GB2189605A - Pollutant detector - Google Patents

Abstract

A method of detecting pollution in an aqueous liquid flow comprises adding a compatible electron transfer mediator to a portion of the liquid. The mixture is passed into a sensor cell in which is contained a living bacteria means, an activity of which is stimulated at least periodically. The level of said activity is measure at an electrode in the cell by means of electron transfer thereto by said mediator. Measurement is preferably carried out in a flow cell. The mediator, which is preferably one or more of potassium ferricyanide and and dimethyl or para benzoquinone is mixed with the aqueous liquid and passed through inlet 1 to carbon or platinum electrode 4 held at 400-550 mV relative to silver/silver chloride reference electrode 5. Electrode 4 is surrounded by bacteria retained by a fine nylon mesh on a translucent alumina filter and the bacteria activity is stimulated by LED light source 6, or, for some bacteria, by glucose and/or fructose. Various suitable bacteria are specified and the detector may be used to determine water pollution. <IMAGE>

Description

SPECIFICATION Pollutant detector The present invention relates to a pollutant detector and a method of detecting pollution. More particularly, but not exclusively, the invention relates to a method and apparatus for detecting pollution in aqueous liquids, for example water supply, by use of a bacterial sensing means.

Thirty percent of water supplies in the United Kingdom are drawn directly from rivers, which may on occasion be subject to pollution. Such pollution may cause shut down of the water purification works if detected in time. If the pollution is not detected in time, the water supply may be contaminated. Many of the pollutants which may enter rivers are toxic chemicals, e.g.

herbicides, which may not degrade during the normal water purification process.

Typical pollutants which may be found in water supplies and which may be detected by the present invention include: Diesel oil Oil Petroleum Aniline Metal Salts Agricultural chemicals e.g. of Cd, Cu, Cr and particularly such commerical herbicidal formulations as: DIOUAT PARAQUAT MORPHAMQUAT BROMOXYNIL DICHLOBENIL DIPHENATRILE IOXYNIL AMETRYNE ATRATONE ATRAZINE AZIPOTRYN BLADEX CHLORAZINE CYPRAZINE DESMETRYNE IPAZINE METHOPROTYN PROMETRONE PROMETRYNE PROPAZINE SIMAZINE SIMETONE SIMETRYNE TERBUTRYN TRIETAZINE BENZOMARC BENZTHIAZURON BUTURON CHLOROBROMURON CHLOROXURON CHLORTOLURON CYCLURON DCU DIFENOXURON DIURON (DCMU) FENURON FLUOMETURON ISONORURON KARBUTILATE LINURON METHABENZTHIAZURON METOBROMURON METOZURON MONOLINURON MONURON MONURON-TCA NEBURON NOREA SIDURON TRIMETURON ACROLEIN PYRAZON 6706 9785 9774 BROMOCIL LENACIL DP-733 ISOCIL TERBACIL There is a need for rapid, simple and low cost toxicity screening procedures which can be used to assess the impact of the increasing number of chemicals on the quality of aquatic and soil environments. This need is seen nowhere more acutely than at intake protection for drinking water, especially when raw water is being abstracted from lowland rivers at risk from industrial or agricultural pollution. Such water may, in some circumstances, be collected, treated and piped to a householder in less than four hours. In such situations monitors must be capable of responding rapidly, perhaps within 30 minutes.

The success of the Ames test for rapid screening of chemical mutagens has stimulated research on the use of bacteria for rapid and inexpensive screening of chemical toxicity. During recent years, for example, the use of microorganisms in aquatic tests has rceived increasing attention. Techniques are becoming more diversified and have been developed to fulfull various screening functions. Amongst these are the Spirillum volutans inhibition test, the triphenyltetrazo liumchloride (TTC), and resazurin, dehydrogenase activity inhibition tests, the Microtox assay protocol and the activated sludge respiration inhibition test. Although enjoying some degree of success, these tests suffer from two major disadvantages. First they require a certain amount of skilled manipulation and are therefore time consuming.Secondly, they cannot easily be applied on line thus making real time analysis impractical. Biosensors, which are devices that transduce a selective biochemical response into an electrical signal can offer practical alternatives in environmental monitoring particularly in terms of cost, ease of manufacture, ability to reduce the test system's complexity to a minimum and suitability for on line applications. A further advantage offered by biosensors is that a specific determination can be carried out in multi-component solutions thereby alleviating the need for any complicated and time consuming separation procedures.

In general terms a biosensor consists of a biological component such as an enzyme, tissue slice or whole cell held in close proximity to the surface of a transducing element. In this configuration the biocatalyst not only confers selectivity on the device but also produces or consumes a species which can be detected by the transducer.

Carbon dioxide specific electrodes have been used in conjunction with immobilised bacteria to produce a whole cell biosensor capable of toxin detection. This particular approach makes the possibility of on-line monitoring more feasible. Where herbicide disturbance of photosynthetic electron transport (PET) systems is being monitored, the biocatalyst must incorporate complete photosystems capable of carrying out the Hill reaction. Isolated thylakoids, chloroplasts and intact photosynthetic prokaryotic or eukaryotic cells are capable of acting as such biocatalysts.

However complexities of preparation and poor stability of isolated membranes and organelles reduce the attractiveness as potential biocatalysts. Cyanobacterial and algal cells, however, are easily maintained in axenic culture and harvested to give uniform batches of material. The absence of membrane bound organelles in cyanobacteria makes these organisms particularly suitable for biosensor use.

The use of whole cells in biosensors offers the additional advantages of increased stability and ease of immobilisation. Mediators, such as soluble, low molecular weight redox couples, are provided to interact at sites along the ET chain and become reduced, in effect acting like terminal electron acceptors. Subsequent reoxidation at the working electrode results in a steady flow of current which is measured by the external circuitry. The magnitude of the current is proportional to the photosynthetic activity of the organisms and any perturbations in the ensuing current/time curve are used to indicate the presence of a pollutant.

The cells may be presented to the electrode immobilised onto the surface of bacteriological filters, although, it is important that the filters allow diffusion of both toxicant and mediator molecules to the cells and diffusion of reduced mediator to the electrode surface.

According to a first aspect of the present invention there is provided a method of detecting pollution in an aqueous liquid flow comprising the steps of adding to a portion of the liquid a compatible electron transfer mediator, passing the mixture into a sensor cell in which is contained a living bacteria means, stimulating an activity of said bacteria means, measuring the level of said activity at an electrode in the cell by means of electron transfer thereto by said mediator.

According to a second aspect of the present invention there is provided a sensor to detect the presence of a pollutant in an aqueous liquid, comprising an electrode, living bacteria means, means to stimulate activity of said bacteria means, a compatible mediator to transmit an indication of the activity level of said bacteria means to said electrode, and means to measure an electrical parameter of said electrode for determination as to normality of the bacteria activity.

In either first or second aspects of the invention, the activity monitored of the bacteria means may be respiration and/or photosynthesis, either of which involves electron transfer.

The mediator preferably operates by electron transfer from the respiratory or photosynthetic pathway of the bacteria means to the electrode, the current flowing from which acts as a measure of the bacteria activity.

The mediator may be potassium ferricyanide, dimethylbenzoquinone or p-benzoquinone either alone or in admixture.

The preferred bacteria means is a cyanobacterium, for example Synechococcus, or alternatively a Eubacteria, for example, E.coli.

An alternative is a eukaryotic alga, for example chlorella.

The means to stimulate activity of the bacteria means may be a light source, for example a bright light emitting diode disposed close to the bacteria means or a high intensity light source and light guide.

The bacteria means may be retained on a translucent alumina filter. In this case, the bacteria may be held by means of a fine nylon mesh.

Alternatively, the bacteria means may be provided with an energy source, for example glucose and/or fructose.

The determination of normality may be by comparison with the electrical parameter output of a control electrode in a sensor having aqueous liquid of predetermined purity.

The electrical current may be measured as derived preferably by means of a carbon or platinum electrode poised in the region of 400-550 mV with reference to silver/silver chloride.

The bacteria means may be an 0.2 ,um bacteriological filter membrane which may be held against electrode surface by nylon mesh.

Alternatively, the bacteria means may be held in a reservoir through which the aqueous liquid and mediator are passed prior to their passing to the electrode.

Embodiments of the invention will now be more particularly described by way of example and with reference to the accompanying drawings, in which: Figure 1 shows diagrammatically a form of flow cell for use in the invention; Figure 2 shows schematically the electron transfer events during photosynthesis of cyanobacteria and the sites of access of mediators; Figure 3 shows graphically the results of a series of control sensor runs with various combinations of bacteria/alga and mediator with LED stimulation; Figures 4 and 5 show continuations of the runs of Fig. 3 with the addition of a pollutant; Figure 6 shows a graphical comparison of a control run and a run with the addition of DCMU pollutant for a Synechococcus sensor; Figure 7 shows graphically the results of a control sensor run of E.coli and mediator with glucose stimulation;; Figure 8 shows an electrochemical cell which may be magnetically stirred; Figure 9 shows schematically a tangential flow electrochemical cell for online use; Figure 10 shows graphically the results of photosynthetic activity monitored in Synechococcus (a) and Chlorella (b) immobilised on cellulose acetate filters and mounted in the flow cells, using ferricyanide and ferricyanide/p-benzoquinone mediators. The working potentials in both instances are 400 mV vs the Ag/AgCI reference electrode.Illumination was provided by LEDs; (+) = LEDs on, (-) = LEDS off; Figure 71 shows graphically the response of a Synechococal biosensor incorporating cells harvested from cultures in stationary phase, immobilised onto alumina filters, and incorporated in a flow cell with a mediator solution of ferricyanide in Bgll. Again illumination was provided by LEDs; (+) = LEDs on, (-) = LEDs off. The working electrode potential was 400 mV vs Ag/AgCI. (The spike is the result of introducing an air bubble into the mediator flow stream to the sensor.); Figure 12 shows graphically the effect of the herbicide DCMU on the response from a Synechococcal biosensor using cells immobilised on cellulose acetate filters with ferricyanide as the mediator.DCMU was added to the reservoir (+DCMU) to give a final herbicide concentration of 233 ppb. Both sensors experienced the same light and dark regimes, illumination being by LEDs; (+) = LEDs on. (-) = LEDs off. The potential at the working electrode was 400 mV vs Ag/AgCI; Figure 13 shows graphically the effect of chlortoluron on the response from a Synechococcal electrode, using cells immobilised on alumina filters with potassium ferricyanide as the mediator.

(+) = illumination on, (-) = illumination off and (+Chl) = an addition of chlortoluron to the electrochemical cell resulting in a final concentration of 2 ppm; Figure 14 illustrates graphically reversible inhibition by the herbicide chlortoluron of a Synechococal biosensor. The cells are immobilised onto alumina filters and the sensor was operated in the stirred glass cell using ferricyanide as mediator at a working potential of 550mV vs Ag/AgCl. Following a one hour flushing of the sensor in mediator free Bgll repression by chlortoluron was partially reversed. Photosynthetic electron transport is again monitored using ferricyanide and a second response to chlortoluron obtained.Chlortoluron was added at (+Chl) to give a final concentration of 2ppm; and Figure 15 shows graphically the effect of linuron on the response obtained from a whole cell biosensor incorporating Synechococal cells immobilised onto alumina filters, with ferricyanide as mediator and poised at 550 mV vs a silver chloride reference. Illumination was provided by an external light source; (+) = light on, (-) = light off. Linuron additions (+ Lin) were made as indicated resulting in final concentrations of 17 ppb and 34 ppb.

Referring now to Fig. 1, a sensor cell comprises an inlet 1 and outlet 2 for passage of aqueous liquid mixed with mediator. A water jacket 3 is provided to maintain an even temperature, if this should be desired. The liquid flows over an electrode 4 which is a carbon or platinum electrode poised at 400 mV with respect to a silver/silver chloride reference electrode 5.

An intense LED 6 is disposed adjacent the electrode 4, spaced in a preferred embodiment 3mm therefrom. Fine nylon mesh holds a 0.2 jtm bacteriological filter against the electrode 4.

Captured on the filter, on the illuminated side, i.e. remote from the electrode, and protected by a dialysis membrane, is a bacteria or alga (herein generally referred to for convenience as bacteria means) which may be: Synechococcus PCC-630 1 Anabaena cylindrica PCC-7122 Anabaena Variabilis ATCC-29413 Chlorella sp.

Escherichia coli Bacillus licheniformis Bacillus subtilis Agrobacterium tumifaciens The first four of the above bacteria means may be stimulated by illumination of the LED, or the final six of the bacteria means may be stimulated by adding glucose and/or fructose to the input liquid.

The sensor cell is arranged with a number of others to receive selectively a portion of the input liquid. The effective life span of the bacteria means varies with species, age and metabolic status, but will require renewal after a number of hours or days, use. In order to provide continuous detection, the input flow can be switched to a renewed cell while the bacteria means in a "used" cell is changed. Further cells can be provided to act as control cells to which liquid of known or standard purity is fed. The cells may be renewed by exchanging a fresh supply of bacteria for the used bacteria or alternatively by supplying the cell with a saline or mineral solution containing nutrient over a "recovering phase", the length of which may vary.

The mediator mixed with the input is potassium ferricyanide, dimethylbenzoquinone and/or pbenzoquinone.

Fig. 2 shows schematically the electron transfer events during photosynthesis by cyanobacteria. The electrons captured by the mediator are transmitted to the electrode 4 where the current can be measured.

Referring now to Figs. 3 to 5, four runs are plotted graphically as follows: A-Cyanobacterium-Synechococcus with 5mM potassium ferricyanide mediator; B -Eukaryotic alga-chlorella also with 5mM potassium ferricyanide mediator; C-Synechococcus with mixture of 5mM ferricyanide and 0.5 mM p-benzoquinone; and D-Chlorella with the same mediator mixture as in C.

At intervals, once the bacteria means had a stabilised activity, the LED was illuminated, as indicated by points L. It was switched off again at D. As can be seen, in all except one case, illumination produced increased activity (as measured at the electrode) fairly quickly.

At dimming the activity, again with the one exception, diminishes. The pattern in each case is however, regular. The exceptional case would seem to show that eukaryotic alga will not function properly in the absence of p-benzoquinone.

In Figs. 4 and 5, a herbicide 3-(3,4-dichlorophenyl)-1,1-dimethylurea (DCMU) was added to a mediator reservoir at the points marked +, to give a final concentration of 0.75 x 10-6 M (c.

175 ppb). The effects can be seen clearly from the Figures.

The effect is shown even more clearly in Fig. 6 which shows the effect of addition of 1 x 10-6 M (c. 233 ppb) DCMU at the point marked + to a Synechococcus sensor, as opposed to a control Synechococcus sensor. The addition gives a marked reduction of bacterial activity within 2 or 3 minutes.

Finally Fig. 7 shows graphically the activity of a Eubacteria-Escherichia Coli as transferred by a ferricyanide mediator, when activated at intervals at points t by addition of glucose to the input, and deactivated at l by returning the input to a mediator/saline mixture.

The method and apparatus disclosed provide a useful means of determining quickly whether or not a pollutant is present in the water intake. When a signal is received that a pollutant may be present, the flow can be interrupted and the nature of the pollutant determined by analytical or other techniques.

The invention has been described with reference to particular bacterial and alga. Others may of course be applicable, as may other compatible mediators.

Since the mediator transfers the activity of the bacteria to the electrode, it is not strictly necessary for the bacteria to be immediately adjacent the electrode. The bacteria may be contained in a reservoir remote from the electrode and the mediator/input liquid passed from the reservoir to the electrode.

Fig. 9 shows a tangential flow electrochemical cell in which the analyte, mixed with a mediator, flows across a sensor electrode while being periodically illuminated by means of the LED.

The remaining Figures show graphically the results of using various bacteria means in the method of the invention. One point to note is that, as shown in Fig. 11, an air bubble may cause a spike. Thus it is advisable that the sample should first pass through a bubble trap to eliminate any such complications.

It is possible to maintain the shelf life of the bacteria means used for a period of up to many months. The cells can be presented to the sensor electrode captured in calcium alginate.

Bacterial or cyanobacterial cells may be harvested during logarithmic growth and mixed with sodium alginate in a concentration in the range of 1-4% weight/vol. to give the desired cell density. The sodium alginate is then solidified, by imersion in 0.5 molar calcium chloride solution at 4"C. for six hours, into beads or sheets. Following rinsing in distilled water, the alginate is air dried for twelve hours at 25"C. and stored in sealed glass containers until needed.

Prior to use on the sensor electrodes, the alginate is re-hydrated in an appropriate bathing medium (such as BG 11, or 0.9% saline solution).

Referring now more particularly to Fig. 10, the response obtained with Synechococcus and Chlorella based biosensors to a regime of light and dark periods in the presence of either ferricyanide or a mixture of ferricyanide and pbenzoquinone is shown. The inability of ferricyanide to access eukaryotic photosynthetic activity is clearly seen as no response is recorded from the Chlorella electrodes following illumination. A cocktail containing a lipophilic mediator is required to access the chloroplast ET chains. It is believed that benzoquinone alone will act as an efficient mediator in this system.

With both Synechococcus and Chlorella based sensors, the use of pbenzoquinone results in a sequential decline in the maximum current obtained during the period of illumination, ultimately resulting in total loss of response.

In the case of the Synechococcus biosensors, PET chain events can be monitored using ferricyanide alone, with a less marked decline in the light response. A working life for Synechococcal biosensors of 3-4 days continuous use, if not more, can be achieved using ferricyanide.

Typically with Synechococcus, if cells are used immediately after harvesting from illuminated batch culture, a substantial dark respiratory response is observed (early portion of the curve) which must be exhausted before the sensor is used for photosynthetic monitoring. This response is not seen however, with Chlorella or indeed in Synechococcus when tbenzoquinone is present in the electrolyte.

Of the three cyanobacteria tested in this investigation, Synechococcus was found to be the most successful, whilst PET activity could be monitored in both Anabaena cylindrica and Anabaena variabilis, the use of these filamentous forms proved more difficult in terms of obtaining optimum loading levels on the filters. In addition, the response times were longer with these organisms than with Synechococcus. Based on these and other results the unicellular cyanobacterium Synechococcus PCC 6301 was selected as the biocatalyst for use in all subsequent work.

Fig. 11 shows typical results obtained from a Synechococcus probe produced from cells harvested from cultures during their stationary phase of growth. Following illumination, such sensors showed a rapid increase in current which reached a maximum peak value. The current then declined to a lower but more stable value. This type of overshoot effect, which is only seen with cells harvested from older cultures, is evident in all subsequent light stimulations. The magnitude of the overshoot together with the steady portion of the curve is consistent throughout the sensor run. The spike towards the end of the sensor run demonstrates the effect of an air bubble introduced into the flow stream. In order to guard against this during on line applications, a bubble trap on the delivery line to the flow cell needs to be fitted.

To confirm that the sensor response was indeed the result of PET, the effects of DCMU, a known cyanobacterial photosynthetic inhibitor was investigated. Using Synechococcus biosensors in the tangential flow ce!l configuration, the response to the addition of 1 x 10 6 M (233 ppb) DCMU was compared to a control sensor, as is shown in Fig. 12. Following a short transit time taken to deliver the herbicide to the biosensor from the receiving reservoir, there was a rapid cessation in PET activity; as evidenced by the drop in sensor current which mirrors that seen on the removal of illumination to the control sensor. So long as the DCMU is present in the flow stream no recovery in the light response could be obtained from the "poisoned sensor".

Subsequently, two other members of the DCMU family of herbicides, chlortoluron and linuron, were investigated. The effect of an addition of chlortoluron to give a final herbicide concentration of 2 ppm, on a Synechococcus biosensor shows in Fig. 13 an equally rapid fali in photosynthetic ET activity.

The reversibility of the effects of these herbicides is shown in Fig. 14 where chlortoluron inhibition of Synechococcal photosynthesis at a concentration of 2 ppm is reversed following a one hour recovery period during which the sensor is flushed with fresh Bgll medium. Similar results, both in terms of inhibiting the photosynthetic response and recovery after washing were obtained for the herbicides DCMU and linuron.

Fig. 15 shows the results of a two pulse addition of Linuron, final concentration 17ppb and 34ppb. Although the decrease in current is not as rapid as that seen for higher concentrations (200 ppb-2 ppm) it is nevertheless substantial and can be used as an indication of the presence of herbicide pollution, even at these extremely low levels.

The particular mediators disclosed are those presently though to be most advantageous.

However, in order to ensure the successful detection of a particular herbicide it is important to choose the correct mediator molecule. In order to detect the presence of a PSII inhibitor for example, the mediator must be reduced downstream from the herbicide's site of action. Similarly, in order to monitor the presence of a PSI blocker, the mediator should be reduced at the distal end of the PET i.e., near the terminal electron acceptor site. In this context ferricyanide is a good choice for a mediator, not only because of its chemical stability in aqueous solution and straightforward electrochemistry at the working electrode, but also because it interacts with the PET at the end of PSI.Consequently, by using ferricyanide as the mediator it should be possible to monitor all herbicides which prevent the flow of electrons through PSII and those which interrupt electron flow through PSI before the ferrodoxin, the putative ferricyanide reduction site.

These sensors may be applied to detecting a wide range of herbicide families such as nitriles, triazines, bis-carbamates, pyridazines, nitrophenyl ethers, uracils and anilides.

The sensor configuration shown in Fig. 9 could have an important role in detecting pollution plumes in river water, and is intended to supplement existing monitoring systems. Biosensors have the advantages of being cheap to produce, the sensor electrode being a disposable device, and requiring minimal pretreatment, i.e. removal of suspended solids and air bubbles from the sample flow to the sensor. The system can also be fully automated, opening up the possibility of locating sensors at remote sites.

The invention will be further described with reference to the following Example.

EXAMPLE Carbon foil indicator electrodes were manufactured in the following manner. 5mm diameter discs were punched out of a 1 mum thick strip of porous graphite, washed twice in acetone (30 minutes for each wash with gentle agitation), once in boiling distilled water (1 hour), and dried in air oven at 100"C. Electrical contact to the discs was made by cementing lengths of 0.2mm wire to the surface of the graphite with a drop of electrically conducting (silver loaded) epoxy resin.After curing, the graphite discs were potted in a mixture of 9 parts epon resin (grade 815) and 1 part triethylenetetramine catalyst at a temperature of 60"C. Immediately prior to use, the electrodes were conditioned electrochemically in 0.1 M potassium dihydrogen orthophosphate by cycling the potential between -0.8 and +1.or vs Ag/AgCI 100 times at a sweep ?ate of 1V/sec.

Anabaena variabilis (ATCC 29413) and Synechococcus (PCC 6301), Anabaena cylindrica (PCC 7122) and Chorella sp. were used. Cultures of anabaena were maintained on agar plates made from BG11O medium, nitrate-free BG11 medium supplemented with 1.5% (w/v) bacteriological agar. Synechococcus and Chlorella cultures were maintained on agar suppiemented BG11 medium. Liquid batch cultures were prepared by transfer of inoculum from plate cultures to 50 ml sterile BG1 lo or BG11 medium in conical flasks. Liquid cultures were maintained at 25"C., with 5.0 W m 2 illumination to a mid exponential phase (OD663= 1.0) and the cells were harvested by rapid centrifugation (12,000 rpm for 60 secs.).The cells were resuspended in 5081 of fresh culture medium and applied to the surface of 5mm diameter bacteriological filter discs.

Two different techniques were employed in the present Example. Initially cells were loaded onto 0.2ijm cellulose acetate filter dsics. The filters were then placed onto the sensor electrode with the cell loaded surface exposed. Dialysis membrane was used to sheath the electrode and hold the filter and cells in place. Alumina filters which are translucent when wet can now be used to replace cellulose acetate. In this case the alumina filters are presented to the electrode with the cell layer against the electrode surface. The filter is held in place by a fine nylon mesh, removing the need for a dialysis membrane. Prior to use, the assembied electrodes were conditioned by soaking in BG11 medium for approximately 10 minutes.

Claims (35)

1. A method of detecting pollution in an aqueous liquid flow comprising the steps of adding to a portion of the liquid a compatible electron transfer mediator, passing the mixture into a sensor cell in which is contained a living bacteria means, stimulating an activity of said bacteria means, and measuring the level of said activity at an electrode in the cell by means of electron transfer thereto by said mediator.

2. A method as claimed in claim 1, wherein the activity monitored of the bacteria means is respiration and/or photosynthesis.

3. A method as claimed in claim 2, wherein the mediator operates by electron transfer from the respiratory or photosynthetic pathway of the bacterial means to the electrode, the current flowing from which acts as a measure of the bacteria activity.

4. A method as claimed in claim 3, wherein the mediator is potassium ferricyanide, dimethylbenzoquinone or p-benzoquinone eithr alone or in admixture.

5. A method as claimed in any one of the preceding claims, wherein the bacteria means is a cyanobacterium.

6. A method as claimed in claim 5, wherein the cyanobacterium is Synechococcus.

7. A method as claimed in any one of claims 1 to 4, wherein the bacteria means is a Eu bacteria.

8. A method as claimed in claim 7, wherein the Eubacteria is E.coli.

9. A method as claimed in any one of claims 1 to 4, wherein the bacteria means is a eukaryotic alga, with the proviso, when dependent on claim 4, that the mediator is not potassium ferricyanide alone.

10. A method as claimed in claim 9, wherein alga is chlorella.

11. A method as claimed in any one of claims 5, 6, 9 or 10, wherein the means to stimulate activity of the bacteria means is a light source.

12. A method as claimed in claim 11, wherein the light source is a bright light emitting diode disposed close to the bacteria means.

13. A method as claimed in any one of claims 7 to 10, wherein the bacteria means activity is stimulated by provision of an energy source.

14. A method as claimed in claim 13, wherein the energy source is glucose and/or fructose.

15. A method as claimed in any one of the preceding claims, wherein the bacteria means are retained on a translucent alumina filter.

16. A method as claimed in any one of the preceding claims, wherein the determination of normality is by comparison of an electrical parameter output of the electrode with that of a control electrode in a sensor having aqueous liquid of predetermined purity.

17. A method of detecting pollution in an aqueous liquid flow substantially as described herein with reference to the accompanying drawings.

18. A sensor to detect the presence of a pollutant in an aqueous liquid, comprising an electrode, living bacteria means, means to stimulate activity of said bacteria means, a compatible mediator to transmit an indication of the activity level of said bacteria means to said electrode, and means to measure an electrical parameter of said electrode for determination as to normality of the bacterial activity.

19. A sensor as claimed in claim 18, wherein the mediator is potassium ferricyanide, dimethylbenzoquinone or p-benzoquinone either alone or in admixture.

20. A sensor as claimed in claims 18 or 19, wherein the bacteria means is a cyanobacterium.

21. A sensor as claimed in claim 20, wherein the cyanobacterium is Synechococcus.

22. A sensor as claimed in claims 18 or 19, wherein the bacteria means is a Eubacteria.

23. A sensor as claimed in claim 22, wherein the Eubacteria is E.coli.

24. A sensor as claimed in claims 18 or 19, wherein the bacteria means is a eukaryotic alga, with the proviso, when dependent on claim 19, that the mediator is not potassium ferricyanide alone.

25. A sensor as claimed in claim 24, wherein the alga is chlorella.

26. A sensor as claimed in any one of claims 20, 21, 24 or 25, wherein the means to stimulate activity of the bacteria means is a light source.

27. A sensor as claimed in claim 26, wherein the light source is a bright light emitting diode disposed close to the bacteria means.

28. A sensor as claimed in any one of claims 22 to 25, wherein the activity of the bacteria means is stimulated by provision of an energy source.

29. A sensor as claimed in claim 28, wherein the energy source is glucose and/or fructose.

30. A sensor as claimed in any one of claims 18 to 29, wherein the bacteria means are retained on a translucent alumina filter.

31. A sensor as claimed in any one of claims 18 to 30, further comprising a control electrode in a sensor having aqueous liquid of predetermined purity, to allow the determination of normality by comparison with the electrical parameter output thereof with that of the electrode.

32. A sensor as claimed in any one of claims 18 to 31, wherein electrical current is measured as derived by means of a carbon or platinum electrode poised in the region of 400-550 mV with reference to silver/silver chloride.

33. A sensor as claimed in any one of claims 18 to 32, wherein the bacteria means is an 0.2 um bacteriological filter membrane held against the electrode surface by nylon mesh.

34. A sensor as claimed in any one of claims 18 to 32, wherein the bacteria means are held in a reservoir through which the aqueous liquid and mediator are passed prior to their passing to the electrode.

35. A sensor substantially as described herein with reference to the accompanying drawings.