GB2474228A - Microfluidic device for removing oil from oil separated aqueous sample droplets - Google Patents

Abstract

The device comprises a sample channel 11 with an oil outlet channel and an aqueous flow outlet channel. The dimension of the oil outlet channel opening is such as to prevent water from passing through it. The micro-fluidic device also includes an encapsulation device 1 for separating portions of a pre-separated aqueous analyte stream 4 passing along a conduit using oil 2. The oil separated droplets 12 being fed into the aforementioned device, with the aqueous material 14 being fed into a further separation device 21. The analyte 4 is preferably peptide (protein), nucleic acid (DNA, RNA), amino acid or cells. Preferred pre- and post-separation devices 21 include high pressure liquid chromatography (HPLC), liquid chromatography (LC), isoelectric focussing, capillary electrophoresis (CE), capillary gel electrophoresis (CGE), isotachoporesis or micellar electrokenetic chromatography, coupled with ultraviolet (UV), fluorescence, phosphorescence, staining, radioactivity, refractive index or vibrational infrared (IR) spectroscopy detector.

Description

2-DIMENSIONAL SEPARATION

TECHNICAL FIELD

This invention relates to a microfluidic device for compartmentalising chemically separated components in 2-dimensional (2D) separations and for interfacing two orthogonal separation methods.

BACKGROUND ART

Chemical separation techniques play an important role in proteomics, genomics, metabolomics and a range of other biochemical fields, such as forensics. Proteomics is an important emerging area for bioanalytical is chemistry and deals with the large-scale study of proteins, particularly the study of their structures and functions. The most widely used analytical methods in protein chemistry are based on electrophoresis and chromatography.

In liquid chromatography (LC), a sample to be analysed is forced through a column that is packed with irregularly or spherically shaped particles or a porous monolithic layer (stationary phase) by a liquid (mobile phase). When high pressure is used to drive the liquid through the column, the technique is referred to as high performance liquid chromatography (HPLC).

The relative polarity of the mobile and stationary phases is selected in order to achieve the best possible resolution of analyte components. The liquid emerging from the column is termed the eluate'. Molecular components within the eluate will be ordered by their respective polarity.

Gas chromatography (GC) is a separation technique in which the mobile phase that passes through the column is a gas, e.g. helium. The stationary phase (often a liquid silicone-based material) may either be a solid matrix inside a packed column or adhered to the inside of a small-diameter a capillary column. The high temperatures used in GC make it unsuitable for high molecular weight biopolymers or proteins, as the heat will denature them.

Analyte diffusion in the mobile phase results in peak broadening and consequently lower resolution of analyte components. This can result from simple longitudinal diffusion of an analyte band within the column. A further contributor to peak broadening includes the parabolic flow profile observed with pressure driven flow. This contribution can be reduced by decreasing the column diameter or the particfe size in the stationary phase.

Another cause of peak broadening in liquid chromatography is diffusion of the analyte into stagnant pores and/or in stationary phase material. Large particles and loosely packed columns further contribute to analyte dispersion and peak broadening ("Eddy diffusion"). This factor is absent in open tubular columns (e.g. in capillary electrophoresis).

Besides factors arising from the column, band broadening is also caused by the volume of the analytical apparatus and will depend on the injected volume, the volume of connective tubing, and detector volume and response time. These extra-column contributions become more significant on downscaling columns, although injection variance generally improves as the scale is decreased. Thus the smaller the apparatus, the more important for good separation the connectors between the column and detector become.

A general problem observed with elution is that if one solvent or one temperature is used, then either earlier peaks will be well resolved and later peaks will be broad, or earlier peaks will be broad and later peaks will be well resolved. In order to address this, solvent strength can be varied in LC (gradient elution) and a temperature ramp may be applied in GC.

Electrophoresis is the migration of dispersed materials relative to a fluid under

the influence of a uniform electric field.

In capillary electrophoresis (CE) a sample containing analyte is injected into a capillary to which an electric field is applied. CE is used to separate ionic species based on their charge and frictional forces. Electrically charged analytes move in a conductive liquid medium under the influence of an electric field. The technique of capillary electrophoresis (CE) was designed to separate species based on their size to charge ratio in the interior of a small capillary filled with an electrolyte.

The electrolyte solution may be propelled by electroosmotic flow (EOF) and the different ions are separated by electrophoresis. Generally, a flat flow profile is achieved due to the high voltage applied and this allows for good resolution of analyte peaks.

The electrolyte solution moves towards the cathode. Negatively charged analyte molecules move towards the positively charged anode and positively is charged analyte molecules move towards the negatively charged cathode.

The velocity of particle movement depends on both charge and frictional forces. That is, small, highly charged particles are more mobile. The mobility of analyte molecules is also affected by the pH of the buffer or electrolyte.

Capillary gel electrophoresis (CGE) is a technique used to separate ionic analyte components based on their molecular weight.

The gel used in the capillaries is a crosslinked or entangled polymer, for example, polyacrylamide or agarose. Separation of the analyte components occurs when an electric field is applied, as the gel obstructs the migration of larger molecule to a greater degree than smaller molecules.

CGE is used in various biochemical fields including genetics, forensics, molecular biology and microbiology. CGE is used for the separation of protein molecules, deoxyribonucleic acid (DNA), ribonucleic acid (RNA), or oligonucleotides, using an electric current applied to a gel matrix.

The results can be analysed quantitatively by visualizing the gel with UV light and a gel imaging device. The intensity of the band or spot of interest is measured using computer software and can be compared against known markers loaded on the same gel.

Nucleic acids migrate from negative to positive electrodes, due to the negative charge carried by the sugar-phosphate backbone. Gel electrophoresis of large DNA or RNA is usually done by agarose gel electrophoresis.

Proteins can have varying charges and complex shapes. Therefore, they may not migrate into the polyacrylamide gel at similar rates, or at all, when placing a negative to positive electromotive force (EMF) on the sample. Proteins may be denatured in the presence of a detergent such as sodium dodecyl sulfate/sodium dodecyl phosphate (SDS/SDP), which coats the proteins with a negative charge. However, positively charged proteins or peptides may is also be used.

Where a single technique is not able to separate some analytes, a series of unresolved peaks may subsequently be resolved by using a second separation method having different physico-chemical properties from the first method. Since the mechanism of retention on this second solid support is different from the first dimensional separation, it can be possible to separate compounds that are not resolved by one-dimensional chromatography.

After a first separation, which may for example be by HPLC, GO, or CE, the eluate is directed onwards to a second separation. In the second dimension, separation is carried out orthogonally to the first separation. For example, if an electric potential is applied in both the first and the second separation, in the second separation, it will be applied at a 90 degree angle from the first field. Thus, component peaks observed in the first separation can be further analysed and previously masked peaks can be resolved into separate component peaks.

The most important tool in separating very complex mixtures of proteins is 2D gel electrophoresis.

If negatively charged, the proteins migrate towards the side of the gel where for example the positive etectrode is placed, at a velocity proportional to their mass-to-charge ratio. Migration through the gel is slowed by frictional forces.

The polymeric gel separates the proteins on the basis of their molecular weight. Smaller proteins are able to move more easily through the gel and reach lower regions of the gel than larger proteins, which are retained higher in the gel.

The result of this second, orthogonal separation is a gel with proteins spread out on its surface. These proteins can then be detected by a variety of known means, such as staining, e.g. coomassie blue or ethidium bromide staining (for DNA detection). Analyte detection may also be carried out by any of the following methods: UV (fixed or variable wavelength), fluorescence, vibrational spectroscopy, electrochemical (for electro-active compounds), refractive index, evaporative light scattering, mass spectrometry, or radioactivity.

At the current time, 2D polyacrylamide gel electrophoresis methods are considered to be the gold-standard in separating complex protein mixtures.

However, other 2D column-based separation techniques including LC-LC, LC-CE and CE-CE, have also been developed to provide for automated analysis of a broader range of biomolecules.

In summary, 2D separation techniques allow for analyte component resolution according to two molecular properties.

Another useful resolution technique is termed "heart cutting", which refers to the technique of isolating of a pair (or more) of unresolved salutes from one separation and placing the mixture on another column which does resolve them. That is, if in a first separation a major component masks the presence of a minor component, the fraction of the solute comprising these analyte components is collected and re-run using either the same separation technique (in the case that the masking occurred merely because of the greater quantity major component) or a different separation technique having is a different property (in the case that the masking occurred due to similar retention times of the components on a specific stationary phase). This could, for example, be to first run the sample on a purely dispersive stationary phase, followed by a polar stationary phase.

There has been a continued desire for using smaller sample quantities, which has led to research being carried out into transferring many known analytical techniques to a sub-millimetre scale, e.g. Lab-on-a-Chip technology.

Converting 2D methods to a capillary format is highly attractive since capillary techniques require only minute quantities of sample. Further, the high surface-to-volume ratio of the capillary results in rapid heat dissipation.

Maintaining an even temperature minimizes heat-induced peak broadening (known as Ohmic or Joule heating) in electrophoresis.

Over the past few years many research groups have tried to transfer 2D separation techniques to a microfluidic, i.e. microscale, format but this has often proved difficult due to the problems associated with transfer of material between the two dimensions.

However, recent advances in microfluidic control architecture have significantly improved the ability to work with materials on a microscale, for example facilitating the precise temporal and spatial manipulation of single droplets and functions such as sorting, splitting and merging for complex analyses.

Edgar et a!. (2006) [J. S. Edgar, C. P. Pabbati, R. M. Lorenz, M. Y. He, G. S. Fiorini and D. 1. Chiu, Anal. Chem., 2006, 78, 6948-6954] attempted to use capillary efectrophoresis (CE) coupled with laser induced fluorescence to analyse the contents of single femtolitre-volume aqueous droplets. However, since both the droplet and the carrier oil are infused into the separation channel, it was found to be difficult to prevent wetting of the separation channel, even with good pressure control. Introduction of oil into the separation column means that electroosmotic flow (EOF) cannot be recovered once the hydrophobic channel surfaces have been exposed to the oil. As a result, this disclosure does not demonstrate the capability of continuously loading droplets into the separation channels or columns, thereby making the platform inappropriate for any 2D separations.

Roman et a!. (2008) [G. T. Roman, M. Wang, K. N. Shultz, C. Jennings and R. T. Kennedy, Anal. Chem., 2008, 80, 8231-8238] disclosed a method for sampling and electrophoretic analysis of aqueous plugs segmented in a stream of immiscible oil. In that method, an aqueous buffer and an oil stream flow parallel to each other to form a stable virtual wall in a microfabricated K-shaped fluidic element. As aqueous sample plugs in the oil stream make contact with the virtual wall, i.e. the phase boundary, coalescence occurs and the sample is electrokinetically transferred to the aqueous stream. However, such a virtual wall is not easy to maintain, and the sampling ratio cannot exceed 5% of the original volume. Therefore this method cannot achieve whole loading of the droplet sample for further separation. The inability to analyse whole droplets makes this method unsuitable for most applications in separation science. This is because analysing only a sample of a droplet assumes a uniform dispersion of all analyte molecules within the droplet.

However, such uniformity cannot be assumed as, after a first separation, droplets may contain analyte components in bands rather than being uniformly dispersed within a droplet. This would most notably be the case in protein or peptide analysis, and with other large molecules, and polar analyte components.

By integrating droplet generation induced by electroosmotic flow (EOF) with chemical separation, Edgar et a!. (2009) [J. S. Edgar, G. Milne, Y. Q. Zhao, C. P. Pabbati, D. S. W. Lim and D. T. Chiu, Angewandte Chemie-International Edition, 2009, 48, 2719-2722] have used droplets in an immiscible oil phase to spatially confine components separated by CE. This approach overcomes molecular diffusion by confining the separated bands in a series of droplets.

This disclosure does not teach the removal of oil from the eluate and consequently, second dimension separations were not achieved.

X. Niu et a!. [X. Niu, S. Gulati, J. B. Edel and A. J. deMello, Lab on a Chip, 2008, 8, 1837-1841] disclose a method of controllably merging aqueous microdroplets within a segmented flow microfluidic device. By using a merging chamber containing rows of pillars separated by distances smaller than the droplet dimension, the surface tension can be altered in such a way as to allow one or more droplets to merge into one larger droplet. A first droplet enters the merging chamber and will slow down and stop within the pillar section. A successive droplet entering the pil'ar arrangement will merge with the first droplet. The first droplet will continue to merge with successive droplets until the surface tension is overwhelmed by the hydraulic pressure.

Once this occurs, the new combined droplet will exit from the pillar arrangement and proceed out of the merging chamber into an exit channel.

The resulting combined droplets are still segmented by the carrier oil.

The approach of the present invention addresses all of the above-mentioned problems, can achieve whole loading of the droplets and also avoid oil contamination of the separation channels. Importantly, the operation of the present invention only relies on the design of the microfluidic device; in other words, it is passively controlled and easy to achieve, which simplifies the operation of the system dramatically. In the present invention, the microfluidic device enables the direct interfacing of a first and a second separation dimension, without losing the resolution achieved by the first separation.

SUMMARY OF THE INVENTION

In recent years, a diversity of two-dimensional (2D) separation methods have been developed in an effort to efficiently probe and resolve complex analytical mixtures, such as proteins expressed by cells or organisms. By coupling orthogonal separation techniques, enhanced analytical resolution and peak capacities defined by the product of the component peak capacities can be provided. Unfortunately, realising high analytical performance is far from trivial, due to the inherent difficulties in spatially and temporally coupling two different separation techniques (using for example traditional valves) when only extremely small amounts of sample are available.

The need to integrate multiple operations and achieve high throughput analysis suggest that chip-based or microfluidic platforms may be suitable for performing the comprehensive analyses demanded in -omics sciences.

However, transferring materials between two orthogonal dimensions has remained a significant challenge.

Droplet-based microfluidics has recently emerged as a valuable instrumental platform for performing high throughput chemical and biological experiments.

In such systems, droplets are made to spontaneously form when laminar streams of aqueous reagents are injected into an immiscible carrier fluid, either at a 1-junction or in a flow focusing geometry.

The present invention relates to a new approach for coupling two separation techniques (or mechanisms) using a dynamic microdroplet interface. This approach does not sacrifice resolution in any single dimension and allows analysis of nanoliter to femtoliter volumes without the need for valves.

The present invention addresses the problems occurring when transferring io materials between two orthogonal dimensions by using microdroplet-based microfluidics as an efficient interface in two-dimensional separations.

The present invention efficiently integrates two separation dimensions utilising droplet generation after the first dimension, with oil depletion and droplet is merging prior to the second dimension. This combination forms a fully functional droplet connector for two-dimensional separations. It further allows for a first and a second separation to have a direct interface rather than merely being carried out sequentially.

After an analyte solution undergoes a first separation, the analyte solution flows along a first channel, which is connected to a second channel, for example via a 1-junction. In this application, a "T-junction" is defined to comprise a first, straight channel, and a second channel that joins onto the first channel at any angle between 100 and 170°, preferably between 30° and 150°.

An oil flows along the second channel. When the analyte flow enters the oil channel, it forms droplets that are carried along in the oil flow. The size of these droplets may be controlled to be of micro-, nano-, pico-, or femtolitre volume. The oil prevents the components within the droplets from diffusing outside of the droplet. The separated components from the first dimensional separation contained in these droplets, may be transferred by the carrier oil along the oil flow channel to a second separation device. Prior to the analyte solution being fed into the second separation apparatus, the carrier oil is removed and the droplets that were previously dispersed in the oil phase are merged into an aqueous stream. Because the oil is only removed immediately prior to the second separation, there is virtually no opportunity for molecular diffusion of the analyte components to occur within the analyte solution. Accordingly, the resolution achieved by the first separation is not lost during the transfer of the analyte between the first and the second separation apparatus.

Alternatively, the separated components from the first dimensional separation may be stored as contained in the specific droplets dispersed in the oil phase in a microfluidic droplet connector, for later use in further separation techniques or further analysis.

Using two-dimensional capillary separation techniques, for example HPLC- CE, nanolitre-sized droplets can be used as an effective tool in coupling two-dimensional separations. Using a microfluidic droplet connector, the peaks from a first dimension separation can be segmented and stored into nanoliter volume droplets. After oil filtering and droplet merging, these droplets can be loaded individually into a second dimension for comprehensive or heart-cutting' separations resulting in potential enhancements in resolution.

Thus, the present invention makes two-dimensional separation possible, wherein the droplets obtained from a first separation are able to be transferred to a second dimensional separation method directly upon removal of the carrier oil. Analyte molecules can be encapsulated and stored inside the droplets, without evaporation or contamination between droplets.

BRIEF DESCRIPTION OF THE FIGURES

Figure 1 shows (a) a first dimension capillary HPLC separation and segmentation; and (b) droplet recombination to a continuous flow followed by the second dimension CE separation.

Figure 2 shows tubing containing segmented droplets.

Figure 3 shows a section of a fabricated device.

Figure 4 shows the complete merging of droplets with a continuous stream in the second dimension CE separation.

Figure 5 shows five sequential CE separations Figure 6 shows (a) a two-dimensional separation of a peptide mixture and (b) a chromatogram from the first dimension LC separation.

Figure 7 shows the UV absorbance electropherogram of a representative droplet from the second dimension separation described in Figure 6(a).

Figure 8 shows the heart cuffing separation of yeast cell proteins in (a) chromatogram from a capillary HPLC separation of yeast cell lysate together with the segmented droplets and (b) CE separation of the selected droplets.

DETAILED DESCRIPTION

In the description below, the separation of a complex protein mixture is used as an example of an analyte solution to describe the invention. However, any mixture of components suitable for separation by any of the methods described above may be used. In the example below, the first separation is achieved by LC, for example HPLC. The second separation is carried out by CE. However, other separation methods and combinations of separation methods may be used.

Although the term "microdroplet" usually refers to a droplet between 1 microlitre and I millilitre in volume, in the description below, the term "microdroplet" or droplet" incorporates droplets of micro-, nano-, pico-, femto-, and attolitre volume. In any given experiment, all of the droplets will be of substantially the same size, e.g. all droplets generated in a particular experiment will be of substantially the same volume.

The two microfluidic devices are designed for analyte compartmentalization and oil depletion (Figure 1(a) and (b)). Both devices are fabricated in a material that has a particular wettability with respect to the continuous phase.

That is, the oil phase should wet the channel surfaces and the aqueous droplets should not wet the channel surfaces. Suitable materials include glass, plastics material, and polymers, including polymeric organosilicon compounds. The devices may, for example, be fabricated from polydimethylsiloxane (PDMS) using known soft lithography methods. The interfacial tension between POMS and water is approximately 40 mN/rn at room temperature. The interfacial tension between PDMS and oil is approximately 1-39 mN/rn at room temperature.

The first device (1), shown in Figure 1(a), includes an oil inlet (2) connected to an oil flow channel (3), a sample channel (4) for eluate from the first separation, and a 1-junction (5) connecting the sample channel with the oil flow channel. Optionally an additive channel (6) is further provided.

The second device (10), shown in Figure 1(b), comprises a k-shaped' channel: A sample delivery channel (11) carrying the oil encapsulated droplets of analyte (12) is connected via a T-junction (13) to an aqueous flow channel carrying aqueous solution (14). Immediately prior to the T-junction, the device comprises one or more pillar elements (15) at the opening (16) to an oil outlet channel (17). The aqueous flow channel is connected at a first end (18) to an aqueous solution reservoir (19) located upstream of the 1-junction, and at a second end (20) to a second analytical separation means (21).

In the first device, the T-junction the angle between the oil flow channel and the sample channel is preferably between 100 and 90°, such as between 30° and 800, for example between 40° and 70°.

In the T-junction in the second device, the aqueous flow channel is preferably at an angle of between 80° and 110° to the sample delivery channel, more preferably at an angle of between 85° and 95°, most preferably at an angle of approximately 90°.

The oil outlet channel is preferably at an angle of between 80° and 110° to the sample delivery channel, more preferably at an angle of between 85° and 95°, most preferably at an angle of approximately 90°.

The first and second device may be connected directly using hydrophobic tubing (25) or by a channel to provide for on-line 2D separations. Alternatively, a longer length of hydrophobic tubing can be used to collect all of the droplets from the first dimension generated using the first device and store them in the tubing in sequence for further analysis. This is shown in Figure 2. In Figure 2, coloured dye was added to the eluate via the additive channel, to increase the colour contrast between the droplets and the oil.

The tubing is preferably made of polytetrafluoroethylene (PTFE) and has an inner diameter of between 1pm to 1000pm, for example between 5Opm to 500pm, for example approximately 300pm. The tubing has a length of between 1cm and 1000cm, for example between 10cm and 100cm. This tubing may then subsequently be connected to the second device for off-line separation. The first and second devices described can thus be provided as a single device or as two separate devices, which may be connected or separate.

In use, the oil flows from the oil inlet (2) along the oil flow channel (3), towards a T-junction (5) with a channel (4) carrying the eluate from the first separation.

The oil inlet is located upstream of the 1-junction with the eluate channel.

The eluate from the first separation is channelled into the oil flow channel, thereby causing droplets of eluate to be encapsulated in the oil. The droplets are then carried along separated by the oil, as the oil flows along the oil flow channel in the direction of the second separation device. The oil optionally also flows through tubing between the first and second devices, if tubing is present.

As shown in Figure 1(b), the microfluidic device incorporates a means to remove the oil (15). Oil removal and merging of droplets into an aqueous phase are crucial processes prior to the second separation apparatus using droplets, since oil contamination alters the surface chemistry of the capillaries and channels and defunctionalises the separation.

A passive pillar array may be used to filter out the oil and induce merging between the droplet and the continuous flow in a channel carrying aqueous solution. An exemplary pillar arrangement is shown in detail in Figure 3. One or more pillar elements (15) produced during the manufacture and made of the same fabrication material as the microfluidic device. This may for example be PDMS, using standard soft lithography. The pillar elements are crucial in evacuating oil and loading analyte droplets into the aqueous flow channel (14).

One or more pillars are located in the opening of the oil outlet channel. A pillar should comprise a flat surface that is substantially in line with the wall of the sample delivery channel (11) adjacent to the opening (16) of the oil outlet channel (17). The rest of the pillar is located inside the oil outlet channel. The device may comprise 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 pillars, for example 1, 2, 3 or 4 pillars.

Two syringe pumps are operated in constant volume mode to drive both the droplets and the continuous oil flow. During operation, one syringe pump allows fine delivery of single droplets downstream to the pillars, while the other is used to aspirate oil from the oil outlet.

The oil removal is achieved by the distance between adjacent pillar elements and the distance between pillar elements and adjacent channel walls being of such a size that that the aqueous droplets are not able to pass through the pillar elements into the oil outlet channel. The total size of the opening (i.e. the distances between the pillar elements and / or between pillar elements is and the channel walls) must be sufficiently large to enable all of the oil to pass into the oil outlet channel.

The combination of oil having a lower surface tension than water, the channel surfaces being lipophilic, the size of the aqueous droplets, and the syringe aspiration in the oil flow outlet, means that the oil is able to pass through in between the pillars into the oil outlet channel, whereas the aqueous droplets are not.

To generate droplets, the oil should have greater wettability to the microfluidic device materials than the solvent or buffer used.

If electrophoresis or any other separation method using an electrical current is used as the first and / or second separation apparatus, the oil should be non-conductive to avoid any potential contribution or interference when conducting 3o online experiments. For use in cellular assays the oil is preferably gas permeable. The oil is preferably transparent to aid detection.

Representative oils useful as a carrier liquid include carbon-based oils, silicone-based oils, and fluorinated oils. Representative examples of oils useful in the invention include embryo-tested mineral oil, light mineral oil, heavy mineral oil, PCR mineral oil, AS4 silicone oil, AS 100 silicone oil, AR2O silicone oil, AR 200 silicone oil, AR 1000 silicone oil, AP 100 silicone oil, AP 1000 silicone oil, AP 150 silicone oil, AP 200 silicone oil, CR 200 Silicone oil, DC 200 silicone oil, DC702 silicone oil, DC 710 silicone oil, octanol, decanol, acetophenone, perfluoro-oils perfluorononane, perfluorodecane, perfluorodimethylcylcohexane, perfluoro-1 -butanesulfonyl fluoride, perfluoro- 1-octanesulfonyl fluoride, perfluoro-1 -octanesulfonyl fluoride, nonafluoro-1 -butanesulfonyl chloride, nonafluoro-tert-butyl alcohol, pertluorodecanol, perfluorohexane, perfluorooctanol, perfluorodecene, perfluorohexene, perfluorooctene, fuel oil, halocarbon oil 28, halocarbon oil 700, hydrocarbon oil, glycerol, 3M Fluoriner.TM. fluids (FC-40, FC-43, FC-70, FC-72, FC-77, FC-84. FC-87, FC-3283), soybean oil, castor oil, coconut oil, cedar oil, clove bud oil, fir oil, linseed oil, safflower oil, sunflower oil, almond seed oil, anise oil, clove oil, cottonseed oil, corn oil, croton oil, olive oil, palm oil, peanut oil, bay oil, borage oil, bergamot oil, cod liver oil, macadamia nut oil, camada oil, chamomile oil, citronella oil, eucalyptus oil, fennel oil, lavender oil, lemon oil, nutmeg oil orange oil, petitgrain oil, rose oil, tarragon oil, tung oil, basil oil, birch oil, black pepper oil, birch tar oil, carrot seed oil, cardamom oil, cassia oil, sage oil, cognac oil, copaiba balsam oil, cypress oil, eucalyptus oil, dillweed oil, grape fruit oil, ginger oil, juniper oil, lavender oil, tovage oil, majoram oil, mandarin oil, myrrh oil, neroli oil, olibanum oil, onion oil, paraffin oil, origanum oil, parsley oil, peppermint oil, pimenta leaf oil, sage oil, rosemary oil, rose oil, sandalwood oil, sassafras oil, spearmint oil, thyme oil, transformer oil, verbena oil, and rapeseed oil.

The oil preferably has a viscosity of between 5 -500 cP at room temperature, for example between 5 -300 cP at room temperature.

While the carrier oil phase is filtered sideways to the oil outlet channel, the aqueous droplet containing analyte sample is driven upwards. Because no surfactant exists in the oil phase, the aqueous droplets merge in a facile and automatic manner with the aqueous solution in the channel carrying aqueous solution. This aqueous solution may, for example, be buffer or electrolyte in a CE channel.

The process of the droplets merging into the aqueous flow is shown in Figure 4. In the channel carrying the oil-encapsulated droplets of aqueous analyte (29), the length of the aqueous droplet or plug' (30) must be greater than the combined length of the opening (32) of the oil outlet channel (33) and the length of the channel wall between the top edge of the entrance to the oil outlet channel and the edge of the entrance to the aqueous channel. The length of the intervening oil plug (31) is not critical to the oil removal.

Each aqueous droplet must fill the entire inner diameter of the channel carrying the oil-encapsulated droplets (29) in order to prevent oil being pushed into the aqueous flow channel. When the droplet begins to merge into the aqueous flow, the droplet blocks the thin channels (34) between the pillars (35) in the opening (32) of the oil outlet channel due to the length of the droplet and due to surface tension.

Further pushing of the oil flow manoeuvres the rest of the aqueous droplet (30) into the aqueous flow channel (36). This pressure is applied by attaching a syringe to an inlet of the channel or tubing between the first and second device, or to the channel carrying the oil encapsulated droplets of analyte.

Significantly, once the droplet has been completely injected into the aqueous flow channel, no oil enters into the aqueous flow channel due to a surface tension barrier. As the subsequent aqueous droplet is pushed up towards the aqueous flow channel, all of the oil that was located between the two droplets is pushed through the pillars into the oil outlet channel.

In summary, due to the size of the aqueous droplets, the small gaps between the pillars, the pillars having a flat surface in one plane with the channel wall adjacent to the opening to the oil outlet channel, and the hydrophobicity of the channel walls and the pillars, the aqueous sample droplets flow past the opening of oil outlet channel. Thereby, the oil flows into the oil outlet channel and the aqueous solution is injected into the aqueous flow channel and carried to a second separation apparatus.

Fluidic channels on the first and second microfluidic device are between 1 pm to 1000pm wide, for example between 5Opm to 500pm wide, for example between 50pm and 250pm wide, for example 120pm wide, but dilate to 300pm where capillary and tubing are inserted. Fluidic channels on the first and second microfluidic device are between 1 pm to 500pm high, for example between 5Opm to 250pm high, for example between 100pm and 200pm high, is for example 150 pm high.

The total length of the opening to the oil outlet channel is between 5Opm and 1000pm, for example between 100pm and 500pm, for example between 150pm and 300pm. The width of the oil outlet channel is between 200pm and 1000pm, for example between 300pm and 600pm, for example between 400pm and 500pm.

The length of the pillar surface that lies in the same plane as the wall of the sample delivery channel is between 10pm and 150pm, for example between 2Opm and 120pm, for example between 5Opm and 100pm, for example between 5Opm and 75pm, for example 6Opm. The length of the gap between adjacent pillars or between a pillar and the wall of the oil outlet channel is between 1pm and 100pm, for example between 25pm and SOpm, for example between 35pm and 45pm, for example 4Opm. The length of the pillars extending into the oil outlet channel is between 100pm and 1000pm, for example between 200pm and 600pm, for example 400pm.

The width of the inlet of the aqueous flow channel is between 4Opm and 150pm, for example between 5Opm and 100pm, for example between 5Opm and 75pm, for example 6Opm. The length of the wall of the sample delivery channel between the end of the opening to the oil outlet channel that is closest to the inlet of the aqueous flow channel and the inlet is between 1pm and 100pm, for example between 10pm and 5Opm, for example between 25pm and 45pm, for example 4Opm. The width of the aqueous flow channel is between 100pm and 500pm, for example between 150pm and 350pm, for example between 200pm and 300pm, for example 250pm.

Droplet volume may range from attolitres to microlitres, preferably droplets are in the range of nanolitre to femtoliter volume. The diameter of a droplet, when it is in its natural spherical state is greater than the width and height of the sample delivery channel. Similarly, the inner diameter of the tubing should be smaller than the diameter of a droplet in its natural spherical state.

The length of one gap between adjacent pillars or between a pillar adjacent to the wall and the wall of the oil outlet channel determines the required droplet size. Accordingly, the length of a droplet in the aqueous delivery channel should be at least 2pm. For example, a droplet of the volume of 52,4 1 Q12 litres may have an diameter of approximately 5Opm and thus a length of approximately 120pm in the sample delivery channel, depending on the width of the sample delivery channel. The droplet will thus block the gaps between pillars or gaps between the pillar and the wall of the outlet channel in the opening to the oil outlet channel until the leading end of the droplet is positioned at the aqueous flow channel inlet.

To minimise dead volumes, surface attachment of the molecules to channel surfaces and droplet polydispersity, all capillaries and tubing were inserted into the channels in a planar fashion. Moreover, no surface treatment of the channel surfaces is required and that the connection between the first and second device is valve-free.

The fractionation resolution can be precisely adjusted by variation of the flow ratio, channel geometry and interfacial tensions. According to Gidding's Criterion for two-dimensional separations, which states that no resolution gained' in the first dimension may be lost in the second dimension, each peak in the first dimension should be sampled at least three times. That is, for each peak observed in the first dimension at least 3 microdroplets should be generated. Close inspection of the LC chromatogram in Figure 6(b) shows that each peak extends over more than 12 seconds. By varying the flow rates io of the oil and the aqueous eluate, channel geometry and interfacial tensions, the size of the droplets, and thus the number of droplets generated per observed peak can be varied and controlled.

In the present invention, the oil flow rate may be between 0.1 and 100 pL/min, for example between 0.1 and 50 pL/min, for example between 0.1 and 10 pL/min. The flow rate of the aqueous eluate may be between 0.1 and 100 pL/min, for example between 0.1 and 50 pL/min, for example between 0.1 and 10 pL/min. The relative flow rates of the oil and of the aqueous analyte eluting from the first separation determines the size of the droplet.

Importantly, once the droplets are formed, encapsulated analytes cannot diffuse beyond the droplet boundary. This is because the oil surrounding the droplet does not allow the analyte to diffuse out of the droplet. Thus "re-mixing" effects, which are problematic in continuous flow based 2D separations, are prevented and the tubing can be directly connected to the second device for separation or be stored for further analysis.

Example

A rapid LC separation of a peptide mixture was carried out on a capillary column packed with poly(styrene-divinylbenzene) (PS-DVB) particles, using acetonitrile/phosphate (AcN/phosphate) (5 mM, pH 3), 30:70 v/v as the elution solvent.

Other eluent solvents may be used. The eluent is selected based on the analyte components to be separated. Suitable solvents include aqueous organic solvent for chromatographic separations, for example, aqueous solutions of acetonitrile, methanol, ethanol, or tetrahydrofuran for reversed phase chromatography. Suitable solvents for ion exchange chromatography and size exclusion chromatography include salt solution, such as, sodium chloride. Suitable solvents for micelle electrokinetic chromatography include buffer with added surfactant, for example, Sodium Dodecyl Sulfate (SDS), or cetrimonium bromide (CTAB). For electrophoretic separations suitable solvents include aqueous buffer solutions, for example, phosphate, or citrate for zone electrophoresis, and for gel electrophoresis linear polymer solutions of polyethylene glycol (PEG), or poly(ethylene oxide) (PEO) are suitable.

Due to the heat-sensitive nature of the PS-DVB polymer particles used for the chromatographic separation, a single particle fritting method was used to fabricate the end frits in a sinter-free way, and a transparent coated fused silica capillary, 100 pm inner diameter, 365 pm outer diameter, was slurry packed with PLRP particles of 10 pm in diameter. The capillary has a packed length of 10cm and a total length of 15cm.

A 0.5 ml plastic syringe was connected to the inlet end of the capillary column via a 1 cm long PTFE sleeve, id 300 pm. During operation, the syringe was used as the mobile phase reservoir. Once a sample was loaded on to the head of the column, the syringe was mounted to a Harvard PhD 2000 syringe pump, which was used as the high pressure supply for capillary LC separation. The outlet end of the capillary column was inserted into the T junction of the first device, as shown in Figure 1(a), for droplet fractionation of the capillary LC eluate. Tubing filled with FC4O oil was connected to the outlet end of the oil flow channel, and a syringe pump was operated in refill mode for aspirating the droplets into the tubing.

The oil flow rate was maintained at 3.5 pL/min, with an aqueous eluate flow rate of 0.2 pL/min and the droplet generation frequency was 0.25 Hz. Within a time frame of 100 seconds, the HPLC eluate was thus converted into 25 droplets with individual volumes of approximately 10 nL.

The separation was complete within 100 seconds. For these separation conditions only two peaks and a low lying hump' were obtained for the five-component mixture, as shown in Figure 6(b).

The rapid elution of the peptides is primarily a result of their positive charge state, which is buffered by the phosphate solvent at pH 3, and decreases their hydrophobic interaction with the stationary phase. The co-elution of species in the two impure peaks is a result of similar hydrophobicity of the components, which limits resolution in this fast LC separation.

Experiments repeated with and without connecting the LC capillary to the first device or removing the collection tubing generated reproducible chromatograms, indicating that droplet generation does not affect the separation in the first dimension.

The second separation was carried out by capillary electrophoresis. An open CE buffer reservoir (5mm in diameter), connected directly with a first end of the CE channel, is punched directly on the microfluidic device, filled with buffer and connected to a high voltage power supply via a platinum (Pt) electrode. The second end of the CE channel is connected directly to a fused silica capillary with an inner diameter of 100 pm. This capillary is immersed in a buffer reservoir connected to ground.

While the second dimension separation is performed, the subsequent droplets remain in a queue within the cartridge tubing. Such queuing does not lead to problems associated with anafyte diffusion since analytes are already "locked" into each droplet vessel. Indeed, the droplet-on-demand operation between the two dimensions detaches the separation stages both temporally and spatially, without sacrificing separation performance. This in turn allows independent optimisation of each separation dimension.

A 20cm capillary was used. Once the sample enters the capillary, a 10 kV high voltage was applied to initiate the electrophoresis, with UV absorbance detection (at 214 nm) being performed downstream of the capillary, a short distance before the second end of the capillary.

The resolution of charged species in CE is related to both the time that the analytes experience the electrophoretic stimulus and the capillary length. A capillary of between 10cm and 100cm, for example 10-50cm, for example 20 cm long is used to perform the CE separation. The voltage may be from 0.1- 50kV, for example 5-30kV, such as 5-1 5kV.

An evaluation of the reproducibility of the second dimension separation (using identical droplets segmented from the original peptide mixture) is shown in Figure 5. The excellent reproducibility of sequential droplet separations enabled the comprehensive analysis of all the fractions obtained from the first HPLC separation.

Figure 6(a) shows a 2D LC-CE pseudo-gel map of the peptide mixture.

Figure 7 shows the electropherogram of a representative droplet of Figure 6(a). Accordingly, all five peptides and possible isomers were resolved and displayed based on the orthogonal selectivity of LC and CE.

The peak capacity achieved in the system of the example was about 2000.

The narrow elution window in the first dimension leads to a peak capacity of only 25, and in the CE stage a reasonably high peak capacity of approximately 75 was obtained.

Accordingly, the present invention enables analyte droplets to be transferred across the interface between a first and a second separation device, without the analyte component separation that has been achieved in the first separation being lost during the transfer across the interface. This is because the analyte components are retained within nanolitre droplets during the transfer to the next separation device. The oil separating the droplets from each other is removed directly prior to the droplets entering the second separation device. Because of the extremely short time between the analyte droplets merging into an aqueous flow and the second separation being carried out, the components are not able to remix between former droplets, and molecular diffusion is minimal. Accordingly, the resolution achieved by the first separation is maintained as the initial resolution position upon which the second separation method can build.

is For more complex mixtures, such as proteomic samples, a detailed analysis of certain sub-groups (e.g. proteins that are slightly different due to post translational modifications) is often more important than achieving a comprehensive 2D map.

In this regard, a "heart-cutting" 2D separation has also been pursued with the droplet platform described herein. In this method only certain fractions of interest are submitted to the second dimension for further analysis. These experiments are illustrated in Figure 8, which shows a heart-cutting 2D separation of crude lysate of yeast cells. A stepwise gradient separation was conducted in the first dimension LC separation. Four plugs of mobile phase with increasing AcN concentrations (20, 40, 60 and 80%) were utilised sequentially. The eluate was segmented within 40 minutes into 260 droplets containing proteins that ranged from highly hydrophilic to highly hydrophobic.

Droplets in elution plugs of 20% and 40% AcN, containing relatively hydrophilic proteins, were selectively chosen to perform the second dimension CE separation.

The present invention describes experiments on droplet-mediated two-dimensional separations. Further optimisation of each dimension can be performed in isolation to increase component resolution and decrease separation time.

The unique properties of droplet-based microfluidics provide significant advantages when interfacing orthogonal separation techniques. First is the ability to segment, encapsulate and manipulate the separated samples within nL-sized droplets. The generation and merging of droplets occurs in a passive io manner, without the need for valves or electrokinetic driven flows. The benefits of adjustable droplet volumes ranging from nanolitres to femtolitres are also significant. The ability to finely slice' peaks in the first dimension ensures that no chemical or biological information will be lost. Furthermore, the ultra small volume of samples that can be handled is especially important in many modern day applications such as single cell proteomics.

After the second separation has been carried out, the components of the analyte can be made visible by various methods. The analyte molecules may fluoresce under ultraviolet light, for example due to fluorescence or phosphorescence, optionally caused by fluorescent or phosphorescent labels, or the analyte molecules contain radiolabels. In these cases a photograph under UV light can be taken, or an autoradiogram can be recorded. Detection may also be carried out by refractive index. The molecules may also be stained, for example with ethidium bromide, silver, or coomassie blue dye, to make them visible. Other suitable detection methods may also be used.

The method of this invention may further comprise the use of molecular weight size markers, for example where the second separation is carried out by gel electrophoresis. Suitable markers include prestained markers, such as ColorBurstTM markers, Color markers, Prestained Blue marker; unstained markers, such as SigmaMarkersTM, Ultra-Low Range marker, Peptide molecular weight marker; and specialty markers, such as ChemichromeTM Western Control, ChemichromelM Ultimate, ProteoProfileTM PTM Marker, biotinylated markers, 2D electrophoresis marker, fluorescent markers, recombinant markers, silver stain markers, SYPRO® marker, Non-Denaturing molecular weight marker, and IEF marker (3.6-9.3) (all available from Sigma Aldrich).

In a resulting 2D gel, bands or spots in different lanes that are located at the same distance from the top of the gel, contain molecules that passed through the gel with the same speed, usually meaning that they are approximately the same size. If a marker is run on one lane in the gel parallel to a sample containing undetermined components, the observed bands of the known marker can be compared to the bands of the undetermined components in order to determine their size.

Optionally, reagents can be added via an additive channel and mixed into the droplets with ease. These additives are provided in aqueous solution and are therefore able to easily combine with the aqueous droplets. Hereby many sample preparation steps can be integrated into the droplet connector.

Examples of additives include analyte solution, one or more buffer additives needed for the second dimension, radioactive labels, fluorescent labels, affinity ligands, or dye to stain the protein in a second dimension gel electrophoresis separation.

Radioactive labels and fluorescent labels enable radioactive or fluorescent detection downstream. If both types of label are added, it is possible to carry out, for example, UV detection after the first separation (in the first dimension) and laser-induced fluorescence (LIF) after the second separation (in the second dimension). This is important for detecting low abundance proteins, which have limited copy numbers.

The additive channel can also be used to introduce detergent to denature proteins in an online style. This can enable gel electrophoresis in the second dimension. Suitable denaturing reagents include SDS Towbin Transfer Buffer (SDS TTB). Such a separation can also provide molecular weight information of proteins which is useful for identification of proteins Other samples besides protein or peptide containing samples may be analysed using the method of this invention. These include protein fragment, nucleic acid, DNA, RNA, amino acid, organic molecules, inorganic molecules and cells.

Although the present invention is described in detail above in relation to a two-dimensional capillary HPLC-CE 2D system, this approach can be widely adapted to couple a range of other separation techniques. Consequently, droplet-based interfaces could become key components in 2D or multi dimensional separations. Thus, the first and second separation techniques can be independently chosen from any of the following separation techniques including: high-performance liquid chromatography (HPLC), ion exchange chromatography, size exclusion chromatography, bioaffinity chromatography, reverse phase chromatography, liquid chromatography, low-pressure liquid chromatography, open tubular liquid chromatography, magnetic chromatography, capillary electrophoresis, capillary zone electrophoresis (CZE), isotachophoresis (lIP), capillary isoelectric focusing (ClEF), capillary gel electrophoresis (CGE), micellar electrokinetic chromatography (MEKC), electrokinetic chromatography (EKC), micro emulsion electrokinetic chromatography (MEEKC), non aqueous capillary electrophoresis (NACE), and capillary electrochromatography. Preferred separation techniques include High Pressure Liquid Chromatography (HPLC), Liquid Chromatography, Isoelectric Focusing, Capillary Electrophoresis, and Capillary Gel Electrophoresis.

Examples of column material useful in, for example, HPLC, include any separatory material capable of separating a chemical mixture, including stationary phases modified with C18, C8, phenyl, silica, pentafluorophenyl, cyano, and amino groups. HPLC columns may also include embedded polar groups, ion-exchange phases, and reverse-phase amides.

A significant benefit of the current approach relates to the fact that droplets are separated using an electrically non-conductive oil. Accordingly, any electrical field is effectively insulated between the dimensions, providing additional flexibility in connecting two-dimensional CE separations.

Claims (29)

1. CLAIMS: 1. An oil removal device for removing oil from a flow of oil-separated aqueous sample droplets, the oil removal device comprising: (I) a sample delivery channel for receiving the flow of the oil-separated sample droplets; (ii) an oil outlet channel for carrying away oil removed from the flow; (iii) at least one restricted opening between the sample delivery channel and the oil outlet channel; (iv) an aqueous flow channel for conducting a flow of aqueous liquid; and (v) an aqueous flow channel inlet between the sample delivery channel and the aqueous flow channel, which restricted opening is located downstream of the oil channel opening, for introducing the aqueous sample droplets into the aqueous flow channel; wherein the dimensions of the restricted opening are such that the oil flows into the oil outlet channel, but the aqueous sample droplets do not.
2. 2. A device according to claim 1, wherein the distance between (i) the aqueous flow channel inlet, and (ii) the end of the restricted opening that is furthest away from the inlet, is smaller than the length of an aqueous sample droplet.
3. 3. A device according to claim I or 2, wherein the oil removal device comprises one or more pillar elements at the opening of the oil outlet channel.
4. 4. A device according to claim 3, wherein in the opening of the oil outlet channel, the one or more pillar elements creates a flat surface substantially in line with the wall of the sample delivery channel.
5. 5. A device according to any one of claims 1-4, wherein the device is made from a material which is hydrophobic.
6. 6. A device according to any one of claims 1-5, wherein the oil outlet channel is located in a side wall of the sample delivery channel.
7. 7. A device according to any one of claims 1-6, wherein the aqueous flow channel inlet is located at the end of the sample delivery channel.
8. 8. A system for transferring a stream of aqueous material separated in a first separation apparatus to a second separation apparatus, the system comprising: (i) an encapsulation device for encapsulating successive portions of the stream received from the first separation apparatus as droplets between plugs of oil to provide a flow of oil-separated sample droplets; (ii) a conduit for transporting the flow of oil-separated sample droplets; and (iii) an oil removal device for receiving the flow of oil-separated sample droplets, the oil removal device comprising: (a) a sample delivery channel for receiving the oil-separated sample droplets from the conduit and delivering the sample droplets into an aqueous liquid stream flowing in a channel connectable to the second separation apparatus; (b) an oil outlet channel for carrying away removed oil; (c) at least one restricted opening between the sample delivery channel carrying the oil-separated sample droplets received from the conduit and the oil outlet channel; and (d) an aqueous flow channel.
9. 9. A system according to claim 8, wherein the conduit is of a sufficient length to retain or store all of the oil-separated sample droplets.
10. 10. A system according to claim 8, wherein the oil removal device is as claimed in any of claims 1-7.
11. 11. A system according to claim 8-10, wherein the device is provided in two separate parts connectable by the conduit.
12. 12. A system according to claim 11, wherein the first and second separation apparatus are each an apparatus for resolving a plurality of components from a mixture.
13. 13. A system according to claim 11 or claim 12, wherein the first and second separation apparatus each carry out a different separation technique.
14. 14. A method for transferring an aqueous analyte solution between a first separation apparatus and a second separation apparatus, wherein: (i) encapsulating the analyte solution as droplets by an oil; (ii) conducting said oil and droplets of analyte solution from the first separation apparatus flow along a conduit; wherein each of said droplets is of such dimensions that it fills the inner volume of said conduit; and (iii) removing the oil from the analyte before the analyte reaches the second separation apparatus; wherein the oil removal step comprises: (a) pushing the oil-encapsulated droplets is pushed along said conduit, passing an oil outlet channel opening, towards an inlet of an aqueous flow channel, (b) said droplet being of such a length that when the leading end of the droplet is positioned at the aqueous flow channel inlet, it completely blocks the oil outlet channel opening, and (c) forcing the oil that separates two successive droplets through the oil outlet channel opening into the oil outlet channel, after the preceding droplet has started to enter into the aqueous flow channel.
15. 15. A method according to claim 14, wherein the oil in the conduit does not io enter the aqueous flow channel, due to surface tension effects at the aqueous flow channel inlet.
16. 16. A method according to claim 14 or 15, wherein the conduit is a channel or tubing.
17. 17. A method according to claim 14 -16, wherein the conduit is a length of tubing able to receive all of the oil encapsulated analyte droplets.
18. 18. A method according to claim 17, wherein the oil encapsulated analyte droplets may be stored in the tubing.
19. 19. A method according to any one of claims 14-18, wherein the oil encapsulated analyte droplets are transferred through the conduit to the second separation apparatus.
20. 20. A method according to any one of claims 14-19, wherein the oil has a surface tension to the device material that is less than the surface tension between water and the device material.
21. 21. A method according to any one of claims 14-20, wherein the oil is electrically non-conductive.
22. 22. A method according to any one of claims 14-21, wherein pressure is applied to drive the droplets and the oil flow downstream to the aqueous flow channel inlet, and suction means are applied to the oil outlet channel to aspirate oil from the oil outlet.
23. 23. A method according to any one of claims 14-22, wherein the first and second separating apparatus are different separation techniques.
24. 24. A method according to any one of claims 14-23, wherein the analyte io solution comprises a plurality of molecules or macromolecules.
25. 25. A method according to any one of claims 14-24, wherein the analyte solution comprises a plurality of molecules or macromolecules, selected from any of the following: protein, protein fragment, peptide, nucleic acid, DNA, RNA, amino acid, organic molecules, inorganic molecules and cells.
26. 26. A method according to any one of claims 14-25, wherein each separation apparatus is independently selected from: High Pressure Liquid Chromatography, Liquid Chromatography, I soelectric Focusing, Capillary Electrophoresis, Capillary Gel Electrophoresis, Isotachophoresis, and micellar etectrokinetic chromatography.
27. 27. A method according to any one of claims 14-26, further comprising the use of detection means for either one or both of the first and second separation means.
28. 28. A method according to claim 27, wherein the detection means are independently selected from: UV detection, fluorescence, phosphorescence, staining, radioactivity, refractive index detection, and vibrational spectroscopy.
29. 29. A method according to any one of claims 14-28, wherein any one or more of the following agents is added in an aqueous solution via an additive channel connected to the conduit: a fluorescent agent or label, a radioactive label, a dye, a buffer additive and an affinity ligand.