between 0.1 and 1 Hz and an amplitude between 1 kV and 2 kV. Under such condition, the droplet may break in less than 1 minute. The lowest the voltage applied, the “cleanest” the splitting but the longer it may take. The droplet length may be about the length of gap 6.

8: OTHER EXEMPLARY EMBODIMENTS OF THE INVENTION

(47) As illustrated in FIGS. 9 and 10, the microchannel portion 50 between the electrodes 4 may form a T-intersection with a transverse channel 51. After coalescence of the droplets caused by the electric field between the electrodes 4, the resulting droplet 52 may be driven in the transverse channel 51, for example by using a syringe pump connected to the transverse channel 51.

(48) As illustrated in FIGS. 11 and 12, the droplet splitting may be performed by extracting a droplet 53 from a relatively large mass of fluid 54 by applying the electric field between the two electrodes 4.

(49) Thus droplets 53 can be formed when desired, for example by programmably controlling the electrodes 4.

9: ANOTHER EXEMPLARY EMBODIMENT OF THE INVENTION

(50) FIG. 13 shows a device 60 according to the invention, said device 60 comprising a microchannel 61 connected in a portion 62 to first and second side channels 63 and 64.

(51) Portion 62 may for instance be situated substantially at the middle of the microchannel.

(52) In the present embodiment, the microchannel 61 may have a thickness of about 100 μm and a width of about 300 μm and the side channel 63 a thickness of about 100 μm and a width of about 50 μm.

(53) The side channel 63 is connected to a delivery system 66 comprising a syringe pump having a reservoir 67 containing a solution of surfactant of oleic acid and SDS in hexadecane, at a concentration superior to the critical micellar concentration.

(54) The microchannel 61 is filled with a solution containing hexadecane containing SPAN® 80 at a concentration adjusted to avoid interaction of aqueous droplets with the microchannel walls.

(55) The side channel 64 is connected to a springe pump 68 in aspiration mode configured to aspirate the solution from the microchannel 61.

(56) Two droplets 70 of a 5×TBE Buffer are introduced and displaced in the microchannel 61 by its both ends. The aspiration of droplets 70 by both ends is synchronized so that the droplets 70 arrive from both sides at the same time at the portion 62. When the droplets 70 are in the portion 62, a solution of surfactant contained in the reservoir 67 is delivered by the delivery system 66 into the portion 62 of the microchannel with a predetermined flow rate such as droplets 70 coalesce. The optimal flow rate may be determined by progressively increasing the flow until droplets coalesce at each collision, which sets the optimal flow rate.

(57) In another embodiment, the solution of surfactant may be delivered in pulses synchronized with the arrival of the pair of droplets at the connection portion 62.

(58) The resulting droplet 71 is collected in syringe pump 68 for further use, or e.g. transferred to another microchannel for detection.

10: EXAMPLE OF FORMATION AND TRANSPORTATION OF REGULAR ARRAYS OF WATER DROPLETS IN A FLUORINATED OIL CONTAINED IN A FLUOROPOLYMER CAPILLARY TUBE

(59) Trains of droplets are created by using a Y-connector (Upchurch Scientific) connected to an electro-pinch valve (NResearch, Caldwell N.J.). One entry to the Y-connector was filled with TBE 5× buffer (0.45M Trisbase, 0.45M boric acid, and 0.01M EDTA, Sigma) dyed with 0.25 wt % bromophenol blue for ease of observation. The other side of the Y-connector and the test capillary (PFA, i.d. 800 μm, Upchurch Scientific) were primed with either the bulk fluorinated oil FC-40 (3M) or FC-40 containing various amounts of a fluoroalcohol surfactant (1H,1H,2H,2H perfluorodecan-1-ol, Fluorochem). Droplet trains were created by cycling the electrovalve with a LabView program while aspirating with a computer-controlled Harvard milliliter module syringe pump. A typical cycle consists of 6 seconds of aspiration from the TBE line and 8 seconds of aspiration from the FC-40 line. The droplets were observed with a binocular microscope (Olympus) and recorded using a CCD camera (Hitachi) and WinTV.

(60) The stability of droplet trains in the fluidic system is first tested. In previous work on segmented flow PCR, cross-contamination between droplets was attributed to droplet instability and the formation of small satellite droplets. We not only wanted to look for droplet breakage in individual droplets, but also to observe the overall stability of droplet trains containing several hundred droplets. The stability of such trains is essential for high-throughput applications of our technique.

(61) When the droplets are entrained in pure FC-40, it has been observed that they occasionally stick to the walls. This is unexpected, since both the walls and the carrier fluid are fluorinated, and it has been expected that the walls will be strongly wetted by the FC-40. The sticking has been attributed to imperfections (either roughness or chemical inhomogeneities) in the capillary walls, which would be expected in bulk capillaries manufactured by an extrusion process. The layer of FC-40 between the droplets and the capillary wall is very thin, and small perturbations to the wall surface could disrupt the lubrication flow. In any event, once one droplet becomes entrained on the wall, even temporarily, the train as a whole loses its stability. The trailing droplet collides with the entrained droplet and exchanges fluid, the entrained droplet is released from the wall, and the trailing droplet becomes entrained. This process proceeds ad infinitum and would be catastrophic in any PCR application.

(62) The fluoroalcohol surfactant is then added to FC-40 in the range of 0.5-3.0 wt %. Upon making trains containing over 200 droplets, no transient pinning to the walls, satellite droplet formation, or instabilities in the droplet train are observed.

11: MEASUREMENT OF INTERFACIAL TENSION BETWEEN A WATER DROPLET AND A SOLVENT CONTAINING FLUOROSURFACTANT

(63) Interfacial tension measurements were made using a homemade drop-volume tensiometer. The FC-40/surfactant drop is dispensed into a reservoir of TBE 5× buffer from a 0.8 mm ID Teflon® capillary tube. Using a 50 μL Hamilton gastight syringe and a Hamilton PSD/2 syringe pump at maximum resolution, the drop volume could be increased in 25 nL increments with arbitrary waiting times between steps. To allow for equilibration of the surfactant, we typically waited 30 seconds between steps. The tension measurements are the average of at least 15 different drops, corrected for wetting of the Teflon® tip.

12: EXAMPLE OF PCR AMPLIFICATION OF DNA IN A DEVICE ACCORDING TO THE INVENTION, INVOLVING A CHANNEL MADE OF BULK FLUOROPOLYMER

(64) A PCR device 80 is depicted in FIG. 15, said device 80 comprising a 4 cm diameter copper cylinder 81 machined into three pieces corresponding to the denaturing 82, annealing 83, and elongation 84 regions. The elongation region 84 is twice the size of the other two regions, thus occupying half the cylinder 81. The three regions are isolated one from another by polycarbonate sheets 85, which are affixed between the pieces of the copper cylinder. The two ends of the heater are capped with a polycarbonate cylinder to provide structural stability while maintaining isolation between the two zones. Three small holes 88 per quarter cylinder are drilled through the entire device to provide vents for air cooling. Two small holes 89 (for the thermocouples) and one larger central hole 90 (for the heater) are drilled partially through the cylinder in each temperature region. The heating and thermocouple holes are not open to the air-cooling side of the heater. Each region is covered with a thin layer of aluminum foil in contact with the respective region of the copper cylinder to provide heating from above. The foil layers are covered by a polycarbonate shell and a layer of cotton, thereby insulating the cylinder and providing a uniform temperature across the capillary. A small turbine blows ambient air through the three ventilation holes per quarter-cylinder, allowing for better temperature control and uniformity.

(65) Each region includes a resistance heater and two Pt-100 thermocouples. The resistance heaters are located in the center of their respective zone, while the thermocouples are located near the interface between zones. The heating elements and thermocouples are connected to custom electronics. The thermocouples communicate with a custom PID program control written in LabView via a Keithley 2701 multimeter. The temperature of each zone can be set arbitrarily. For the experiments discussed here, it has been used a denaturing temperature of 94° C., an annealing temperature of 55.5° C., and an elongation temperature of 72° C. With our design, a temperature difference of ±0.2° C. across each zone is achieved.

(66) A 4.5 meter long transparent PFA capillary tube 92 (i.d. 800 μm, Upchurch Scientific) enters the cylinder through a groove in the denaturing region, providing an initial denaturation step of approximately 1 minute. The capillary is then wound 35 times around the cylinder, corresponding to 35 PCR cycles. The capillary exits the heater through a hole in the extension segment, providing approximately 30 seconds of additional extension on the 35.sup.th cycle.

(67) PCR Amplification: The template is a 2823 base pair DNA fragment of Litmus 28i (New England Biolabs). This fragment is amplified on 572 base pairs from base 2008 to base 2580 using Eurogentec primers (lower primer 5′-CGC-ATT-GCG-GTA-TCT-AGA-ACC-GGT-GAC-GTC-3′ (SEQ ID NO: 1), upper primer 5′-AGC-TTG-GAG-CGA-ACG-ACC-3′ (SEQ ID NO: 2), Eurogentech Oligold). A 50 μL PCR mix is prepared using the Ready Mix Taq reaction mixture (Sigma) according to the manufacturer's specifications with the maximum concentration of template and primers.

(68) The carrier fluid is a bulk fluorinated oil FC-40 (3M) containing 0.5%-1.0% wt. Fluoroalcohol surfactant (1H,1H,2H,2H perfluorodecan-1-ol, Fluorochem). The surfactant prevents the transient adsorption of droplets to the capillary walls. The 2 μL aqueous droplets are injected into the inlet by aspirating from the capillary outlet using a Hamilton PSD/2 pump and a 100 μL Hamilton gastight syringe. The drops are separated one from another by a 5 μL FC-40 spacer. After injecting the desired number of droplets, the outlet is disconnected from the Hamilton pump and the inlet is then connected to a computer-controlled Harvard milliliter-module pump with a 5 mL Hamilton gastight syringe. The droplets are circulated at 0.1 cm/s.

(69) The droplets are collected at the outlet and analyzed by gel electrophoresis on a 1 wt. % agarose gel in 0.5×TAE buffer. A control amplification sample is made by amplifying the remaining volume of the 50 μL PCR mix in a classic PCR thermal cycler (Perkin Elmer) with a cycle of 1 minute at 94° C., followed by 35 cycles of 94° C. for 30 seconds, 55° C. for 30 seconds, and 72° C. for 1 minute. This mimics the cycling in our continuous-flow PCR, although the lag time for heating and cooling the classic cycler means that the total amplification is approximately twice as long as our flow device. A 2 μL aliquot of the amplified control system is used for the gel electrophoresis.

(70) It has been made a train of five droplets, each one containing the PCR mix and template. The result of this successful amplification in all droplets is depicted in FIG. 16. Lane 1: 1 kbp DNA ladder (New England Biolabs), Lane 2: 24, control sample of mix with DNA, Lanes 3: Droplet 1, Lane 4: Droplet 2, Lane 5: Droplet 3, Lane 6: Droplet 4, Lane 7: Droplet 5.

(71) The degree of amplification in the device according to the invention (lanes 3 to 7) is comparable to that obtained in the conventional thermal cycler (lane 2).

EXAMPLE 13: STUDY OF CONTAMINATION BETWEEN DROPLETS IN A CHANNEL MADE OF BULK FLUOROPOLYMER

(72) The system has been tested for cross contamination between droplets. All conditions are identical to Example 12, except that two separate PCR mixes are made; a first mix contains the template, primers and Ready Mix reaction mixture and the second mix is identical except that it does not have any template. Five droplets are aspirated, but only the third droplet contains the template. In order to avoid contamination from the tip itself, the latter has been washed it in distilled water between each droplet injection. FIG. 17 shows the gel electrophoresis result from this experiment. Lane 1: 1 kbp DNA ladder (New England Biolabs), Lane 2: 5 μL control sample of mix without DNA, Lane 3: 2 μL control sample of mix with DNA, Lane 4: Droplet 1 (no template), Lane 5: Droplet 2 (no template), Lane 6: Droplet 3 (with template), Lane 7: Droplet 4 (no template), Lane 8: Droplet 5 (no template). There is no observable contamination between different droplets—the only droplet exhibiting any DNA amplification is the third droplet, which contained the DNA.

EXAMPLE 14: EXAMPLES OF DROPLET TRANSPORT IN A CHANNEL MADE OF A NON-FLUORINATED MATERIAL WITH AND WITHOUT COATING WITH A LAYER OF FLUORINATED MATERIAL

(73) In this series of example, the microchannel is made of Silicon tubes (inner diameter 0.8 mm) commercialized by Cole Parmer. The carrier fluid is FC-40 (3M), and the droplets are made of the aqueous buffer TBE 5×. In all cases, a train of droplets is created following the same protocol as described in Example 13, and the shape and migration of droplets in the tube is directly observed and photographed with a binocular and CCD camera.

14A: DROPLETS OF TBE IN PURE FC40 IN UNTREATED SILICONE TUBE

(74) In some instances, the droplets appeared to have a spheroidal shape (FIG. 18a) which indicates that the walls were not wetting. However, after forming a droplet train with several hundred droplets, the walls became wetting at some point during the train (FIG. 18b). In some instances, the first droplets looked like FIG. 1b. It has been presumed that the initially non-wetting behavior observed in some instances is due to a residual product from the manufacturing process. As many droplets are moved through the system, this unknown product is removed from the walls. To test this hypothesis, the tubes are washed with 5 volumes of distilled, purified water. Afterwards, the first droplets always have a behavior like FIG. 18b. For all subsequent experiments, the tubes are always with 5 volumes of distilled, purified water to remove any variations in the initial conditions.

14B: DROPLETS OF TBE IN FC40 WITH FLUOROSURFACTANT ADDED, IN UNTREATED SILICONE TUBE

(75) The behavior of the droplets is tested with the addition of various wt % of a fluorosurfactant, 1H,1H,2H,2H perfluorodecan-1-ol, Fluorochem. The surfactant reduces the interfacial tension between the droplets and FC-40 but does not affect the solid-liquid tension. As a result, the droplets are destabilized and break into many small droplets (FIG. 18 c, d, e) The nature of the breakup depends on the surfactant concentration but even at extremely low levels of fluorosurfactant the droplets are still unstable.

14C: DROPLETS OF TBE IN FC40 WITHOUT FLUOROSURFACTANT ADDED, IN A SILICONE TUBE. TREATED BY SILANIZATION IN ARGON ATMOSPHERE

(76) The tubes were silanized while isolated in an Argon atmosphere. A small quantity of 1N HCl (Sigma) was heated to approximately 60 C on a hotplate. One end of a cleaned silicon tube was connected to a syringe of at least double the volume of the tube and the other end was placed in the warm HCl solution. The HCl was aspirated into the tube until the syringe was partially filled. The HCl was left in the tubes for 5 minutes, during which we occasionally oscillated the syringe pump to provide local mixing. The HCl was then evacuated from the tube and the tube was dried with a flow of Argon. We then connected a new syringe to one end of the tube and placed the other end of the tube in a solution of fluorosilane and spectroscopic grade methanol (Sigma). The fluorosilane is usually 1H,1H,2H,2H,-perfluorooctyltrimethylsilane (Fluorochem). 1H,1H,2H,2H-perfluorodecyltriethoxysilane (Fluorochem) has also been tested and essentially the same results are achieved. All the results shown here are for 1H,1H,2H,2H,-perfluorooctyltrimethylsilane (Fluorochem), which we chose to use because it is less expensive. The silane solution is aspirated into the tube and left for 5 minutes, during which we occasionally oscillated the syringe pump to provide local mixing. The silane solution was evacuated from the tube and the tube was dried with a flow of Argon. The dried tube was then placed in an oven at 110 C for approximately 20 minutes to fix the silanes. The tube was then washed with several volumes of methanol and FC-40 before performing the droplet test.

(77) The results of droplets in Argon-silanized tubes are shown in FIG. 19. For vol % greater than 7.5, we observed no pinning for trains including at least 200 droplets. At 5 vol %, the droplets appeared to be initially nonwetting, but eventually pinned to the wall. For lower vol %, the droplets always wetted the wall. We concluded that 7.5 vol % fluorosilane is necessary to form a stable, nonwetting surface when the reaction is performed under Argon.

14D: DROPLETS OF TBE IN FC40 WITHOUT FLUOROSURFACTANT ADDED, IN A SILICONE TUBE. TREATED BY SILANIZATION IN AIR ATMOSPHERE

(78) To simplify the silanization procedure, the tubes are silanized in air. In order to better preserve the pure silane, the silane/methanol mixture is first performed in Argon but the remainder of the reaction is then performed in a hood using essentially the same protocol as above. The vol % of silane and the time that the silane was allowed to react with the tube (reaction time) are varied. In one instance, the HCl activation step is replaced with plasma activation.

(79) For a 10 vol % silane and 5 minute reaction time, the silanization in air was indistinguishable from the case in Argon (FIG. 20a). At 7.5 vol % silane and a 5 minute reaction, the droplets initially appeared nonwetting but eventually wet the walls (FIG. 20b). It has been found that at 5 vol % silane the droplets still wet the walls, but when the reaction time is increased to 30 minutes, the droplets were nonwetting throughout the entire train (>200 droplets, FIG. 20c). The same parameters (5 vol % silane, 30 minute reaction) are tested but the tube is placed in plasma for 1 minute rather than filling it with HCl. As in FIG. 20d, the plasma activation did not result in a nonwetting surface. Since the plasma must diffuse through the tube for activation, it is less efficient than the liquid activation by HCl, which can be pumped through the tube. It has been tested further lowering the silane concentration to 2.5 vol % silane while retaining the 30 minute reaction time. As in FIG. 20e, the walls are slightly wetting.

(80) The silanization in air probably produces a less uniform coating on the surface than in Argon, since the water in air competes with the silane for the activated surface sites. In essence, the air silanization does not reduce the water-solid tension as much as the Argon silanization. However, the air protocol is much simpler to perform and more amenable to automation.

14E: DROPLETS OF TBE IN FC40 WITH FLUOROSURFACTANT ADDED, IN A SILICONE TUBE. TREATED BY SILANIZATION IN AIR ATMOSPHERE

(81) It has been checked if fluorosurfactants, which strongly reduce the FC-40/water interfacial tension, are sufficient to overcome the nonuniformities (and concomitantly lower water-solid tension) that arises from the air-silanized surface. As indicated in FIG. 21, a small concentration of surfactant (0.5 wt %) is sufficient to prevent wetting at both 2.5 vol % silane and 5 vol % silane, even after forming a train of 200 drops.

EXAMPLE 15: VERIFICATION BY QUANTITATIVE PCR, OF CONTAMINATION BY DNA BETWEEN DROPLETS TRANSPORTED IN DEVICES ACCORDING TO THE INVENTION PREPARED IN EXAMPLE 14

(82) To test whether unpinned droplet shapes do not lead to contamination, a set of experiments has been performed using quantitative PCR to make a sensitive test of the DNA concentrations in different droplets.

(83) The tubes were prepared according to one of these protocols: 1. No Silane: Tubes were washed with several volumes of distilled water only. Carrier fluid is FC-40 (3M), prepared according to Example 14A. 2. Silane: Tubes were washed and then silanized in air according to the protocol in Example 14D with 5 vol % silane and 30 minute reaction time. The carrier fluid is FC-40. 3. Silane+Surfactant: Tubes were washed and then silanized in air according to the protocol in Example 14E. The carrier fluid is FC-40 with 0.5 wt % 1H,1H,2H,2H perfluorodecan-1-ol. 4. Bulk fluoropolymer+Surfactant. A train of droplets is prepared according to example 13: Teflon capillaries were used as supplied by the manufacturer. The carrier fluid is FC-40 with 0.5 wt % 1H,1H,2H,2H perfluorodecan-1-ol.

(84) One end of the tube was connected to a Y-connector, and the outlets of the Y-connector were selected using an electro pinch valve. One outlet goes to a Harvard millilitre module syringe pump and a 5 ml Hamilton gastight syringe and the second outlet goes to a Hamilton PSD/2 syringe pump with a 100 μl Hamilton gastight syringe. The Hamilton PSD/2 was used to make all of the droplets (by aspiration) or dispense the droplets from the tube (by pumping). The Hamilton millilitre module was used for droplet oscillation inside the tube. Prior to each experiment, the capillary was filled completely with the carrier fluid and the open end was placed in a reservoir of carrier fluid.

(85) The droplets are mixtures of Taqman PCR mix for quantitative PCR containing Gold Taq polymerase enzyme (Applied Biosystems), qPCR Core Reagent Kit (Eurogentec), specific primers, and a fluorescent probe (3′-ATCTGCTGCATCTGCTTGGAGCCCA-5′ (SEQ ID NO: 3), Applied Biosystems). “Mix” samples contained all of the components for PCR except for the template. The “DNA” samples contain cDNA isolated from cell line A549 at a concentration of 6.25 ng/μl. The fragment is amplified on 149 bp corresponding to the RPLPO gene using Proligo primers (upper primer 3′-GGCGACCTGGAAGTCCAACT-5′ (SEQ ID NO: 4); lower primer 3′-CCATCAGCACCACAGCCTTC-5′ (SEQ ID NO: 5)). We made two reservoirs for each experiment, one reservoir with sufficient mix for each control droplet (typically 30 μl) and a second reservoir with sufficient mix and template for the cDNA droplets (typically 22 μl).

(86) It has been tested for contamination during injection by the following procedure. 2 μl has been aspirated from the DNA reservoir and then 4 μl from the carrier fluid reservoir. The tip was then washed by dipping it in a reservoir of distilled water and drying with a ChemWipe. The subsequent 5 drops were formed by aspirating 2 μl from the mix reservoir and 4 μl from the carrier fluid reservoir. After all of the droplets were formed, we reversed the procedure and collected each droplet in a separate Eppendorf tube. The eppendorf tubes were stored at −80 C prior to the quantitative PCR.

(87) The contents of each drop were analyzed by quantitative PCR on a Taqman 7700 qPCR machine (Applied Biosystems). In these experiments, a value of 35 indicates that there was no detectable amount of cDNA in the droplets (i.e. 35 cycles of amplification without a fluorescence signal above the noise threshold), and each integer increment corresponds to a halving in the mass of cDNA.

(88) Table 1 presents the results from the inlet contamination experiment. There was significant contamination in the untreated capillary, as would be expected from the droplet shape. There was also contamination in the silanized capillary. However, there was no contamination in the mix reservoir, so we could conclude that the contamination occurs from droplet transport inside the capillary. For the silanized capillary with fluorosurfactant, we observed some contamination in the first wash droplet and equivalent contamination in the mix reservoir. This led us to believe that the contamination occurred at the tip during transfer, and that more vigorous washing of the tip after the aspiration of the DNA droplet should be sufficient to eliminate contamination at the inlet using a silanized capillary and fluorosurfacant. There was no contamination using the Teflon tip.

(89) We then tested the ability of the fluorosurfactant and silanized capillary to prevent contamination while the droplets were in transit. Using essentially the same aspiration procedure as above with the fluorosurfactant and a silanized capillary, we first made two mix droplets, then a cDNA drop, and then two mix droplets. The difference between the present injection protocol and the one above is that we now washed the tip in two separate distilled water reservoirs after injecting the DNA drop, with the aim of reducing the contamination at the tip. At the end of the procedure we had a 2 μl cDNA drop with two 2 μl mix droplets on either side of it (Control 1 and 2 leading, Control 3 and 4 trailing), where each droplet was spaced by 4 μl of the carrier fluid.

(90) We then aspirated using the Harvard millilitre module so that the droplets were in a straight section of the capillary positioned below an Olympus binocular microscope. Using computer control, we oscillated the Harvard millilitre module so that it pulled the droplets 5 cm in the capillary at an average velocity of 1 mm/sec, and then pushed the droplets 5 cm at an average velocity of 1 mm/sec, so that a single cycle resulted in no net displacement of the droplets. We performed 50 such cycles, so that the total distance travelled by the droplets (5 m) is comparable to that required in our continuous flow PCR machine. By oscillating the droplets, rather than pushing them at a constant rate through a 5 m capillary, we simulate the conditions in a high throughput operation. We occasionally observed the droplets with the microscope to ensure that they were not wetting the walls.

(91) After completing the 50 cycles, we collected the droplets in individual eppindorf capillaries. After ejecting each droplet, we aspirated a 2 μl wash droplet of distilled water to clean the tip. The wash droplets were also collected. All of the droplets and the mix reservoir were analyzed by quantitative PCR as above.

(92) Table 2 presents the results of the quantitative PCR. There was no detectable contamination in any of the control drops. Moreover, there was no detectable contamination in the wash droplets. There was some contamination in the mix reservoir, but this did not lead to contamination in any of the control droplets. We conclude that the combination of the fluorosurfactant and a silanized capillary should be sufficient to prevent contamination in high throughput applications.

(93) We performed essentially the same oscillation experiment using a Teflon capillary and fluorosurfactant. The only difference is that in this experiment we did not wash the tip when collecting the droplets or use the extra wash droplets.

(94) The results with a Teflon tube and fluorosurfactant are presented in Table 3. As in the inlet contamination test, there is no contamination in the mix, indicating that the washing is sufficient to remove the DNA from the tip. There is some contamination in drop 2. We believe that this contamination is due to not washing the tip when collecting the droplets. Drop 2 exits the system immediately after the cDNA drop. It is possible that a small part of the cDNA drop may have become entrained on an imperfection in the tip surface, which would then be transferred to Drop 2. An automated droplet collection or in-line detection procedure should eliminate this source of contamination.

(95) TABLE-US-00001 TABLE 1 Quantitative PCR results for contamination at the inlet. 35.00 corresponds to no detectable cDNA, and each integer decrement corresponds to a doubling in the relative cDNA mass. No Silane Silane Silane + Surfactant Teflon + Surfactant Q PCR Q PCR Q PCR Q PCR Droplet Value Droplet Value Droplet Value Droplet Value cDNA 19.90 cDNA 20.32 cDNA 20.29 cDNA 19.74 Wash 1 32.99 Wash 1 32.18 Wash 1 32.89 Wash 1 35.00 Wash 2 35.00 Wash 2 33.66 Wash 2 34.36 Wash 2 35.00 Wash 3 32.15 Wash 3 34.91 Wash 3 34.77 Wash 3 35.00 Wash 4 32.54 Wash 4 35.00 Wash 4 35.00 Wash 4 35.00 Wash 5 35.00 Wash 5 34.92 Wash 5 35.00 Wash 5 35.00 Mix 33.15 Mix 35.00 Mix 33.13 Mix 35.00 Reservoir Reservoir Reservoir Reservoir

(96) TABLE-US-00002 TABLE 2 Quantitative PCR results for contamination during cycling in a silanized capillary with fluorosurfactant. 35.00 corresponds to no detectable cDNA, and each integer decrement corresponds to a doubling in the relative cDNA mass. Droplet Q PCR Value Control 1 35.00 Control 2 35.00 cDNA 20.17 Control 3 35.00 Control 4 35.00 Wash 1 35.00 Wash 2 35.00 Wash 3 35.00 Wash 4 35.00 Wash 5 35.00 Mix Reservoir 33.90

(97) TABLE-US-00003 TABLE 3 Quantitative PCR results for contamination during cycling in a Teflon capillary with fluorosurfactant. 35.00 corresponds to no detectable cDNA, and each integer decrement corresponds to a doubling in the relative cDNA mass. Droplet Q PCR Value Control 1 35.00 Control 2 30.19 cDNA 19.83 Control 3 35.00 Control 4 35.00 Mix Reservoir 35.00

EXAMPLE 16: DESIGN AND FABRICATION OF A CONTAMINATION FREE T-CONNECTOR

(98) This example illustrates the conception of a PDMS T-connector 100 with a very low dead volume allowing to connect without contamination three different entries compatible with the use of pinch valves.

(99) The connector 100 is made in PDMS (KODAK SYLGARD 184) with a 1:10 proportion of curing agent. The mould 101 used for manufacturing the connector 100 is a PMMA parallepiped (inner length and width 9 mm, inner height 8 mm). 800 μm inner diameter holes 102 are drilled on the center of three faces 103 of the parallepiped. A first 5 mm piece 104 of Teflon® capillary 105 (outer diameter 800 μm) is introduced inside the end of a 3 cm long second Teflon capillary (outer diameter 1.5 mm, inner diameter 800 μm) and sticks out on 1.5 mm. Three pieces are made this way and are introduced and maintained inside the parallepiped so as to form a T with the smaller tubing facing each other (see FIG. 22). The PDMS 106 is then degassed for 20 minutes, poured in the so constructed mould and put in a 65° C. oven for three hours.

(100) After three hours, the Teflon® pieces 104 and 105 are taken off by gently pulling on them, and the obtained T-connector is taken out of the mould 101. The T-connector 100 has three ports 108, each comprising coaxial cylindrical hollow portions following each other, the outer one 110 having 1.5 mm diameter on a length of 3 mm, the inner one 111 having 800 μm diameter on a length of 1.5 mm. 5 cm silicone capillary tubes 109 (Cole-Parmer; outer diameter 1.8 mm, inner diameter 800 μm) coated with silicon rubber (Dow Corning) are then introduced on 3 mm (corresponding to the outer cylinder) in each hole of the T-connector, and the connector is strengthened by adding more rubber around the tube near the entry of the PDMS T. Due to silicon and PDMS elasticity, it is possible to push the 1.8 mm outer diameter silicon tubing in the 1.5 mm diameter hole of the T, thus providing a good tightness of the junction. The obtained connector is put in the oven for two hours.

(101) The T-connector 100 with silicon capillary tube is further silanised with the method described in Example 14. A solid connector with very low dead volume and no leakage even under high pressure is obtained with this method. Furthermore, the use of silicon tube allows using pinch valves which have no dead volume.

EXAMPLES OF APPLICATIONS

(102) A device made in accordance with the invention may be used to carry out, for instance: mixing, nucleic acid screening, nucleic acid amplification, e.g. by PCR, NASBA, rolling circle amplification RNA reverse transcription genotyping, proteomic analysis, transcriptome analysis, crystallization, and in particular protein crystallisation, searching and evaluation of pharmaceutical targets, pharmaceutical hits or leads, or drugs, enzyme-protein reaction, antigen-antibody reaction, screening of libraries of chemical of biological products, high throughput screening, drug delivery, diagnosis, analysis or lysis of at least one living cell or dead cell, analysis of microorganisms, chemical reaction, reactive-catalyzer reaction, polymerization reaction, fusing particles, for example colloids, to form a chain, preparation of colloids, emulsions, vesicles, in particular monodisperse colloidal objects, preparation of nanoparticules or microparticles, environmental control, detection of pollutants, control of an industrial process.