It is a frequent requirement to determine both the quantity and type of biotic or abiotic particles such as microorganisms present in a sample of liquid or gas. It is known to use the phenomenon of dielectrophoresis ('Dielectrophoresis', H A Pohl, Cambridge University Press, 1978) to separate out and to collect particular types of organisms from a sample suspension.

Dielectrophoresis has been defined as the motion of a polarised but uncharged particle situated within a region of non-uniform electric field. When a particle is placed within a region of electric field, a dipole may be induced within that particle in addition to any permanent dipole that may be present within that particle. If the particle is within a region of non-uniform electric field there may be an imbalance in the forces exerted upon either end of the net dipole within the particle owing to one end of the dipole being situated in a region of higher electric field than the other end, even though the particle remains electrically neutral and has no excess charge. Depending upon the directions of the dipole moment and the electric field vector, the imbalance of forces exerted upon the particle may lead to a net attractive force being exerted upon the particle inducing a translational motion towards areas of highest electric field non- uniformity (referred to as positive dielectrophoresis), or may lead to a net repulsive force being exerted upon the particle inducing a translational motion towards areas of lowest electric field non-uniformity (referred to as negative dielectrophoresis). If the net force exerted upon the particle is zero, then no induced motion will occur. In addition, rotational motions may be induced upon the particle when the electric field acts to align the dipole moment. This phenomenon is referred to as electrorotation (W M Arnold et al; J Electrostat; 21,2-3,151- 191,1988).

It is known to use positive dielectrophoresis to aggregate particles from a suspension in areas of high electric field non-uniformity. These areas of high electric field non-uniformity are typically generated by metal electrodes constructed as a pair, number of pairs, or in another

design. The collected particles can then be released and counted as they pass under a microscope (A P Brown et al; Microbios; 91,51-65,1997). This technique involves placing the sample in a reservoir and pumping the liquid from the reservoir across a pair of spaced electrodes providing a non-uniform electric field gradient. Because different particles (e. g. microorganisms) respond to different frequencies in a non-uniform electric field, the electrodes are supplied with a voltage having a frequency corresponding to a selected type and these particles are then attracted to one or other of the electrodes under positive dielectrophoresis (X B Wang et al; J Phys D; 26,8,1278-1285,1993). The liquid is cycled for a period sufficiently long for substantially all the selected organisms to be collected at the electrodes.

Discontinuing the voltage releases the particles back into the circulating liquid. A microscope over the electrodes or downstream of the electrodes can be used to count the number of passing particles in a representative cross-section of the flow and from this the total number of particles can be determined.

While the above technique is adequate to determine the concentrations of relatively large particles or cells, there is a problem with submicron sized particles and chemicals since the resolution of a microscope is not adequate to enable a satisfactory quantitative count of these to be made. This is also true of most measurement devices and systems used for detecting or quantifying dielectrophoresis, such as the evanescent imaging system described by Hughes (M P Hughes et al; Meas Sci Tech; 10,8,759-762,1999), the laser based system described by Pohl (US Patent 4,326,934; 1982), the spectrophotometric method described by Betts (UK Patent GB 2266153B; 1993) and the general method described by Betts (UK Patent WO 91/08284; 1991).

The problem is solved by determining the concentration of particles by measuring the effect that they have on the electrical characteristics of the electrodes, namely the amount by which they change the impedance of the electrodes. A comparison can be effected between the change in voltage across a pair of electrodes supplied with the sample liquid and a pair of reference electrodes supplied with a reference liquid, to determine the net change in impedance caused by the aggregation of particles at the pair of electrodes supplied with the sample liquid.

Alternatively, the change in voltage across a pair of electrodes supplied with the sample liquid can be monitored over time to determine the change in impedance caused by the aggregation of particles at the electrodes over time relative to impedance of the electrodes prior to the aggregation of particles at the electrodes.

Alternatively, the electrodes in either case may be configured as a number of pairs or in another fashion designed to manipulate the particles in the suspension in to a region where impedance measurements may be most effectively taken or particle aggregation is most efficient or designed for an other purpose.

It is an object of the invention to provide an improved method and apparatus for sampling microorganisms or detecting bodies. The term"body", as used hereafter, is defined as non-exclusively including abiotic or biotic particles, biological cells, viruses, viroids, prions, subcellular organelles, chemical and biological molecules.

According to the present invention there is provided a method of determining the quantity of bodies present in a fluid sample, the method comprising the steps of circulating the fluid sample through a region of non-uniform electric field density produced by an electrode structure, circulating a reference fluid having similar electrical characteristics to the carrier fluid of the sample through a region of non-uniform electric field density produced by a second electrode structure having a similar configuration to said electrode structure, energising both electrode structures with a first voltage having a predetermined frequency and for a sufficient period selected to attract a predetermined variety of bodies in the sample to the first electrode structure, superimposing on both electrode structures a second voltage having a second frequency selected so as to impose substantially no translational force on said predetermined bodies in said sample, measuring the voltages at said second frequency across said two electrode structures, processing said measurements to determine the differences in capacitance and conductance between said two electrode structures directly related to the quantity of bodies built up on said electrodes.

According to the present invention there is further provided an apparatus for determining the quantity of bodies present in a sample, the apparatus comprising a support defining first and second similar fluid flow channels through similar regions of non-uniform electric field density respectively produced by first and second electrode structures of similar configuration, first circulation means for circulating a sample containing said bodies through said first channel and second circulation means for circulating a fluid of similar electrical characteristics to the fluid of said sample but without said bodies through the second channel, circuit means connecting said first electrode structure in series with a first capacitor, said second electrode structure in series with a second capacitor and both said circuit series in parallel with each other, a first AC source for supplying a first voltage at a first frequency across the parallel circuit, said frequency being selected to cause a predetermined variety of bodies to be attracted to said first electrode structure, a second AC source for supplying a second voltage at a second frequency across said parallel circuit, said second frequency being selected so that it imposes substantially no force on said predetermined bodies in said sample, means for measuring said second voltage, a third voltage at said second frequency appearing across said first electrode structure and a fourth voltage at said second frequency appearing across said second electrode structure and the relative phase between the third and fourth voltages and the second voltage, and means for processing said measurements to determine the difference in the capacitive and conductive elements of the impedances of the two electrode structures.

According to the present invention there is still further provided a second method of determining the quantity of bodies present in a fluid sample, the method comprising the steps of circulating the fluid sample through a region of non-uniform electric field density produced by an electrode structure, energising said electrode structure with a first voltage having a predetermined frequency selected so as to impose substantially no translational force on said predetermined bodies in said sample, measuring the voltages at said frequency across said electrode structure, then superimposing on the electrode structure a second voltage having a second predetermined frequency for a sufficient period selected to attract a predetermined variety of bodies in the sample to the electrode structure, measuring the voltages at said first

frequency across said two electrode structures, processing said measurements to determine the variation with time of capacitance and conductance of said electrode structure, directly related to the quantity of bodies that builds up on said electrodes.

According to the present invention there is also provided A second set of apparatus for determining the quantity of bodies present in a sample, the apparatus comprising a support defining a fluid flow channel through a region of non-uniform electric field density respectively produced by an electrode structure, circulation means for circulating a sample containing said bodies through said channel, circuit means connecting said electrode structure in series with a capacitor, a first AC source for supplying a first voltage at a first frequency across the circuit, said frequency being selected so that it imposes substantially no force on said predetermined bodies in said sample, a second AC source for supplying a second voltage at a second frequency across said circuit, said second frequency being selected to cause a predetermined variety of bodies to be attracted to said electrode structure, means for measuring said first voltage and a third voltage at said first frequency appearing across said electrode structure and the relative phase between the third voltage and the first voltage, performing said measurements both with the second voltage disabled and at predetermined time intervals with said second voltage enabled, and means for processing said measurements to determine the difference in the capacitive and conductive elements of the impedance of the electrode structure.

Methods and apparatus for sampling microorganisms, whereby a comparison is effected between a pair of electrodes supplied with a sample liquid and a pair of reference electrodes supplied with a reference liquid, will now be described by way of example, with reference to the accompanying diagrammatic drawings, in which: Figure 1 is an electrical and fluid circuit of one embodiment of the apparatus; Figure 2 is a perspective fragmentary section taken through the collection block of the apparatus of Figure 1;

Figure 3 is a graph showing variation in capacitance change recorded with drive voltage source frequency for a variety of sample liquids consisting of differing concentrations of 2 um diameter abiotic latex beads suspended in deionised water; Figure 4 is a graph showing variation in capacitance change recorded with concentration of 2 pm diameter abiotic latex bead sample liquids for a variety of drive voltage source frequencies; Figure 5 is a graph showing variation in capacitance change recorded with drive voltage source frequency for a variety of sample liquids containing differing concentrations of 100 nm diameter abiotic latex beads suspended in deionised water; Figure 6 is a graph variation in capacitance change recorded with concentration of 100 nm diameter abiotic latex bead sample liquids for a variety of drive voltage source frequencies.

The apparatus shown in Figure 1 comprises a collection block 2 in which micron or submicron sized organisms can be collected and measurements taken. The collection block 2 contains a pair of spaced electrodes 4 and 6 lying in common plane and a fluid flow channel 8 positioned to cause liquid to flow across the upper faces of the two electrodes.

The structure can be more clearly seen in Figure 2. As shown, the structure includes an electrically insulating substrate 10 on which the two elongate electrodes 4 and 6 have been deposited in parallel but spaced relationship with each other. Also, deposited on the substrate 10 and extending at right angles to and overlapping the two electrodes 4 and 6 where they traverse them, are a pair of spaced electrically insulating strips 14 and 12. The resulting slot 16 between the two strips 12 and 14 defines the flow channel 8 across the two electrodes 4 and 6.

A further electrically insulating layer (not shown) extends over the strips 12 and 14 and the slot 16 to form the roof of the channel 8. The exposed face of each electrode may be covered with a thin electrically insulating layer, or non-stick or other coating as required.

A reservoir 20, for containing a sample of liquid to be analysed, is connected by a duct 22 to the upstream end of the channel 8. A duct 24 connected to the downstream end of the channel 8 feeds liquid from the channel 8 through a pump 26 back to the reservoir 20.

The pump 26 is advantageously a peristaltic pump to prevent any damage, injury or contamination to the sample liquid. The liquid in the reservoir 20 may be agitated by bubbling air or other gas therethrough or by using a magnetic stirrer or other device to keep the microorganisms in suspension.

This whole arrangement is replicated by a similar reference arrangement in which similar parts are similarly referenced but provided with the suffix A. The reservoir 20A contains a reference liquid selected as will be described in more detail hereinafter and the channel 8A and the electrodes 4A and 6A share the same substrate 10.

The capacitor formed by the electrodes 4 and 6 is connected, in series, with a capacitor 30 to form a potential divider. The output voltage VA is obtained from a centre tap terminal 32 between the two capacitors.

Similarly, the capacitor formed by the electrodes 4A and 6A is connected in series with a capacitor 34 to form another potential divider. An output voltage VB is obtained from a centre tap terminal 36.

The two potential dividers are connected in parallel with each other and an impedance matching resistor 38 across a pair of input terminals 40 and 42. A terminal 41 provides a voltage VIN appearing across the terminals 40 and 42.

An AC probe voltage source 44, supplies a voltage to the primary winding of a transformer 46. An AC drive voltage source 48 is connected in series with the secondary winding of the transformer 46 across the terminals 40 and 42. A first impedance matching resistor 50 is connected in parallel with the source 48 and a second impedance matching resistor 51 is connected in parallel with the secondary winding of the transformer 46.

The drive voltage source 48 is a variable frequency source having an output voltage selected to cause microorganisms to migrate to one or other of the two electrodes 4 and 6.

Typically, the voltage may be of the order of 12 volts and the frequency varied in the range of from 103 to 10 Hz.

The probe voltage source 44 is a fixed frequency source having a voltage sufficiently low as to have little or no effect in attracting microorganisms to one or other of the two electrodes 4 and 6. Typically, the voltage may be of the order of 0.5 volts and the frequency around 810 Hz.

A lock-in amplifier (not shown) is used to monitor the relative changes in magnitude and relative phase of the voltages VA, VB and VIN- It is apparent that with some samples, when microorganisms are being collected on the electrodes 4 and 6, that not only does the permittivity between the electrodes change but also the conductivity, due to liquid being carried between the electrodes being displaced by the microorganisms, which themselves have a different conductivity to the carrier liquid.

Accordingly, the liquid contained in the reservoir 20A is selected to have similar electrical characteristics to the carrier liquid for the microorganisms.

In operation, both pumps 26 and 26A are operated to cycle respective liquids. Then, with the drive source 48 de-energised and probe source 44 energised, the magnitudes and relative phases of the voltages ViN, VA and VB are determined.

The drive source 48 is then energised at a frequency selected to attract a specific organism and, after a period of sufficient length to allow all those organisms to be attracted to one of the electrodes 4 and 6, the magnitude and relative phases of the voltages VIN, VA and VB, are again determined.

The measurements of VIN, VA and VB can be represented as complex numbers VIN*, VA* and VB*.

Assuming that the impedance of the capacitor 30 has a value of ZM and that the impedance of the electrodes 4 and 6 can be represented by ZE, then the following relationship exists: <BR> <BR> VA ZM<BR> . _.<BR> <P>VIN-VA ZE ZE can also be expressed in terms of admittance YE where and admittance is related to the electrode conductance G of the electrodes 4 and 6 where YE = G +jcoC where C = electrode capacitance and co (= 2sf) is the angular frequency of the probe source voltage.

Using the formula ZM = (jcoC)'whereCM is the capacitance of the capacitor 30, it can be seen that VA C G CL CM The same mathematical process can be used to determine the values CA and GA of the reference electrodes 4A and 6A. 4

This then allows the change in capacitance change AC and conductance change AG in electrodes 4 and 6 to be determined. These values are then adjusted for any changes due to the energisation of the electrodes 4 and 6 by the drive source 48 and the resultant AC and AG were found to be directly related to the quantity of organisms built up on the electrodes. The correlation of AC and AG with the quantity of organisms is a significantly more accurate correlation than that found using the change of capacitance AC with the quantity of organisms with conductivity changes not accounted for (as presented in K R Milner et al; Electron Lett; 34,1,66-68,1998).

In another embodiment of the invention, all of the components shown in Figure 1 are used except those provided with the suffix A, that is the reservoir 20A, the pump 6A, the duct 24A, the channel 8A, the electrodes 4A and 6A, the capacitor 34 and the centre tap terminal 36.

In this fashion the circuitry is identical to the previous method, with the exception of having no reference electrodes or associated components.

As described previously, an AC probe voltage source 44, supplies a voltage to the primary winding of a transformer 46. An AC drive voltage source 48 is connected in series with the secondary winding of the transformer 46 across the terminals 40 and 42. A first impedance matching resistor 50 is connected in parallel with the source 48 and a second impedance matching resistor 51 is connected in parallel with the secondary winding of the transformer 46.

The drive voltage source 48 is a variable frequency source having an output voltage selected to cause microorganisms to migrate to one or other of the two electrodes 4 and 6.

Typically, the voltage may be of the order of 12 volts and the frequency varied in the range of from 103 to 107 Hz.

The probe voltage source 44 is a fixed frequency source having a voltage sufficiently low as to have little or no effect in attracting microorganisms to one or other of the two

electrodes 4 and 6. Typically, the voltage may be of the order of 0.5 volts and the frequency around 7 kHz.

A lock-in amplifier (not shown) is used to monitor the relative changes in magnitude and relative phase of the voltages VA and VIN.

In operation, pump 26 is operated to cycle the sample liquid. Then, with the drive source 48 de-energised and probe source 44 energised, the magnitudes and relative phases of the voltages Vi and VA are determined.

The drive source 48 is then energised at a frequency selected to attract a specific organism and the magnitude and relative phases of the voltages V, N and VA (or of VA alone) are repetitively determined at predetermined time intervals, until a period of sufficient length has elapsed to allow all those organisms to be attracted to one of the electrodes 4 and 6.

The mathematical process described earlier can be used to determine the values C and G of the electrodes prior to particle attraction and at predetermined time intervals while particle attraction is occurring, whereby VA -r'p A -'"VA TIME=T VA IN *ATIME=T TIME=T This then allows the net capacitance change AC and conductance change AG in electrodes 4 and 6 after a predetermined time to be assessed. The resultant AC and AG were found to be directly related to the quantity of organisms built up on the electrodes. It is also possible to differentiate the capacitance change and conductance change data with respect to time to provide values for DC/8t and agnat respectively, which were found to be directly

related to the rate at which organisms built up on the electrodes. Both of these methods may also be used to infer the concentration of particles in the liquid sample.

In yet another embodiment of the invention, the impedance spectrum of particles held at a pair of electrodes supplied with a sample liquid is monitored. This method may be used with either of the apparatus layouts described earlier, that is with a single collection block supplied with a sample liquid, in which micron or submicron sized organisms can be collected and measurements taken, along with associated pump and duct configurations and appropriate electrical power supplied and components; alternatively with two collection blocks, one supplied with a sample liquid, in which micron or submicron sized organisms can be collected and measurements taken and the other supplied with a reference liquid selected to have similar electrical characteristics to the carrier liquid for the microorganisms, along with associated pump and duct configurations and appropriate electrical power supplied and components.

In operation, pumps are operated to cycle the liquids. The drive voltage source is then energised at a frequency preselected to attract a specific organism to the electrodes. With the organisms held at the electrodes, the capacitive and resistive components of the electrodes across a frequency range are then determined using the mathematical process described earlier, by measuring the magnitudes and phases of voltages Vs and VA (and VB if a pair of electrodes supplied with a reference liquid is used) for a variety of probe voltage source frequencies. In this manner, dielectric spectroscopy (H P Schwan in'Electrical Properties of Tissue and Cell Suspensions, Vol 5, Advances in Medical and Biological Physics', S H Lawrence and C A Tobias, Academic Press, New York, 1957) is performed on the organisms held in place at the electrodes. If a reference liquid is also used, a comparison may be effected between the dielectric spectrum of the pair of electrodes supplied with a sample liquid and the pair of reference electrodes supplied with a reference liquid, The resultant AC and AG spectra were found to be both characteristic of the organisms

in the liquid sample and also directly related to the quantity of organisms built up on the electrodes.

It will be appreciated that while these methods and apparatus for the quantitative determination are described in connection with organisms which are of submicron size, they can equally be used for larger microorganisms which can be viewed and counted using a microscope.

The results are not only more accurate than in the situations where a microscope count is effected since the measurement is made on the basis of the total number of organisms, not just a representative sample (e. g. 1 in 40 as in A P Brown et al ; Biosens Bioelectron; 14,3,341- 351,1999) but also much speedier since no time is spent in the laborious counting of the microorganisms after release. Furthermore, the whole apparatus is much simplified since no microscope is involved merely electronic circuitry for determining the value of AC and AG.

EXAMPLE 1 A number of cell characterisation spectra were obtained using the above experimental technique whereby a comparison was effected between a pair of electrodes supplied with a sample liquid and a pair of reference electrodes supplied with a reference liquid. A variety of sample liquids consisting of differing concentrations of optically resolvable 2 zm diameter abiotic latex beads suspended in deionised water were used in conjunction with a similar number of reference liquids comprising deionised water alone. Each characterisation spectrum was obtained by varying the frequency of the drive voltage source and recording the resulting variations in voltages developed across the measurement capacitors at the frequency of the probe voltage source with a lock-in amplifier. Figure 3 illustrates the resulting capacitance change spectra. The quantitative nature of the technique is demonstrated by plotting the capacitance change values for each sample liquid recorded at a discrete frequency of the drive voltage source in order to assess the relationship between sample liquid concentration (and hence, indirectly, number of particles collected at the pair of electrodes supplied with a sample liquid) and induced capacitance change. Figure 4 illustrates the resulting trends for drive

voltage source frequencies of 4 kHz and 40 kHz. It will be apparent that the concentration of a sample liquid containing an unknown number of 2 um latex beads could be determined by measuring the capacitance change value recorded at one of these, or any other, drive voltage source frequency and referring to Figure 4 or another calibration curve determined for the drive voltage source frequency in use. It will also be apparent that the concentration of a sample liquid containing an unknown number of another particle could be determined by measuring the capacitance change value recorded at a discrete drive voltage source frequency and referring a similar calibration curve developed for the particle in question.

EXAMPLE 2 A number of cell characterisation spectra were obtained using the above experimental technique whereby a comparison was effected between a pair of electrodes supplied with a sample liquid and a pair of reference electrodes supplied with a reference liquid. A variety of sample liquids consisting of differing concentrations of 100 nm diameter abiotic latex beads, which cannot be optically resolved, suspended in deionised water were used in conjunction with a similar number of reference liquids comprising deionised water alone. Each characterisation spectrum was obtained in similar fashion as in Example 1. Figure 5 illustrates the resulting capacitance change spectra. The quantitative nature of the technique is demonstrated in similar fashion as in Example 1. Figure 6 illustrates the variation in measured capacitance change with sample liquid concentration for drive voltage source frequencies of 1.6 kHz and 6.3 kHz. It will be apparent that the concentration of a sample liquid containing an unknown number of any particle type, either optically resolvable or not, could be determined by measuring the capacitance change value recorded at a discrete drive voltage source frequency and referring a similar calibration curve developed for the particle type in question.

It will also be apparent to those skilled in the art that many modifications and changes could be made to this specific apparatus without departing from the scope of the invention. For example, other electrode configurations could be used to produce the non-uniform electric field, other fluid flow channels could be constructed over said electrodes, or other frequencies or magnitudes of the probe and drive voltage sources could be employed.