ELECTROCHEMICAL BIOSENSOR STABILITY
This invention relates to an electrochemical biosensor of the kind wherein an enzyme-containing reagent system is applied to an electrically conductive support member and which responds amperometrically to the catalytic activity of the enzyme in the presence of its substrate.
Immobilisation of an enzyme upon an electrode impedes conduction of electrons between the active site of the enzyme and the electrode. Electron carriers or mediators have been employed to overcome this disadvantage and to enhance the reliability and sensitivity of the electron transfer .
Mediator assisted biosensors do not give consistent results upon repeated use. especially those made by screen printing methods. Commercial products are designed for use once or twice only, after which they must be discarded.
According to a first aspect of the present invention an electrochemical biosensor comprises: a conductive support member; an enzyme system including a mediator immobilised on the support member: and a covering membrane. characterised in that the membrane incorporates a quantity of a mediator.
The mediator contained in the membrane may be the same or different to that of the enzyme system. Preferably the mediators are the same.
According to a second aspect of the present invention an electrical biosensor comprises: a conductive support member: an enzyme system incorporating a mediator immobilised on the support member: and a conductive membrane: characterised in that the biosensor provides a
substantially quantitative and perometric response when used for 50 repeat tests with an analyte.
We have found surprisingly that provision of a quantity of mediator within the membrane provides a biosensor which may be used repeatedly, for example preferred biosensors may be used 1000 times or more, without significant reduction in the quantitative and amperometric response. Biosensors in accordance with this invention may be incorporated into automatic analytical equipment for example for monitoring analytes in a sample stream, or implantation to allow in vivo metabolic analysis. Further applications include blood analysis for example for glucose, lactate, creatinine, urea, pyruvate etc. The invention may also be used for process control for industrial fermentation and brewing, amino acid production and for food processing, for example for monitoring glucose levels in meat. Monitoring of mammilarian cell cultures is also facilitated.
Membranes which may be applied to the enzyme containing layer of the biosensor act as a diffusion barrier, preventing loss of the components of the enzyme containing layer.
The membrane may be composed of various materials. Preferred materials include ethyl hydroxyethyl cellulose, ethyl cellulose, cellulose acetate, polyvinylchloride, polyurethane. polycarbonate, cellulose nitrate, polyurethane, cellulose acetate, aryl polyethers for example polyether sulphone or polyether ketone including functional ised polyaryl ethers, such as sulphonated polyether sulphonate.
Anionic or cationic ion exchange membrane materials may be used. For example Nafion membranes may be used to retain the mediator in the charged state.
In most applications non-ionic membranes are preferred.
Membrane thickness of 0.1 to 100 microns may be used.
In preferred embodiments of the invention wherein the
enzyme system is partially soluble in aqueous solvents, the membrane and mediator are preferably soluble in a non aqueous solvent. This prevents leaching of the enzyme reagents during casting of the membrane. It also prevents solubi 1 isation of the printed electrode matrix. The solvents for the membrane and mediator may be a mixture of two or more solvents, one of which may be a good solvent for the mediator and another being a good solvent for the membrane material, the volatilities of the solvents being selected so that the concentration of the mediator across the membrane may be selectively controlled.
The concentration of the mediator in the membrane may be up to 50m . g . preferably 5 to 20mg. for example lOmg.g .
The mediators which may be employed may be selected from several classes i Heterofulvalene Pi-donors for example tetrathiafulvalene and tetraselenafulvalene. ii Metallocenes for example ferrocene, nickelocene and selenocene. iii Ouinones for example benzoquinone. iv Metal complexes based on platinum group metals or transition group elements and organic ligands: for example platinates. ruthenates and tungstates. v Organic metal complexes for example TCNQ complexes. TCNQ derivatives and NMP+NCNQ" and TTF+TCNQ~- type charge transfer complexes.
The enzyme containing layer may comprise conductive carbon particles coated with the enzyme, mediator and a binder. The binder may be chemically similar to the membrane material. For example a .- droxyethy1 cellulose binder may be used in conjuction with an ethyl hydroxyethyl cellulose membrane material. This affords compatibility between the enzyme layer and the membrane and good adhesion .
A thin layer of non-mediator membrane material (EHEC)
may be cast over the electrode before the mediator- containing membrane. This provides good adhesion but prevents excess mediator leaching from the membrane into the electrode. Covering membranes may also be required to confer biocompatibi 1 it , antifouling or diffusion properties needed to suit the application.
Membranes may be applied by spin or dip coating or by ink jet printing. Ink jet printing has the advantage that the mediator-doped membrane is only applied to selected regions of the substrate. The mediator may also be included in precast membranes which are subsequently applied to the electrode. Such precast membranes may comprise cellulose nitrate, cellulose acetate polycarbonate, polyamide or other commercially used membrane materials. Precast membranes may be secured to the electrode by conventional methods, eg by adhesive or using physical attachments.
The invention is further described by means of example but not in any limitative sense, with reference to the accompanying drawings of which:
Figure 1 illustrates repeated test of a TTF membrane covered TTF/GOD electrod in accordance with this invention in 11 mM glucose:
Figure 2 illustrates calibration of TTF membrane covered and free printed TTF/GOD electrodes;
Figure 3 illustrates repeated test of mediator free membrane covered TTF/GOD electrodes in 11 mM glucose;
Figures 4 to 7 illustrate calibration curves and repeat test results for the electrodes:
Figure 8 illustrates calibration curves and repeat test results for the electrodes of Example 3;
Figure 9 illustrates plan and sectional views of a preferred spein or dip coated electrode in accordance with this invention; and
Figure 10 illustrates plan and sectional views of a
second preferred electrode.
Graphite powder grade T * Q was obtained from Morganite Electrical Carbon Ltd. Swansea, Vales. Micronised silver chloride was obtained from MCA Services. Cambridge, England. Polyviny1chloride substrate (0.650mm thick) was obtained from Mazzucchelli Limited, Wellington, England. Silver (Electrodag 477 SS RF) and carbon (Electrodag 423 SS) inks were purchased from Acheson Colloids Co. Plymouth. England. Hydroxyethyl cellulose, ethylene glycol and diethyleneglycol monoethyl ether acetate were obtained from Fluka Chemicals Limited. Glossop, England. Ethyl hydroxy ethyl cellulose (high viscosity grade) was purchased from Polysciences . Inc, Varrington, USA. All other chemicals were of reagent grade.
1. The working electrode (WE) ink was prepared as follows:
(i) Weigh 133.6mg TTF (Aldrich) in to a glass uni ersal bottle (in 30ml capacity) and add approximately 5ml diethyl ether (Fluka) .
(ii) Weigh 2g TJQ graphite (Morganite) into a large, high-sided glass petri dish.
(iii) Ada the TTF dissolved in diethyl ether to the graphite. Wash out any remaining TTF from the universal bottle with approximately 4ml diethyl ether and add the washings to the graphite.
(iv) In a fume hood, mix the TTF and graphite using a spatula until all the diethyl ether has evaporated and the TTF is deposited on to the
(v) Weigh 536mg GOD (Glucox-PS:Sturge) into a straight-sided, wide-necked screw top jar
( 100ml capacity) . Add 4.67g of a two percent (w/v)) aqueous solution of hydroxy ethyl cellulose (HEC) (Fluka) containing six percent (w/v) ethylene glycol (Fluka) and mix to dissolve the GOD.
(vi) Mix the graphite/TTF into the GOD solution and rotate on an angled (45°) rotary mixer at 4rpm for two to three hours at room temperature.
NB a For dimethyl ferrocene (DMF) the above precedure was repeatedbut DMF was substituted weight-for-weight for TTF.
b For tetracyanoquinodimethane (TCNQ) the above procedure was repeated but substituting TCNQ for TTF in a weight- for-weight manner. In addition, the TCNQ was dissolved in 200ml toluene (rather than in diethyl ether for step (iii) above). In this case, the TCNQ/toulene washings (approximately 20ml) were also added to the T-^Q graphite powder.
2. The reference electrode (RE) ink was prepared as follows:
Micronised silver chloride was added to silver ink to a content of thirty five percent (w/w) and mixed thoroughly. The mixture was thinned to a printable consistency by the addition of nine percent (w/w) diethyleneglycol monoethyl ether acetate.
3. The silver, carbon and insulating inks were used as received.
The sensor consisted of a PVC substrate (0.65mm
thick) upon which was printed a series of layers (see Figure 9) . All layers were printed through polyester screens using a Model 245 screen printer (DEK Printing Machines Ltd, Weymouth , England) . Electrical contact to the WE and RE pads was provided by silver tracks. A carbon layer was positioned over the conductive tracks ensuring that the silver did not make contact with the analyte solution. The WE and RE pads were positioned over the carbon layers. An insulating layer, with openings allowing external electrical contact to one end of the strip and analyte access to the WE and RE at the other, was printed over the whole strip. Four such sensors were printed on each substrate.
An electrochemical interface under IBM-compatible computer control was used to measure simultaneously the current from four strip sensors supported in a thermostatically controlled glass cell containing 20 ml quiescent test solution. A potential of 200mV vs Ag/AgCl (printed) reference electrode was applied to the working electrode and the current was measured after thirty seconds .
Example 1
Standard TTF/GOD printed electrodes (HEC binder: see experimental procedures) were modified by spin-coating with a 6% (w/w) solution of EHEC (ethylhydroxyethyl cellulose) in toluene containing lOmg TTF. The spin-coating was carried out at approx lOOOrpm for two to five seconds. The membrane coated electrodes were dried rapidly (two to five seconds) in a stream of air.
Repeat tests were carried out in llmM glucose in 20mM sodium phosphate (pH7.4) containing 0.1M KCl at ?^°C. The tests were for thirty seconds duration at a poter__ial of +200mV v. Ag/AgCl (printed).
No significant loss in activity of the coated electrodes was observed after 100 assays. The untreated
electrodes, however, lost approximately ninety per cent activity by the fiftieth assay. The results are illustrated in Figure 1.
The calibration curve between zero and 44.41mM glucose had a higher background than the untreated electrodes and little change in signal could be observed between zero and 2.22mM glucose. This may have been due to the high TTF concentration or it may be due to the requirement for some degree of pre-treatment. the coated electrodes showed increasing responses between 22.2 and 44.41mM glucose, whereas the untreated electrodes did not. This flattening of the calibration curve was thought to be due to diffusion limitation to glucose; an expected result (Figure 2).
Simple treatment with an EHEC membrane with no extra TTF present failed to show the same degree of stability (Figure 3) and in the calibration curves (Figures 4 and 5) the zero glucose value was not high as for the TTF- loaded membrane-coated electrodes.
Example 2
Standard TTF/GOD printed electrodes (HEC binder; se previous description of base sensor) were modified by spin- coating with a 6% (w/w) solution of EHEC (ethylhydroxyethyl cellulose) in toluene containing 10 mg g_1 TTF. The spin- coating was caried out at approx lOOOrpm at slow speed for five seconds. The membrane coated electrodes were dried rapidly (two to five seconds) in a stream of air followed by storage in a vacuum at room temperature in the presence of silica gel desiccant for up to two days prior to use.
Repeat tests for stability characterisation were made in 6.67mM glucose in 20 mM sodium phosphate, pH7.40 containing lOOmM potassium chloride. The tests were of thirty seconds duration at a potential o 200mV v. Ag/AgCl (printed) at 25°C in unstirred solution.
The electrodes retained over fifty percent of their initial activity after 500 tests in aqueous 6.67mM glucose.
A calibration curve after 500 samples demonstrated that the electrodes also retained sensitivity to glucose in the range zero to 44mM.
Example 3
Printed electrodes were prer ".red using DMF as the mediator instead of TTF. It was included in the same quantities using the same technique as for TTF (see experimental procedures).
[ ( i ) + (ii) = Treatment A for DMF].
(i) The membrane was formulated as follows:
lOmg DMF was weighed into a scintillation vial insert and lg of six percent (w/w) EHEC in toluene was added. The vial was slowly rotated so as to dissolve the DMF by spin coating.
(ii) The membrane was applied at approx lOOOrpm for two seconds.
(iii) Electrodes were tested as three duplicate pairs - ie. No treatment. Treatment A control (EHEC only) and Treatment A.
The following results were observed:
(i) Untreated DMF/GOD electrodes appeared to fall apart on repeated testing and, therefore, failed to yield a calibration.
(ii) Treatment A control - EHEC (six percent (w/w) ) alone appeared to hold the sensor WE together - sufficiently so as to allow calibration and repeated testing.
(iii) Treatment A produced electrodes with stable responses for over 100 tests in buffered glucose (6.67mM, pH7.40, 25°C) with approximately ninety percent of maximal activity remaining. The control electrodes, however, had only sixty percent activity remaining after 100 tests.
(iv) The calibration curve of electrodes modified by treatment A had lost fifty percent sensitivity but retained the same profile throughout the glucose concentration range. The control electrodes, however, had lost more than ninety percent activity and saturated above lOmM glucose.
The EHEC membrane appeared to:
a. reduce the DMF loss: and b. hold the physical structure together.
The EHEC/DMF layer conferred increased stability on the electrode by providing a reservoir of DMF.
The printed strip electrode shown in Figure 9 comprises a PVC laminar substrate 1 carries conductors 2 connected to contact 3 in the conventional manner. Carbon pads 4 overlie the conductors 2 and are surmounted by working electrode and reference electrode pads 5, 6 respectively. Insulation layers 7 seal the edges of the electrode and mediator loaded membrane 8 is applied over the electrodes 5, 6.
Figure 10 shows an alternative embodiment of the invention wherein the mediator containing membrane is applied to a selected area of the biosensor by ink jet printing. The arrangement is generally similar to that described with reference to Figure 10 with the exception that the mediator containing membrane layer 18 is applied
to overlie the working electrode 15. leaving the remainder of the biosensor uncoated. This results in economical use of materials and allows subsequent processing steps to be performed on the reference electrode or elsewhere on the biosensor. Avtomated manufacturing is facilitated.