The present invention relates to microfluidics. More specifically, the invention relates to interactions between immiscible fluids. In particular, the invention provides a method of generating fluid droplets.

The generation of fluid droplets has become increasingly important in a wide range of applications; for example, fluid droplets are used in many areas of microfluidics including chemical synthesis, drug screening and biomedical testing.

Known methods of generating fluid droplets include using T-junctions and flow focussing. These two methods are based on flowing two immiscible fluids such that they meet at a junction within a fluid channel network, where a deformation of one of the fluids is caused at the interface between the two fluids when they meet. This deformation eventually leads to a droplet breaking off.

T-junctions, as their name suggests, are based on a T-shaped junction in the fluid channel network. Two inlet channels meet orthogonally. Through one channel flows a dispersed fluid requiring emulsification, through the other channel flows a continuous fluid. At the interface where the two the fluids meet, the continuous fluid applies a shear force on the dispersed fluid, causing the dispersed fluid to deform, which ultimately results in droplet formation. The droplets, surrounded by the continuous fluid, then flow out of an outlet channel. The flow focussing method uses cross-shaped junction geometry within the fluid channel network. A dispersed fluid requiring emulsification flows into the junction through a central inlet channel. A continuous fluid flows into the junction through two further inlet channels, one either side of the central inlet channel. An outlet channel lies downstream of the three inlet channels. Due to the geometry of the junction the continuous flow exercises a pinching force on the dispersed fluid. The flow of the continuous fluid into the junction deforms the dispersed fluid as it flows through the junction, causing droplets to form. The droplets, surrounded by the continuous fluid, then flow out of the outlet channel. The above described methods of generating microfluidic droplets suffer from at least the following disadvantages.

Firstly, both methods result in significant dead volume in the fluid network. Dead volume is the portion of the internal flow path, and typically also a syringe providing the flow, of the fluid network that is out of the main flow path. Fluid entering the dead volume, either by flowing or diffusing in, is difficult to recover without interrupting the operation of the fluid network, resulting in some of the fluid in the network being wasted. Furthermore, both methods are generally restricted to the formation of droplets from a single sample per device. This limitation makes it difficult to produce mixtures of different droplets without having to use multiple junctions and a complex network to combine their outlets correctly. This is not helped by the fact that the devices required to form the droplets are complex, requiring micro-scale tubes / channels to be joined together to produce the junctions required for the methods.

In the light of the above-described disadvantages of known methods, the present invention aims to provide an improved method and apparatus for generating fluid droplets.

According to an aspect of the invention there is provided a method of generating fluid droplets, comprising: providing a first fluid within a first fluid conduit, the first fluid containing a surfactant; providing a second fluid within a second fluid conduit, the second fluid being immiscible with the first fluid; and flowing the second fluid into the first fluid conduit; wherein the surfactant in the first fluid acts to reduce interfacial tension between the first and second fluids as the second fluid flows into the first fluid, thereby generating a plurality of droplets of the second fluid.

The present invention simplifies the fluid channel network required for producing droplets. Instead of requiring a multi-channel, sealed fluid network with multiple inlets and outlets, a single channel can be used. Due to this simplicity, the method can be easily parallelised, for example, with multiple parallel lines being driven by a single syringe pump. The creation of droplets in a fluid conduit allows for the easy transfer of the droplets from the apparatus to a storage medium and/or other equipment, such as a thermal cycler. The first and second fluids may have different densities such that gravity acts to generate relative movement between the droplets of the second fluid and the first fluid within the first fluid conduit.

At least part of the first fluid conduit may be inclined away from the horizontal plane, preferably wherein the inclined part contains first fluid into which the second fluid (or droplets thereof) flows. A fluid compartment of a third fluid may be provided in the first fluid conduit, the fluid compartment being arranged to constrain movement of the droplets within the first fluid conduit, wherein the third fluid is immiscible with the first and second fluids. Preferably, the further fluid compartment is arranged as a blocking compartment to constrain the droplets generated from the second fluid within the first fluid conduit; for example, the fluid compartment may span the internal diameter of the first fluid conduit, but for thin films of the first fluid that are present between the fluid compartment and the walls of the first fluid conduit. Preferably, the fluid compartment is provided in the first fluid conduit before flowing the second fluid into the first fluid conduit. A further fluid compartment of a third fluid may be provided in the first fluid conduit after a plurality of droplets has been generated, preferably whereby to isolate said droplets.

The first fluid may have substantially no kinetic energy. Hence the first fluid does not exercise a kinetic force or pressure on the second fluid when they come into contact and formation of the droplets occurs in the absence of kinetic energy imparted from the first fluid onto the second fluid. The lack of requirement of kinetic energy (and optionally momentum) of the first fluid can enable simplification of the fluid channel network required for producing droplets, as no fluid pumping is required for the first fluid, nor are fluidic connections to the first fluid conduit required to enable pumping. The first fluid conduit may be a vessel for containing the first fluid. The first fluid may be at atmospheric pressure, or it may be below atmospheric pressure. The first fluid may be uniformly pressurised. The first fluid may have an insubstantial amount of kinetic energy. Such an insubstantial kinetic energy preferably has no effect on the formation of droplets (though it may have an effect on motion of the droplets in the first fluid). For example the first fluid may have insubstantial kinetic energy from flow of the second fluid that imparts an insubstantial amount of kinetic energy onto the first fluid. The first fluid may have an insubstantial kinetic energy from an insubstantial active flow of the first fluid intended to replenish surfactant in the first fluid. The second fluid may be unforced by motion of the second fluid. Substantially no momentum may be transferred from the first fluid to the second fluid. The first fluid may receive kinetic energy from the second fluid. In the absence of flow of the second fluid no flow may occur in the first fluid. Flow of the second fluid may induce flow of the first fluid. Momentum of the first fluid preferably does not cause formation of droplets. The first fluid preferably does not form a jet or a stream or a vortex.

Preferably, there is no active flow or pumping of the first fluid within the first fluid conduit. In other words, flowing the second fluid into the first fluid conduit and/or providing the third fluid compartment in the first fluid conduit (e.g. pumping the third fluid into the first fluid conduit) may cause the first fluid to flow, but otherwise there is little/no flow due to gravity.

The second fluid conduit may be inserted (at least partially) into the first fluid conduit. The inserted part of second fluid conduit, from which the second fluid emerges, is thereby surrounded by the first fluid (and surfactant) which resides in the first fluid conduit. Preferably, the second fluid conduit has an inner diameter that is narrower than the inner diameter of the first fluid conduit. Preferably, the inner diameter of the first fluid conduit is greater than the outer diameter of the second fluid conduit, such that there is a gap at an interface where the second fluid conduit is inserted into the first fluid conduit, whereby to allow fluid to escape the first fluid conduit.

The first fluid conduit may have an inner diameter of less than 1 mm. The fluid droplets generated may have a diameter of about 100 microns. Preferably, the droplets generated are mono-disperse droplets. The fluid conduits may be fluidly connected in a substantially in-line configuration. The second fluid conduit may be inserted into an end of the first fluid conduit.

The second fluid may be provided in the form of a fluid compartment. The second fluid conduit may be arranged to elongate the fluid compartment as it flows through it from the third fluid conduit. One or more second fluids may be flowed into the first fluid conduit from a plurality of further (second) fluid conduits. A portion of the first fluid conduit may be arranged to be fluidly connected with a plurality of further (second) fluid conduits. Each of the further (second) fluid conduits may be inserted (at least partly) into the first fluid conduit.

The second fluid conduit may be arranged to have a plurality of fluid outlets (or ports) through which the second fluid flows into the first fluid conduit. The second fluid conduit may be arranged to have multiple outlets (or ports) that can each be fluidly connected with the first fluid conduit. Preferably, the multiple ports are provided at an end of the second fluid conduit. Preferably, said end of the second fluid conduit may be inserted (at least partly) into the first fluid conduit. Multiple streams of fluid droplets can thereby be generated, preferably substantially simultaneously, from the second fluid as it flows into the first fluid conduit.

Part of the first fluid conduit may be arranged to provide a manifold-like configuration, such that a plurality of second fluids may be flowed into the first fluid conduit, preferably simultaneously, via a plurality of ports (or orifices) provided in the first fluid conduit. The plurality of second fluids may be the same, different or a combination of second fluids.

The second fluid may be initially provided within a third fluid conduit, wherein the third fluid conduit is fluidly connected to the (or each) second fluid conduit. Preferably, the connection is a fluid tight fit; preferably an interference fit.

According to another aspect of the invention there is provided a method of generating fluid droplets, comprising: providing a first fluid within a first fluid conduit, the first fluid containing a surfactant; providing a second fluid within a second fluid conduit, the second fluid being immiscible with the first fluid; and flowing the second fluid into the first fluid conduit; wherein the surfactant in the first fluid acts to reduce interfacial tension between the first and second fluids as the second fluid flows into the first fluid, whereby a plurality of droplets of the second fluid is generated. Further features are optionally as aforesaid. According to another aspect of the invention there is provided a method of storing a sample of droplets, comprising: providing a fluid conduit containing a first fluid and a plurality of droplets of an immiscible second fluid, wherein the droplets are substantially engulfed within the first fluid; providing a surface that is covered by a third fluid, which is immiscible with the first and second fluids; flowing the contents of the fluid conduit directly onto the surface beneath the surface of the third fluid, such that the first fluid isolates the droplets of the second fluid on the surface, and the third fluid isolates the first fluid on the surface. The tip of the fluid conduit may be inserted into a further fluid conduit having a tip that is wet by the third fluid, and the contents flowed onto the surface via the further fluid conduit.

The method may further comprise flowing onto the surface, spaced from said plurality of droplets a further plurality of droplets of a second fluid engulfed within the first fluid, whereby to create an array of isolated samples of droplets in the storage chamber.

Multiple samples of droplets may be contained in a single fluid conduit, preferably wherein each sample is separated in the fluid conduit by a fluid compartment of a further fluid that is immiscible with the first and second fluids.

Preferably, the droplets of the second fluid are provided in the fluid conduit in an ordered arrangement or array. The fluid conduit may be loaded with one or more samples of droplets created using any of the method steps described above. Preferably, at least part of the surface is substantially planar.

According to yet another aspect of the invention there is provided a method of controlling fluid interactions, comprising: providing a first fluid conduit containing a first fluid; providing a second fluid conduit containing at least one fluid compartment of a second fluid, the second fluid being immiscible with the first fluid; arranging at least part of the first fluid conduit such that it is inclined away from the horizontal plane; and flowing the at least one fluid compartment into the first fluid conduit; wherein the at least one fluid compartment is arranged in the first fluid conduit such that it is not constrained by the inner walls of the first fluid conduit so as to use gravity to control movement of the at least one fluid compartment within the first fluid in the fluid conduit. Preferably, the at least one fluid compartment is arranged such that it is not constrained by the inner wall of the fluid conduit, for example so that the first fluid can flow around it.

The first fluid and second fluid may have different densities. Preferably, the density of the first fluid is greater than the density of the second fluid.

A fluid compartment of a third fluid may further be provided in the first fluid conduit, the fluid compartment being arranged to constrain movement of the at least one fluid compartment within the first fluid conduit.

Preferably, the further fluid compartment is arranged as a blocking compartment to constrain the droplets generated from the second fluid within the first fluid conduit; for example, the fluid compartment may the fluid compartment may span the internal diameter of the first fluid conduit, but for thin films of the first fluid that are present between the fluid compartment and the walls of the first fluid conduit. Preferably, the fluid compartment is provided in the first fluid conduit before flowing the second fluid into the first fluid conduit. The second fluid conduit may contain two or more fluid compartments of the second fluid each being arranged in the first fluid conduit such that they are not constrained by the inner walls of the first fluid conduit, preferably wherein the two or more fluid compartments are capable of merging together. The two or more fluid compartments may be substantially isolated from each other in the second fluid conduit with a fourth fluid, preferably which is miscible with the first fluid.

The fluid compartments in the second fluid conduit may be engulfed within a further fluid compartment of a fifth fluid, which is immiscible with the first, second, third and fourth fluids. Preferably, the further fluid compartment acts as a separating compartment.

The second fluid and fifth fluid may have different densities, preferably whereby the fluid compartments can merge together within the further fluid compartment. Preferably, the fifth fluid is denser than the second fluid. The method may further comprise the step of introducing into the first fluid within said fluid conduit, after the fluid compartment(s) of the second fluid, a further fluid compartment of a third fluid arranged to constrain movement of the at least one fluid compartment within the first fluid conduit.

Two or more separating compartments, each containing at least two fluid compartments, may be flowed into the first fluid conduit, preferably such that the fluid compartments within each separating compartment can merge together before the separating compartments merge together.

Preferably, flowing the second fluid into the first fluid conduit causes the first fluid to flow. The second fluid conduit may be inserted into the first fluid conduit. Preferably, the second fluid conduit has an outer diameter that is smaller than the inner diameter of the first fluid conduit.

Preferably, in each of the above-described methods the fluid conduit is generally straight, or at least has a portion that is generally straight. Preferably, the fluid conduit is a capillary tube. As used herein, the term "fluid conduit" preferably connotes a tube, pipe or channel for conveying fluids.

As used herein, the term "emulsion" preferably connotes a fine dispersion of minute droplets of one fluid in another fluid in which the emulsified fluid is not soluble or miscible.

Any feature in one aspect of the invention may be applied to other aspects of the invention, in any appropriate combination. In particular, method aspects may be applied to apparatus aspects, and vice versa. Furthermore, any, some and/or all features in one aspect can be applied to any, some and/or all features in any other aspect, in any appropriate combination.

It should also be appreciated that particular combinations of the various features described and defined in any aspects of the invention can be implemented and/or supplied and/or used independently. An example of the present invention will now be described with reference to the accompanying figures, in which: Figures 1 A to 1 D show a method of loading a fluid conduit with fluid samples;

Figure 2 shows a fluid conduit loaded with multiple fluid samples;

Figures 3A to 3D illustrate various stages of generating droplets from a fluid sample; Figures 4A to 4C show a method of loading a fluid conduit with fluid samples ready for the generation of droplets;

Figure 5A and 5B show droplets generated from two different samples separated by a blocking compartment;

Figure 6 shows a fluid conduit containing multiple arrays of monodisperse droplets;

Figure 7 shows an arrangement of fluid conduits containing multiple arrays of droplets; Figure 8A shows an example of droplet size distribution;

Figure 8B shows a droplet sample post thermal-cycling;

Figures 9A and 9B show a fluid conduit loaded with droplet samples being introduced to a storage chamber;

Figures 10A to 10E show various stages of storing droplet samples in a storage chamber;

Figures 1 1 A and 1 1 B show an array of different droplet samples arranged in a storage chamber;

Figure 12 shows a plurality of single cell reservoirs;

Figure 13 shows a method of merging fluid droplets;

Figure 14 shows a method of merging fluid droplets within a separating compartment; and

Figure 15 shows a further method of merging fluid droplets within a separating compartment.

Figures 1 A to 1 D show a sample preparation tube 102, being loaded with fluid samples from different fluid reservoirs 106, 1 10, 1 14, provided in a well plate 100, to form a plurality of fluid compartments 104, 108 in the sample preparation tube 102. The fluid samples 104, 108 may be the same or different fluids, preferably aqueous solutions, and may contain the reagents required for various analyses such as digital PCR or single cell analysis. Figure 1A shows a fluid sample being drawn into a sample "preparation" tube 102 from a fluid reservoir 106 to form a fluid compartment 104. Figure 1 B shows an immiscible fluid 1 12 from a different fluid reservoir 1 14 being loaded into the preparation tube 102 after the sample 104. The immiscible fluid 1 12 wets the inner wall of the preparation tube 102. The immiscible fluid 1 12 may be a fluorocarbon. Figure 1 C shows another fluid sample from another fluid reservoir 1 10 being loaded into the preparation tube 102 behind the immiscible fluid 1 12 to provide another fluid compartment 108. Figure D shows the preparation tube 102 loaded with a plurality of fluid samples 104, 108, 1 16, 1 18, all separated by immiscible fluid 1 12. A syringe pump, or similar device, may be used to draw the fluids into the preparation tube 102.

After it has been loaded with fluid samples 104, 108, 1 16, 1 18, the preparation tube 102 is fluidly connected to an "interaction" tube 120 via an interconnecting "delivery" tube 122, as shown in Figure 2. The interaction tube 120 is preferably arranged to be inclined away from the horizontal along at least part of its length. In this embodiment, the interaction tube 120 is generally straight and arranged generally vertically. The preparation tube 102 and delivery tube 122 are arranged to have a fluid-tight fluid connection. The delivery tube 122 has a smaller internal diameter than the preparation tube 102 so that a fluid compartment 104 being flowed into the delivery tube 122 from the preparation tube 102 becomes elongated as it flows through the delivery tube 122. Again, a syringe pump, or similar device, may be used to flow the fluid samples out of the preparation tube 102.

The interaction tube 120 contains a "carrier" fluid 124 that wets its inner walls. The carrier fluid 124 is immiscible with the fluid samples 104, 108, 1 16, 1 18 in the preparation tube 102. Furthermore, the carrier fluid 124 in the interaction tube 120 contains a surfactant (i.e. is "surfactant-laden") to help facilitate the formation of droplets of the fluid samples 104, 108, 1 16, 1 18, as will be described further on. The carrier fluid 124 in the interaction tube 120 is, preferably, similar to the separating fluid 1 12 in the sample tube 102 (i.e. the two fluids should be miscible). In one example, both immiscible fluids 1 12, 124 may be fluorocarbons with the preparation tube 102 containing a mixture of HFE7500 and FC40 at a ratio of 4:1 , and the interaction tube 120 containing pure FC40 as the carrier fluid 124 with 1 % surfactant (e.g. Pico-Surf™). The interaction tube 120 has an inner diameter that is greater than the outer diameter of the delivery tube 122 to provide a gap 126 to allow carrier fluid 124 to escape from the interaction tube 120 when the fluid compartments 104, 108, 1 16, 1 18 are flowed into the interaction tube 120, to avoid an excessive build-up of pressure.

Figures 3A to 3D show the generation of droplets from a fluid sample 104 (in this example in the form of a fluid compartment 104) by flowing the fluid into a surfactant- laden immiscible fluid contained in an interaction tube 120. As the fluid compartment 104 emerges into the interaction tube 120, the surfactant in the carrier fluid 124 reduces interfacial tension between the aqueous fluid sample 104 and the fluorocarbon carrier fluid 124, which creates instabilities at the surface boundary of the fluid compartment 104 that causes the generation of droplets 128 as shown in Figure 3A. The surfactant (a stable amphiphile) surrounds each droplet formed in the interaction tube 120 and thereby acts as a barrier against coalescence of the generated droplets 128.

The carrier fluid 124 wets the inner walls of the interaction tube 120 and thereby prevents the sample fluid 104 from wetting those inner walls, which may otherwise inhibit the surfactant in the immiscible carrier fluid 124 from forming the droplets 128. As shown in Figure 3B, the interaction tube 120 may also contain one or more fluid compartment 130 of a fluid that is immiscible with both the fluid sample 104 and the carrier fluid 124. This fluid compartment may be described as a blocking compartment 130 as it is arranged to constrain / control the flow of the droplets in the carrier fluid 124. For example, the blocking compartment 130 may have a length greater than the inner diameter of the interaction tube 120 such that the immiscible fluid droplets do not flow freely in the interaction tube 120. In other words, the velocity of a blocking compartment is constrained by the mean velocity of the carrier fluid 124 being pumped through the interaction tube 120 since the blocking compartment almost fills the diameter of the interaction tube 120 but for thin films of the carrier fluid 124 that wets the inner walls.

Figure 3B shows droplets 128 of the emulsified fluid compartment 104 rising through the denser carrier fluid 124 until they come to rest against a blocking compartment 130, where they collect together. The size of the resultant droplets 128 is dependent upon the inner diameter of the 'delivery' tube that introduces the fluid compartment 104 into the interaction tube 120, which in this embodiment is the delivery tube 122. Droplets 128 typically have a diameter between about 1 to 2.5 multiples of the internal diameter of the delivery tube 122. The size of the droplets 128 may be controlled by varying the flow rates of the fluid compartment 104, by varying the interfacial tension (e.g. varying the surfactant concentration), or by varying the viscosity/density of the immiscible (carrier) fluids 1 12, 124. During experimentation, it was found that increasing the flow rate by around three-fold did not affect the diameter of the droplets 128 by more than 10% for monodisperse (emulsion) droplets 128. As the droplets 128 form within the interaction tube 120 they consume the surfactant in the carrier fluid 124. As the fluid compartment 104 flows into the interaction tube 120, an equal volume of carrier fluid 124 is displaced by escaping through the gap 126 at the entrance to the interaction tube 120, as can be seen in Figure 3D. If the delivery tube 122 is inserted into the interaction tube 120, the tip of the delivery tube 122 that penetrates into the interaction tube 120 may be cut at an angle to improve the stability of the droplet 128 generation process. The flow of carrier fluid 124 over the tip of the delivery tube 122 as it escapes ensures that the surfactant concentration in the carrier fluid 124 in the vicinity where the fluid compartment 104 emerges into the interaction tube 120 remains sufficiently high to form the droplets 128.

If the internal diameter of the delivery tube 122 is narrower than the internal diameter of the interaction tube 120, the droplets 128 generated from the fluid sample are not constrained by the walls of the interaction tube 120 (e.g. their diameter is less than the internal diameter of the interaction tube 120), and thus movement of the droplets 128 is driven by buoyancy rather than constrained by the inner walls of the interaction tube 120. Furthermore, if the carrier fluid 124 has a greater density than the fluid samples 104 the unconstrained droplets 128 will rise in the carrier fluid 124. Of course, as the fluid sample is flowed into the interaction tube 120, the generated droplets 128 may in any case flow through the carrier fluid 124 towards the blocking compartment 130.

Figure 3D shows a sample emerging into the interaction tube 120, and the resulting droplets 128 rising though the carrier fluid 124 until they reach the blocking compartment 130 (or at least other previously emulsified droplets 128 of the same sample that are constrained by the blocking compartment 130). The generation of droplets 128 here continues until the fluid sample 104 is exhausted. The delivery tube 122 can then be withdrawn from the interaction tube 120 and a new immiscible blocking compartment drawn into the interaction tube 120, followed by more surfactant-laden immiscible carrier fluid.

Figure 4A shows a blocking compartment 130 being loaded into the interaction tube 120 to isolate a previous sample of droplets 104 higher up (not shown) in the interaction tube 120. Figure 4B shows the interaction tube 120 then being loaded with carrier fluid 124 ready for the next fluid sample 108 to be emulsified. Figure 4C shows an emulsified fluid sample 104 sandwiched between two blocking compartments 130, 132 in the interaction tube 120. Once replenished with fresh carrier fluid 124 (and surfactant), the delivery tube 122 can then be re-connected with the interaction tube 120 and another fluid sample flowed into the interaction tube 120, as described above. The process can be repeated as many times as necessary, for example until all of the fluid samples 104, 108, 1 16, 1 18 that were originally drawn into the preparation tube 102 have been flowed into the interaction tube 120.

Figure 5A shows a second fluid sample 108 being flowed into the interaction tube 120 to generate droplets 134 after a first sample of droplets 128 has been generated from a first fluid sample 104. The preparation tube 102 contains further fluid samples 1 16, 1 18 that are waiting their turn to be flowed into the interaction tube 120. Figure 5B shows a close-up view of two the two samples of droplets 128, 134 constrained (and separated) within the interaction tube 120 by blocking compartments 130, 132. Figure 6 shows an interaction tube 120 containing several arrays of mono-disperse droplets 128, 134 that have been generated from different fluid samples 104, 108 using the method described above, with immiscible fluid blocking compartments 130, 132 separating the samples of droplets 128, 134 into discrete compartments to avoid any cross-contamination. The droplet sizes shown are about 90-100 microns in diameter stored within an interaction tube 120 having an inner diameter of 600 micron.

As mentioned before, the above-described process of generating droplets 128, 134 from fluid samples 104, 108 can be performed repeatedly by replenishing the interaction tube 120 for each fluid sample. This allows droplets to be generated from multiple samples in parallel using a single syringe pump. It should also be noted that a delivery tube 122 may not be required for low numbers of fluid samples, or for single sample reactions, for example, and new delivery tubes may be utilised for subsequent samples to prevent cross-contamination. It will be understood by a skilled person that the preparation tube 102 may be configured such that it can be fluidly connected directly to the interaction tube 120, such that in effect it becomes the delivery tube 122. The tubes 102, 120, 122 can be formed using many different geometries to configure them, such as a pulled pipette, or a conical shaped device having a small inlet, for example.

Also, the fluid samples need not be provided as fluid compartments; a continuous supply of a fluid sample could be flowed into the interaction tube 120 for a predetermined time to generate a desired droplet sample, for example. The carrier fluid (and surfactant) would however occasionally need to be replenished, as discussed above.

The interaction tube 120 may receive fluid samples from multiple delivery tubes 122, so that a plurality of samples of drops 128, 134 may be generated substantially simultaneously. The multiple delivery tubes 122 may all be inserted into an end of the interaction tube 120, similar to the method described above.

Alternatively, a delivery tube (not shown) may be configured to have multiple outlets through which the fluid sample can flow into the interaction tube 120. Ideally, the multiple ports would be provided at an end of the delivery tube, with said end of the delivery tube being inserted (at least partly) into the interaction tube 120. Multiple streams of fluid droplets can thereby be generated, preferably substantially simultaneously, from a fluid sample as it is flowed into the interaction tube 120.

Alternatively, an interaction tube (not shown) may be configured to have multiple ports for making fluid connections with, or preferably receiving, multiple delivery tubes 122. For example, a plurality of ports may be provided around the circumference of an interaction tube, each configured to receive a fluid sample from a delivery tube 122, which is preferably inserted into the port to provide a manifold-like configuration to generate multiple arrays of sample droplets. As mentioned before, a syringe pump (not shown) may be used to load the carrier fluid 124 and blocking compartments into the interaction tube 120 and/or to flow the sample fluid compartments 104, 108 from the sample tube 102 into the interaction tube 120 to create the droplets.

Figure 7 shows how the droplets 128, 134 generated from the fluid samples 104, 108 may be stored in the interaction tube 120, where they may undergo thermal cycling for a biological application, e.g. polymerase chain reaction (PCR), as will be described further on. A suitable thermal cycle may alternate between 90-95C, then 40-60C then 72C, for about 30-40 cycles, for example.

Figure 8A shows the droplet size distribution for the droplets 128, 134 generated using the method described above, wherein the preparation tube 102 had an internal diameter of 380 microns, the delivery tube 122 had an internal diameter of 50 microns, and the interaction tube 120 had an internal diameter of 620 microns. The droplets 128, 134 generated had radii ranging from about 44 microns to about 56 microns, with the majority of droplets having a radius of about 48 microns. Figure 8B shows a sample after forty rounds of thermal cycling between 40-95 degrees Celsius. Directly after their generation, droplets 128, 134 may be deposited onto a substrate or into a test tube, for example, and monitored for a short period of time. However, monitoring time is limited because, once the engulfing carrier fluid 124 evaporates, the droplets 128, 134 will rapidly evaporate also. To prevent the carrier fluid 124 from evaporating, it may be covered by an immiscible "storage" fluid 138, which is immiscible with both the droplets 128, 134 and the engulfing carrier fluid 124 within the interaction tube 120. If the sample droplets 128, 134 are an aqueous fluid, and the carrier fluid 124 is a fluorocarbon, then the storage fluid 138 may be a hydrocarbon, such as mineral oil, for example.

Figure 9A shows a storage chamber 136 having a planar surface 140 (e.g. a glass substrate or cell culture well plate) that is covered by storage fluid 138. Figure 9B shows the carrier fluid 124 and a sample of droplets 134 being flowed out of the interaction tube 120 onto the surface 140 of the storage chamber 138, beneath the surface of the immiscible storage fluid 138, to preserve the droplets for long term monitoring. Ideally, the tip of the interaction tube 120 is first brought into close proximity / direct contact with the surface 140 of the storage chamber 136.

Figures 10A to 10C show steps for flowing a sample of droplets 134 and engulfing carrier fluid 124 out of the interaction tube 120 and onto the planar surface 140 of the chamber 136, beneath the surface of the storage fluid 138, for long term monitoring. In Figure 10D, the carrier fluid 124 that wetted the inner wall of the interaction tube 120 has wetted (effectively attached to) the surface 140 of the storage chamber 136 such that the droplets 134 remain engulfed (and are thereby preserved) by the immiscible carrier fluid 124, which is itself preserved by the immiscible "storage" fluid 138 in the chamber 136. Figure 10E shows a plan view of the droplets 134 stored in carrier fluid 124 on the surface 140 in a tightly packed array.

Figure 1 1 A shows a second sample of droplets 128 that has been flowed from the interaction tube 120 onto the surface 140 of the storage chamber 136 to store it beneath the storage fluid 138 next to the previous sample of droplets 134. In this way, multiple samples of droplets 128, 134 may be stored in the same storage chamber 136 in an array, where each sample of droplets 128, 134 is preserved (and isolated from other samples) by the engulfing carrier fluid 124 from the interaction tube 120 such that many samples can be stored together, as shown in Figure 1 1 B.

In this way, pico- to nano-sized sample droplets can be stored in a micro-litre droplet of carrier fluid 124 attached to the surface 140 in an immiscible storage fluid 138. Evaporation is prevented by the volume of immiscible storage fluid 138 which covers the immiscible carrier fluid 124 engulfing each sample of droplets 128, 134. Many such samples of droplets 128, 134 attached to the surface 140 of the storage chamber 136 may be deposited within a single dish, thereby allowing, for example, fluorescent detection to be applied to the samples. Figure 12 shows the use of this storage method for the creation of single cell reservoirs by emulsifying a large sample of cells in media, demonstrating the potential wide diversity of applications. Alternatively droplets may be integrated by flowing them through a smaller tube to form a line of one droplet at a time passing a sensor in a similar fashion to fluorescence activated cell sorting (FACS). In another aspect of the invention, a preparation tube may be loaded with fluid samples that have been merged together to form fluid compartments. The merging of a plurality of sample fluid drops, which may contain the reagents required for various analyses such as digital PCR or single cell analysis, is possible by flowing 'trains' of multiple samples in the form of fluid compartments (or "drops") against gravity.

Figure 13 shows an embodiment where the trains of multiple samples are flowed against gravity in the vertical direction from a sample preparation tube 202 into an interaction tube 204, where they merge when they come into contact due to the absence of a stabilizing amphiphile (surfactant). This is in contrast to the above-described method of generating fluid droplets, where a stable amphiphile surrounds each generated droplet in the interaction tube, thereby preventing the fluid droplets from merging / coalescing once generated. To load the sample tube 202, it is initially filled with a carrier fluid 206 that preferentially wets the inner surface of its walls, such as fluorocarbon. Using suction, a train of fluid sample droplets 208, 210, 212 is drawn from one or more well plate 200 into the sample tube 202. First, a fluid that is immiscible with both the carrier fluid 206 and the aqueous sample droplets 208, 210, 212 is drawn into the sample tube 202 to provide a "blocking compartment", followed by the fluid sample droplets 208, 210, 212 to be merged. The droplets 208, 210, 212 are, preferably, drawn from different reservoirs / well plates 200 into the interaction tube 204, where they are separated by the carrier fluid 206, to prevent cross-contamination.

Figure 13(i) shows the preparation tube 202 having a smaller diameter than the interaction tube 204, such that the preparation tube 202 can be inserted into the interaction tube 204. The interaction tube 204 is also initially filled with a carrier fluid 216 that preferentially wets the inner surface of its walls. As with the droplets generation method previously described, the carrier fluids 206, 216 do not necessarily need to be identical, though they should be miscible. The fluid samples 208, 210, 212 that have been sucked into the preparation tube 202 are then flowed into the interaction tube 204. The first blocking compartment 214 is arranged such that its movement within the interaction tube 204 is constrained by the mean velocity of the carrier fluid 206 within the interaction tube 204. For example, the blocking compartment may be arranged to span the internal diameter of the interaction tube 204, but for thin films of the carrier fluid 206, 216 that are present between the sample tube 202 and the interaction tube 204.

The sample droplets 208, 210, 212 subsequently entering the interaction tube 204 have a length that is less than the internal diameter of the sample tube 202, and therefore the relative density of the carrier fluid 216 in the interaction tube 204 determines whether the sample droplets 208, 210, 212 rise or remain at the junction between the sample tube 202 and the interaction tube 204. If the density of the carrier fluid 216 is greater than that of the first sample droplet 208 then the sample droplet 208 will rise until it abuts the fluid interface of the first immiscible blocking compartment 214, which confines its motion. As subsequent sample droplets 210, 212 enter the interaction tube 204 they rise through the denser carrier fluid 216 until they come into contact with the first (or previous) sample droplet 208 of their train. When the droplets 208, 210, 212 come into contact they may merge spontaneously to form a single combined droplet 226, as shown in Figure 13(iii). Alternatively, the droplets 208, 210, 212 may require the assistance of an external perturbation (not shown). Such perturbation may be thermal, mechanical or electrical such that any number of droplets may be merged one by one or all together, using external perturbation.

Once all of the droplets 208, 210, 212 in the train have merged into a single combined droplets 226, the sequence may be repeated with a (further) sample tube containing first another blocking compartment 224 followed by a train of one or more sample droplets 218, 220, 222, as described above. The method may also be used to load an interaction tube 204 with one or more single samples droplets separated by immiscible droplets. Figure 13(iii) shows the first train of droplets 208, 210, 212 merged (spontaneously) to form a single droplet 226. Figure 13(iv) shows the first two trains of droplets have merged into single droplets 226, 236, and a third train of droplets are about to merge, and a fourth train of three isolated droplets is waiting in the sample tube 202, ready to be flowed into the interaction tube 204. Figure 14 illustrates another embodiment of method of merging droplets similar to the embodiment described above with reference to Figure 13, but where the sample droplets 308, 310, 312 in sample tube 302 are engulfed within a fluid "separating" compartment 336 of an immiscible fluid, which is a different phase to a carrier fluid 316 contained in an interaction tube 304.

In this embodiment, when the sample droplets 308, 310, 312 exit the narrower sample tube 302 and enter interaction tube 304 they become generally spherical and remain enclosed within the separating compartment 336, as shown in Figure 14(i). In this embodiment, the density of the separating fluid 336 is less than that of the sample droplets 308, 310, 312 and also of the carrier fluid 316. Therefore, the separating compartment 336 will rise in the carrier fluid 316 until it reaches a blocking compartment 314, while the sample droplets 308, 310, 312 will merge at the lower end of the separating compartment 336 that engulfs them to form a single combined droplet 326, as shown in Figures 14(ii) and 14(iii). Multiple trains of fluid droplets can be merged within separating compartments 336 using this method, wherein each can have its movement within the interaction tube 304 confined by a blocking compartment 314.

Figure 15 shows yet another embodiment, similar to the embodiment described with reference to Figure 14 in that each train of droplets 408, 410, 412 is engulfed within a separating compartment 436 in the sample tube 402, as shown in Figure 15(i). In Figure 15(ii), the first two droplets 408, 410 have merged into a single combined droplet 426 while still in the sample tube 402, due to the fluid of the separating compartment 436 being denser than the sample droplets 408, 410, 412.

Furthermore, in this embodiment a single immiscible "blocking" compartment 414 is utilised to control only the flow velocity of the first train of droplets in the interaction tube 404, which blocking compartment 414 may then be used to add complexity to the final droplet formation. Figures 15(iii) to 15(v) illustrate how the absence of a blocking compartment between the first train of droplets and subsequent trains of droplets allows a separating compartment 440 in a subsequent train of droplets 442 to catch up and merge with the slower moving separating compartment 436, the velocity of which is constrained by the blocking compartment 414. Once the separating compartments 436, 440 have merged into a single separating compartment 444, the combined droplets 438, 442 from each separating compartment 436, 440 will then merge to form a single combined droplet 446. One or more further blocking compartments may of course be used to separate multiple merged samples.

As mentioned above, the velocity of the merged droplet 438 remains confined by the blocking compartment 414 until, with flow, the next separating compartment 440 catches up the first separating compartment 436 and merges with it, as shown in Figure 15(iv). The lower density of the sample droplets 438, 442 relative to the separating fluid allows the sample droplets 438, 442 to come into contact within the separating compartment 444 and merge, as shown in Figure 15(v).

This process may be used to merge many sample droplets 408, 410 in a single separating compartment 436, for as long as the separating compartment remains attached to the open end of the sample tube 402, as shown in Figure 15(ii). When the separating compartment 436 detaches from the sample tube 402 then another train of droplets may be introduced into the interaction tube 404 and merged.

By having the length of the resultant separating compartment 436 (created at the junction of the smaller diameter sample tube 402 and larger diameter interaction tube 404) less than the diameter of the interaction tube 404, and the density of the carrier fluid 416 in the interaction tube 404 greater than that of the fluid of the separating compartment 436, the droplets will rise (i.e. they will be buoyant) due to gravity and will also enable merging between sequential trains of droplets in this way.

From the example shown in Figure 15, it can be understood that once the (merged) droplets are greater in diameter than the interaction tube 404, the relative density of the fluids has negligible effect (depending on the tube diameter) on the motion of the droplets. Their velocity is then instead primarily controlled by the capillary number of the interaction tube 404 and the viscosity ratio between the droplet 446 and the engulfing fluid 416. The density of the separating fluid 444 here can however be greater/less than the density of the sample drops, so that the sample droplets rise/fall, respectively, in the separating compartments 436, 440 until they abut and merge, as shown in Figures 15(ii) and 15(iii).

These methods of merging droplets, described with reference to Figures 13 to 15, may also be used to load a preparation tube with samples (e.g. fluid compartments) from which to generate droplets using the method described above with reference to Figures 2 to 12. As such, the interaction tube 404 would be the functional equivalent to the sample tube 102, though possibly without the presence of blocking compartments (though they may still be used to separate multiple samples).

The work leading to this invention has received funding from the People Programme (Marie Curie Actions) of the European Union's Seventh Framework Programme (FP7/2007-2013) under REA grant agreement no 333848.