- Email: [email protected]
Improved operational stability of biosensors based on enzyme-polyelectrolyte complex adsorbed into a porous carbon electrode1
Biosensors & Bioelectronics 13 (1998) 1205–1211
Improved operational stability of biosensors based on enzymepolyelectrolyte complex adsorbed into a porous carbon electrode1 V. G. Gavalas a, N. A. Chaniotakis b
a,*
, T. D. Gibson
b
a Laboratory of Analytical Chemistry, Department of Chemistry, University of Crete, 71409 Iraklion, Crete, Greece Enzyme Biotechnology Group, Department of Biochemistry and Molecular Biology, Astbury Building, University of Leeds, Leeds, West Yorkshire LS2 9JT, UK
Received 19 February 1998; received in revised form ??/??/??; accepted 13 May 1998
Abstract A novel porous active carbon is utilized in order to adsorb the diethylaminoethyl-dextran (DEAE-dextran)-enzyme stabilized complexes, for the construction of highly stable biosensors. The interaction of DEAE-dextran with the examined enzymes increases dramatically the operational stabilization of the sensors, without adverse effects on the enzyme activity. At the same time, the porous active carbon allows for high enzyme loading, good electrical contact and low resistance throughout the sensing element. Glucose oxidase and horseradish peroxidase are used as model enzymes in this study to construct biosensors, with very good reproducibility (less than 5% RSD). As a result, the glucose sensor exhibits very long operational stability (over a period of 5 months), while the hydrogen peroxide sensor retains its initial activity after several weeks. 1998 Elsevier Science S.A. All rights reserved. Keywords: Porous carbon electrode; Polyelectrolyte; Operational stability; Glucose oxidase; Horseradish peroxidase
1. Introduction The performance of biosensors depends on the stability of the electrochemical transducer as well as on the retention of the activity of the enzymes employed. During the last few years, many efforts have been directed towards the stabilization of biosensors by careful optimization of these factors. The commonly used electrochemical transducers are either metal electrodes (De la Guardia, 1995; Khan and Wernet, 1997) or carbonaceous materials (Gorton, 1995; Cso¨regi, 1994). The widespread use of carbon paste as the supporting matrix for the construction of biosensors is attributed to the simple construction procedures required, the low background current they exhibit, their ability for surface regeneration, and its very low cost. The prime drawback of carbon paste electrodes is the difficulty in reproducing the composition of the paste, leading to irreproducible sensors. * Corresponding author. Tel.: ⫹ 30-81-393-137; Fax: ⫹ 30-81-210951; E-mail: [email protected] 1 This paper was presented at the Fifth World Congress on Biosensors, Berlin, Germany, 3–5 June 1998.
Simultaneously, carbon paste biosensors usually suffer from short lifetime, which is attributed to leaching of the mediator and/or the enzyme out of the paste, in conjunction to the denaturation of the enzyme within the highly lipophilic environment of the paste. It is thus evident that the immobilization of the enzyme in the sensor element is of crucial importance to the biosensors’ operational and storage lifetime. Various techniques have been applied in order to keep or supply adequate quantities of active enzyme onto the electrode surface. Physical adsorption, covalent attachment, crosslinking, and entrapment in polymer films (Cass, 1990; Emr and Yacynych, 1995) are some of the most commonly used ones. During the immobilization process, it is vital that the enzyme retains its activity and specificity. Recently, it has been shown that certain additives can dramatically increase the stability of the enzymes (Appleton et al., 1997; Lutz et al., 1995; Wang and Naser, 1994; Yabuki et al., 1994), even though this effect is not yet completely elucidated. It is believed though that this increase in the enzyme’s stability is achieved through electrostatic interactions that protect the enzyme, and aid it in retaining its active conformation during immobilization in hostile environments.
0956-5663/98/$ - see front matter 1998 Elsevier Science S.A. All rights reserved. PII: S 0 9 5 6 - 5 6 6 3 ( 9 8 ) 0 0 0 6 6 - 9
1206
V.G. Gavalas et al. / Biosensors & Bioelectronics 13 (1998) 1205–1211
In this paper, we report the use of a novel porous active carbon (Vamvakaki and Chaniotakis, 1996) for the immobilization of polyelectrolyte stabilized enzymes for the construction of biosensors with high operational and storage lifetimes. This is achieved with the physical adsorbed of the enzyme-polyelectrolyte complex into the electrode pores, without using chemical coupling agents, thus avoiding any unnecessary denaturation. The use of the polyelectrolyte diethylaminoethyl-dextran (DEAEdextran) as the enzyme stabilizing polyelectrolyte is used to illustrate the achieved stability of glucose and hydrogen peroxidase biosensors. The combination of the porous active carbon and the positively charged polyelectrolytes is a very promising method for the construction of a variety of biosensors with extended operational stability, good reproducibility and fast response times.
2. Experimental 2.1. Materials Glucose oxidase (GOx) from Aspergillus niger (EC 1.1.3.4, 200 U/mg) and peroxidase (HRP) from horse radish (EC 1.11.1.7, 731 U/mg) were purchased from Fluka. Diethylaminoethyl-dextran (DEAE-dextran) (D9885) was obtained from Sigma. D(+)-glucose, hydrogen peroxide (30%), potassium dihydrogen phosphate and potassium hydrogen phtalate were also purchased from Fluka. The dialysis membrane was a pre-mounted type ‘C’ cellulose acetate. The porous active carbon was obtained from KOH.I.NOOR HARDTMUTH (Austria). The density of the carbon was 1.16 ⫾ 0.03 g/cm−3 with porous volume of 0.336 ⫾ 0.003 cm−3/g. All other reagents used were of analytical grade. 2.2. Measurements and apparatus
2.3. Enzyme immobilization A rod (3.1 mm diameter, 5.0 mm height) of the porous active carbon was prepared by cleaning in a sonicated ethanol bath for 10 min and for 10 min more in water bath. The rod was then dried at 430 K for 30 min, and was ready for use. The enzyme was dissolved in the proper buffer solution to a final concentration of 2500 U/cm−3, and then the appropriate amount of polyelectrolyte was added. The enzyme was allowed to react with the polyelectrolyte for 20 min at room temperature. Finally, the carbon rod was placed into the enzyme solution, and allowed to adsorb enzyme solution for 20 h at room temperature. The concentration of the enzyme in the carbon matrix is the same as the initial enzyme concentration, as shown by absorbance measurements of the enzyme. Since the solution uptake of the carbon is 34% of its weight, the enzyme loading of the resulting electrode is controlled by the initial enzyme concentration in solution. The carbon rode was then removed from the solution, washed thoroughly with buffer solution, and placed in a Teflon holder for testing. The electrical contact was achieved through a platinum wire from the reverse side. During the immobilization of the GOx, phosphate buffer 10 mM, pH 7.5 was used, while a 10 mM, pH 5.0 potassium hydrogen phthalate was the buffer of choice for the HRP sensor. In the case of the glucose biosensor, a dialysis membrane was placed on the outer surface of the carbon rod.
3. Results and discussion 3.1. Glucose biosensor It is well known that GOx catalyzes the following reaction: D ⫺ glucose ⫹ O2→gluconic acid ⫹ H2O2
All experiments were carried out using a three electrode Metrohm 641 VA-Detector, a silver/silver chloride double junction reference electrode (Model 90-02, ORION Research) and a platinum counter electrode. Temperature control at 297.0 ⫾ 0.1 K was achieved with a circulating bath (Model 9105, PolyScience). The flow injection system consisted of a wall-jet flow cell, an injection valve with loop volume 400 l, while the solvent delivery was done using either a sage syringe pump (Model 362, ORION Research) or a peristaltic pump (No 7554-30, Cole-Parmer). The signal was recorded in a personal computer equipped with a 16 bit A/D converter and controlled with a software written in basic. In all experiments nano-pure water ( 苲 18 M⍀, EASYpure model D7033, Barnstead) was used.
while the transduction of the biochemical recognition is achieved via the following electrochemical redox reaction: H2O2→2H+ ⫹ O2 ⫹ 2e− Initially the effect of pH on the sensor’s response was evaluated using 10 mM phosphate buffers applying potential ⫹ 800 mV vs Ag/AgCl. The results obtained are illustrated in Fig. 1. As it can be seen, the biosensor’s response is dramatically increased by changing the pH from 6.5 to 7.5 and then levels off. The operating pH was chosen to be 7.5, even though from the figure the optimum pH seems to be above this value. The apparent increase in the response of the sensor for pH above 8.0
V.G. Gavalas et al. / Biosensors & Bioelectronics 13 (1998) 1205–1211
1207
Fig. 1. Influence of pH on the glucose biosensor. The current refers to injections of 10 mM glucose response. Applied potential ⫹ 800 mV vs Ag/AgCl, in a 10 mM phosphate buffer.
is due to the higher sensitivity of the carbon matrix to peroxide (8.0 ⫾ 0.2 A/mM at pH 7.5 and 11.5 ⫾ 0.4 A/mM at pH 8.5), and not due to the increase of the enzyme efficiency. Fig. 2 shows the hydrodynamic voltammogram obtain with potentials between 0 and ⫹ 800 mV for the glucose biosensor. The response to the enzymatically produced H2O2 increase linearly above ⫹ 500 mV, while even at ⫹ 400 mV the biosensor is analytically usable. Subsequent experiments were carried out at ⫹ 800 mV, in order for the system to be more sensitive to changes of the enzymatic activity. All measurements were obtained in the flow injection system with flow rate 1.25 cm−3/min.
A typical recording and a calibration curve for biosensor containing 1.0% w/w DEAE-dextran are shown in Fig. 3 and Fig. 4 respectively. The response to glucose was linear from 0.5 to 12.5 mM, while the detection limit based on a signal-to-noise ratio of 3 was 0.07 mM. The reproducibility of the measurements was between 0.5 and 4.5% RSD (for six measurements) for all the range of concentrations examined. As it can be seen from Fig. 3 the response time is less than 30 s. The signal reproducibility of four different sensors was then examined. Comparing the response (at 10 mM glucose) and the sensitivity of different electrodes constructed from the same enzyme batch, stabilized with
Fig. 2. Hydrodynamic voltammogram of the glucose biosensor containing 1.0% w/w DEAE-dextran. The current refers to injections of 10 mM glucose response, in a 10 mM phosphate buffer.
1208
V.G. Gavalas et al. / Biosensors & Bioelectronics 13 (1998) 1205–1211
Fig. 3. Typical recording of the GOx ⫺ 1.0% w/w DEAE-dextran stabilized enzyme sensor to 1.0, 5.0, 10.0 mM glucose. The applied potential is 800 mV vs Ag/AgCl, the flow rate 1.25 cm3/min, and the injection volume 400 l.
Fig. 4. Calibration curve of the GOx-based sensors, without (䊊) DEAE-dextran, and with 1.0% w/w DEAE-dextran (䉬). The applied potential is 800 mV vs Ag/AgCl, the flow rate 1.25 cm3/min, and the injection volume 400 l.
1.0% w/w of DEAE-dextran, a relative standard deviation between 1.0 and 2.5% was obtained, a value that suggests a very good sensor reproducibility for fabrication. The effect of the amount of DEAE-dextran added to the enzyme solution was then evaluated, and the results are shown in Fig. 5. An increase in response to glucose with increasing DEAE-dextran amount in the range of 0–2.5% w/w, is observed. Unfortunately, the electrolyte increase is accompanied by an increase in the background current. The electrodes were occasionally tested, and then kept in buffer solution at 277 K between measurements. From Fig. 5 it is clear that the 1.0–1.5%
w/w DEAE-dextran loading gave the sensor with the best long-term stability, while the dramatic decrease in the stability of the sensor containing 2.5% w/w DEAEdextran is unexpected, and cannot be explained at this time. The remaining experiments for the operational and the storage stability of the GOx sensors were thus performed using biosensors containing 1.0% w/w DEAE-dextran. The operational stability of the glucose biosensor under continuous polarization and continuous flow of 5 mM glucose solution was examined, and the results are shown in Fig. 6. The experiment was carried out using a flow rate of 0.50 cm−3/min, resulted in an increase in
V.G. Gavalas et al. / Biosensors & Bioelectronics 13 (1998) 1205–1211
1209
Fig. 5. Effect of the amount of DEAE-dextran, [(+) 0%, (왓) 1.0%, (䊐) 1.5%, (왕) 2.0% and (䊊) 2.5%] to the long-term stability of the glucose biosensor. The electrodes were tested and then stored in buffer solution at 227 K. The current refers to injections of 10 mM glucose response, in a 10 mM phosphate buffer.
Fig. 6. Operational stability of the GOx-based, 1.0% w/w DEAE-dextran stabilized sensor. The sensor was used after a conditioning period of 4 h in buffer. The peaks correspond to injections of 5.0 and 10.0 mM glucose followed by continuous flow of 5.0 mM glucose solution at a flow rate of 0.5 cm−3/min.
the sensors response. The sensors are very stable, since no decrease in the response was observed after 30 h of measurement. The storage stability of the sensors was examined by freeze-drying a batch of freshly made glucose sensors. The dried sensors were kept at room temperature and humidity. After a predetermined period of time a sensor was re-hydrated for 20 h in phosphate buffer, its characteristics were evaluated, and the sensor was discarded. It was found that under these storage conditions the sensor retains its full activity even after 3 months.
3.2. Hydrogen peroxide biosensor The response scheme of the hydrogen peroxide biosensor can be described using the following three reactions (Gorton et al., 1992): H2O2 ⫹ HRP→H2O ⫹ compound ⫺ I where compound-I denotes the oxidized form of the enzyme. compound ⫺ I ⫹ e−→compound ⫺ II
1210
V.G. Gavalas et al. / Biosensors & Bioelectronics 13 (1998) 1205–1211
and compound ⫺ II ⫹ e−→HRP The second and the third reactions are a one-electron step reactions, which can occur with direct electron transfer between the electrode surface and the enzyme intermediates (compound-I, compound-II) (Cso¨regi, 1994). Based on the experience obtained from the GOxbased sensors, immobilization of HRP via physical adsorption into the porous active carbon for the direct measurement of hydrogen peroxide was performed. Initially, the effect of the type of the buffer, as well as the pH on the biosensor’s response of the sensors sensitivity in a batch mode at an applied potential of 0 mV vs Ag/AgCl, was examined. The buffers examined were 10 mM acetate and 10 mM phthalate in the pH range from 4.5 to 5.5. From these experiments, the potassium hydrogen phthalate (KHP) buffer at pH 5.0 was found to be the most appropriate. The concentration of the buffer was then evaluated (Fig. 7), and the optimum concentration was shown to be between 10 and 20 mM. The calibration curve was linear from 4.4 ⫻ 10−5 M to 3.1 ⫻ 10−4 M (1.5–10.5 ppm) for both buffer concentrations, while the detection limit based on a signal-to-noise ratio of 3 was 2.9 ⫻ 10−6 M (0.1 ppm). The reproducibility of the measurements was between 2.0 and 5.0% RSD (for six measurements) for all the range of concentrations examined, while the response time was in the order of 1 min. Subsequent experiments were carried out using 10 mM KHP buffer, pH 5.0 in the flow injection system with a flow rate 0.75 cm−3/min. The effect of the amount of DEAE-dextran added to
the enzyme solution is shown in Fig. 8. The addition of DEAE-dextran to the enzyme increased the long term stability of the sensor. The sensor containing 2.0% w/w DEAE-dextran exhibits initially smaller sensitivity to hydrogen peroxide than the other two containing 0.0% and 0.5% w/w DEAE-dextran, respectively. However, this sensor retains 77% of its activity after a period of 7 weeks, compared to 36% and 58% for the 0.0% and 0.5% w/w DEAE-dextran containing sensors, respectively. An increase in the sensitivity observed for the sensors containing 0.0% and 2.0% after the 40th day is unexpected, and cannot be explained at this time.
4. Conclusions
The combination of an activated conductive porous carbon rod with the diethylaminoethyl-dextran stabilized enzymes was used for the construction of highly stable glucose oxidase and horseradish peroxidase based sensors. The combined use of this novel active carbon and the polyelectrolyte stabilized enzymes result in the construction of biosensors with extended operational stability, good reproducibility and fast response times, offering a novel methodology for the construction of a wide variety of sensors based on enzymes stabilized with electrolytes. Further experiments are aiming in the use of a variety of negatively and positively charged electrolytes for the development of operationally stable biosensors employing less stable enzymes.
Fig. 7. Influence of the KHP concentration on the HRP-based sensor in (䊐) 1.0 mM, (䊊) 5.0 mM, (왕) 10.0 mM and (䉫) 20.0 mM KHP, pH: 5.0 at 0.0 mV vs Ag/AgCl.
V.G. Gavalas et al. / Biosensors & Bioelectronics 13 (1998) 1205–1211
1211
Fig. 8. Effect of the DEAE-dextran, to the long-term stability of the HRP sensor with (䊐) 0%, (䊊) 0.5% and (왕) 2.0% DEAE-dextran respectively. The electrodes were tested and then stored in buffer solution at 227 K. The y-axis represents the sensitivity of the sensor obtained from the calibration curve of the sensor to peroxide concentration from 4.4 ⫻ 10−5 to 3.1 ⫻ 10−4 M, in a 10 mM phthalate buffer.
References Appleton, A., Gibson, T.D., Woodward, J.R., 1997. High temperature stabilization of immobilized glucose oxidase: potential applications in biosensors. Sensors and Actuators B 43, 65–69. Cass, A.E.G. (Ed.), 1990. Biosensors. A Practical Approach. Oxford University Press, New York. Cso¨regi, E., 1994. Construction and miniaturization of peroxidase based amperometric biosensors. Doctoral thesis. Lund University. De La Guardia, M., 1995. Biochemical sensors: the state of art. Mikrochim. Acta 120, 243–255. Emr, S.A., Yacynych, A.M., 1995. Use of polymer films in amperometric biosensors. Electroanal. 7 (10), 913–923. Gorton, L., Jo¨nsson-Pettersson, G., Cso¨regi, E., Johansson, K., Dominguez, E., Marko-Varga, G., 1992. Amperometric biosensors based on an apparent direct electron transfer between electrodes and immobilized peroxidases. Analyst 117, 1235–1241.
Gorton, L., 1995. Carbon paste electrodes modified with enzymes, tissues, and cells. Electroanal. 7 (1), 23–45. Khan, G.F., Wernet, W., 1997. Design of enzyme electrodes for extended use and storage life. Anal. Chem. 69, 2682–2687. Lutz, M., Burestedt, E., Emne´us, J., Lide´n, H., Gobhadi, S., Gorton, L., Marko-Varga, G., 1995. Effect of different additives on a tyrosinase based carbon paste electrode. Anal. Chim. Acta 305, 8–17. Vamvakaki, M., Chaniotakis, N.A., 1996. Solid-contact ion-selective electrode with stable internal electrode. Anal. Chim. Acta 320, 53–61. Wang, J., Naser, N., 1994. Improved performance of carbon paste amperometric biosensors through the incorporation of fumed silica. Electroanal. 6, 571–575. Yabuki, S., Mizutani, F., Katsura, T., 1994. Choline-sensing electrode based on polyethylene glycol-modified enzyme and mediator. Sensors and Actuators B 20, 159–162.
Improved operational stability of biosensors based on enzymepolyelectrolyte complex adsorbed into a porous carbon electrode1 V. G. Gavalas a, N. A. Chaniotakis b
a,*
, T. D. Gibson
b
a Laboratory of Analytical Chemistry, Department of Chemistry, University of Crete, 71409 Iraklion, Crete, Greece Enzyme Biotechnology Group, Department of Biochemistry and Molecular Biology, Astbury Building, University of Leeds, Leeds, West Yorkshire LS2 9JT, UK
Received 19 February 1998; received in revised form ??/??/??; accepted 13 May 1998
Abstract A novel porous active carbon is utilized in order to adsorb the diethylaminoethyl-dextran (DEAE-dextran)-enzyme stabilized complexes, for the construction of highly stable biosensors. The interaction of DEAE-dextran with the examined enzymes increases dramatically the operational stabilization of the sensors, without adverse effects on the enzyme activity. At the same time, the porous active carbon allows for high enzyme loading, good electrical contact and low resistance throughout the sensing element. Glucose oxidase and horseradish peroxidase are used as model enzymes in this study to construct biosensors, with very good reproducibility (less than 5% RSD). As a result, the glucose sensor exhibits very long operational stability (over a period of 5 months), while the hydrogen peroxide sensor retains its initial activity after several weeks. 1998 Elsevier Science S.A. All rights reserved. Keywords: Porous carbon electrode; Polyelectrolyte; Operational stability; Glucose oxidase; Horseradish peroxidase
1. Introduction The performance of biosensors depends on the stability of the electrochemical transducer as well as on the retention of the activity of the enzymes employed. During the last few years, many efforts have been directed towards the stabilization of biosensors by careful optimization of these factors. The commonly used electrochemical transducers are either metal electrodes (De la Guardia, 1995; Khan and Wernet, 1997) or carbonaceous materials (Gorton, 1995; Cso¨regi, 1994). The widespread use of carbon paste as the supporting matrix for the construction of biosensors is attributed to the simple construction procedures required, the low background current they exhibit, their ability for surface regeneration, and its very low cost. The prime drawback of carbon paste electrodes is the difficulty in reproducing the composition of the paste, leading to irreproducible sensors. * Corresponding author. Tel.: ⫹ 30-81-393-137; Fax: ⫹ 30-81-210951; E-mail: [email protected] 1 This paper was presented at the Fifth World Congress on Biosensors, Berlin, Germany, 3–5 June 1998.
Simultaneously, carbon paste biosensors usually suffer from short lifetime, which is attributed to leaching of the mediator and/or the enzyme out of the paste, in conjunction to the denaturation of the enzyme within the highly lipophilic environment of the paste. It is thus evident that the immobilization of the enzyme in the sensor element is of crucial importance to the biosensors’ operational and storage lifetime. Various techniques have been applied in order to keep or supply adequate quantities of active enzyme onto the electrode surface. Physical adsorption, covalent attachment, crosslinking, and entrapment in polymer films (Cass, 1990; Emr and Yacynych, 1995) are some of the most commonly used ones. During the immobilization process, it is vital that the enzyme retains its activity and specificity. Recently, it has been shown that certain additives can dramatically increase the stability of the enzymes (Appleton et al., 1997; Lutz et al., 1995; Wang and Naser, 1994; Yabuki et al., 1994), even though this effect is not yet completely elucidated. It is believed though that this increase in the enzyme’s stability is achieved through electrostatic interactions that protect the enzyme, and aid it in retaining its active conformation during immobilization in hostile environments.
0956-5663/98/$ - see front matter 1998 Elsevier Science S.A. All rights reserved. PII: S 0 9 5 6 - 5 6 6 3 ( 9 8 ) 0 0 0 6 6 - 9
1206
V.G. Gavalas et al. / Biosensors & Bioelectronics 13 (1998) 1205–1211
In this paper, we report the use of a novel porous active carbon (Vamvakaki and Chaniotakis, 1996) for the immobilization of polyelectrolyte stabilized enzymes for the construction of biosensors with high operational and storage lifetimes. This is achieved with the physical adsorbed of the enzyme-polyelectrolyte complex into the electrode pores, without using chemical coupling agents, thus avoiding any unnecessary denaturation. The use of the polyelectrolyte diethylaminoethyl-dextran (DEAEdextran) as the enzyme stabilizing polyelectrolyte is used to illustrate the achieved stability of glucose and hydrogen peroxidase biosensors. The combination of the porous active carbon and the positively charged polyelectrolytes is a very promising method for the construction of a variety of biosensors with extended operational stability, good reproducibility and fast response times.
2. Experimental 2.1. Materials Glucose oxidase (GOx) from Aspergillus niger (EC 1.1.3.4, 200 U/mg) and peroxidase (HRP) from horse radish (EC 1.11.1.7, 731 U/mg) were purchased from Fluka. Diethylaminoethyl-dextran (DEAE-dextran) (D9885) was obtained from Sigma. D(+)-glucose, hydrogen peroxide (30%), potassium dihydrogen phosphate and potassium hydrogen phtalate were also purchased from Fluka. The dialysis membrane was a pre-mounted type ‘C’ cellulose acetate. The porous active carbon was obtained from KOH.I.NOOR HARDTMUTH (Austria). The density of the carbon was 1.16 ⫾ 0.03 g/cm−3 with porous volume of 0.336 ⫾ 0.003 cm−3/g. All other reagents used were of analytical grade. 2.2. Measurements and apparatus
2.3. Enzyme immobilization A rod (3.1 mm diameter, 5.0 mm height) of the porous active carbon was prepared by cleaning in a sonicated ethanol bath for 10 min and for 10 min more in water bath. The rod was then dried at 430 K for 30 min, and was ready for use. The enzyme was dissolved in the proper buffer solution to a final concentration of 2500 U/cm−3, and then the appropriate amount of polyelectrolyte was added. The enzyme was allowed to react with the polyelectrolyte for 20 min at room temperature. Finally, the carbon rod was placed into the enzyme solution, and allowed to adsorb enzyme solution for 20 h at room temperature. The concentration of the enzyme in the carbon matrix is the same as the initial enzyme concentration, as shown by absorbance measurements of the enzyme. Since the solution uptake of the carbon is 34% of its weight, the enzyme loading of the resulting electrode is controlled by the initial enzyme concentration in solution. The carbon rode was then removed from the solution, washed thoroughly with buffer solution, and placed in a Teflon holder for testing. The electrical contact was achieved through a platinum wire from the reverse side. During the immobilization of the GOx, phosphate buffer 10 mM, pH 7.5 was used, while a 10 mM, pH 5.0 potassium hydrogen phthalate was the buffer of choice for the HRP sensor. In the case of the glucose biosensor, a dialysis membrane was placed on the outer surface of the carbon rod.
3. Results and discussion 3.1. Glucose biosensor It is well known that GOx catalyzes the following reaction: D ⫺ glucose ⫹ O2→gluconic acid ⫹ H2O2
All experiments were carried out using a three electrode Metrohm 641 VA-Detector, a silver/silver chloride double junction reference electrode (Model 90-02, ORION Research) and a platinum counter electrode. Temperature control at 297.0 ⫾ 0.1 K was achieved with a circulating bath (Model 9105, PolyScience). The flow injection system consisted of a wall-jet flow cell, an injection valve with loop volume 400 l, while the solvent delivery was done using either a sage syringe pump (Model 362, ORION Research) or a peristaltic pump (No 7554-30, Cole-Parmer). The signal was recorded in a personal computer equipped with a 16 bit A/D converter and controlled with a software written in basic. In all experiments nano-pure water ( 苲 18 M⍀, EASYpure model D7033, Barnstead) was used.
while the transduction of the biochemical recognition is achieved via the following electrochemical redox reaction: H2O2→2H+ ⫹ O2 ⫹ 2e− Initially the effect of pH on the sensor’s response was evaluated using 10 mM phosphate buffers applying potential ⫹ 800 mV vs Ag/AgCl. The results obtained are illustrated in Fig. 1. As it can be seen, the biosensor’s response is dramatically increased by changing the pH from 6.5 to 7.5 and then levels off. The operating pH was chosen to be 7.5, even though from the figure the optimum pH seems to be above this value. The apparent increase in the response of the sensor for pH above 8.0
V.G. Gavalas et al. / Biosensors & Bioelectronics 13 (1998) 1205–1211
1207
Fig. 1. Influence of pH on the glucose biosensor. The current refers to injections of 10 mM glucose response. Applied potential ⫹ 800 mV vs Ag/AgCl, in a 10 mM phosphate buffer.
is due to the higher sensitivity of the carbon matrix to peroxide (8.0 ⫾ 0.2 A/mM at pH 7.5 and 11.5 ⫾ 0.4 A/mM at pH 8.5), and not due to the increase of the enzyme efficiency. Fig. 2 shows the hydrodynamic voltammogram obtain with potentials between 0 and ⫹ 800 mV for the glucose biosensor. The response to the enzymatically produced H2O2 increase linearly above ⫹ 500 mV, while even at ⫹ 400 mV the biosensor is analytically usable. Subsequent experiments were carried out at ⫹ 800 mV, in order for the system to be more sensitive to changes of the enzymatic activity. All measurements were obtained in the flow injection system with flow rate 1.25 cm−3/min.
A typical recording and a calibration curve for biosensor containing 1.0% w/w DEAE-dextran are shown in Fig. 3 and Fig. 4 respectively. The response to glucose was linear from 0.5 to 12.5 mM, while the detection limit based on a signal-to-noise ratio of 3 was 0.07 mM. The reproducibility of the measurements was between 0.5 and 4.5% RSD (for six measurements) for all the range of concentrations examined. As it can be seen from Fig. 3 the response time is less than 30 s. The signal reproducibility of four different sensors was then examined. Comparing the response (at 10 mM glucose) and the sensitivity of different electrodes constructed from the same enzyme batch, stabilized with
Fig. 2. Hydrodynamic voltammogram of the glucose biosensor containing 1.0% w/w DEAE-dextran. The current refers to injections of 10 mM glucose response, in a 10 mM phosphate buffer.
1208
V.G. Gavalas et al. / Biosensors & Bioelectronics 13 (1998) 1205–1211
Fig. 3. Typical recording of the GOx ⫺ 1.0% w/w DEAE-dextran stabilized enzyme sensor to 1.0, 5.0, 10.0 mM glucose. The applied potential is 800 mV vs Ag/AgCl, the flow rate 1.25 cm3/min, and the injection volume 400 l.
Fig. 4. Calibration curve of the GOx-based sensors, without (䊊) DEAE-dextran, and with 1.0% w/w DEAE-dextran (䉬). The applied potential is 800 mV vs Ag/AgCl, the flow rate 1.25 cm3/min, and the injection volume 400 l.
1.0% w/w of DEAE-dextran, a relative standard deviation between 1.0 and 2.5% was obtained, a value that suggests a very good sensor reproducibility for fabrication. The effect of the amount of DEAE-dextran added to the enzyme solution was then evaluated, and the results are shown in Fig. 5. An increase in response to glucose with increasing DEAE-dextran amount in the range of 0–2.5% w/w, is observed. Unfortunately, the electrolyte increase is accompanied by an increase in the background current. The electrodes were occasionally tested, and then kept in buffer solution at 277 K between measurements. From Fig. 5 it is clear that the 1.0–1.5%
w/w DEAE-dextran loading gave the sensor with the best long-term stability, while the dramatic decrease in the stability of the sensor containing 2.5% w/w DEAEdextran is unexpected, and cannot be explained at this time. The remaining experiments for the operational and the storage stability of the GOx sensors were thus performed using biosensors containing 1.0% w/w DEAE-dextran. The operational stability of the glucose biosensor under continuous polarization and continuous flow of 5 mM glucose solution was examined, and the results are shown in Fig. 6. The experiment was carried out using a flow rate of 0.50 cm−3/min, resulted in an increase in
V.G. Gavalas et al. / Biosensors & Bioelectronics 13 (1998) 1205–1211
1209
Fig. 5. Effect of the amount of DEAE-dextran, [(+) 0%, (왓) 1.0%, (䊐) 1.5%, (왕) 2.0% and (䊊) 2.5%] to the long-term stability of the glucose biosensor. The electrodes were tested and then stored in buffer solution at 227 K. The current refers to injections of 10 mM glucose response, in a 10 mM phosphate buffer.
Fig. 6. Operational stability of the GOx-based, 1.0% w/w DEAE-dextran stabilized sensor. The sensor was used after a conditioning period of 4 h in buffer. The peaks correspond to injections of 5.0 and 10.0 mM glucose followed by continuous flow of 5.0 mM glucose solution at a flow rate of 0.5 cm−3/min.
the sensors response. The sensors are very stable, since no decrease in the response was observed after 30 h of measurement. The storage stability of the sensors was examined by freeze-drying a batch of freshly made glucose sensors. The dried sensors were kept at room temperature and humidity. After a predetermined period of time a sensor was re-hydrated for 20 h in phosphate buffer, its characteristics were evaluated, and the sensor was discarded. It was found that under these storage conditions the sensor retains its full activity even after 3 months.
3.2. Hydrogen peroxide biosensor The response scheme of the hydrogen peroxide biosensor can be described using the following three reactions (Gorton et al., 1992): H2O2 ⫹ HRP→H2O ⫹ compound ⫺ I where compound-I denotes the oxidized form of the enzyme. compound ⫺ I ⫹ e−→compound ⫺ II
1210
V.G. Gavalas et al. / Biosensors & Bioelectronics 13 (1998) 1205–1211
and compound ⫺ II ⫹ e−→HRP The second and the third reactions are a one-electron step reactions, which can occur with direct electron transfer between the electrode surface and the enzyme intermediates (compound-I, compound-II) (Cso¨regi, 1994). Based on the experience obtained from the GOxbased sensors, immobilization of HRP via physical adsorption into the porous active carbon for the direct measurement of hydrogen peroxide was performed. Initially, the effect of the type of the buffer, as well as the pH on the biosensor’s response of the sensors sensitivity in a batch mode at an applied potential of 0 mV vs Ag/AgCl, was examined. The buffers examined were 10 mM acetate and 10 mM phthalate in the pH range from 4.5 to 5.5. From these experiments, the potassium hydrogen phthalate (KHP) buffer at pH 5.0 was found to be the most appropriate. The concentration of the buffer was then evaluated (Fig. 7), and the optimum concentration was shown to be between 10 and 20 mM. The calibration curve was linear from 4.4 ⫻ 10−5 M to 3.1 ⫻ 10−4 M (1.5–10.5 ppm) for both buffer concentrations, while the detection limit based on a signal-to-noise ratio of 3 was 2.9 ⫻ 10−6 M (0.1 ppm). The reproducibility of the measurements was between 2.0 and 5.0% RSD (for six measurements) for all the range of concentrations examined, while the response time was in the order of 1 min. Subsequent experiments were carried out using 10 mM KHP buffer, pH 5.0 in the flow injection system with a flow rate 0.75 cm−3/min. The effect of the amount of DEAE-dextran added to
the enzyme solution is shown in Fig. 8. The addition of DEAE-dextran to the enzyme increased the long term stability of the sensor. The sensor containing 2.0% w/w DEAE-dextran exhibits initially smaller sensitivity to hydrogen peroxide than the other two containing 0.0% and 0.5% w/w DEAE-dextran, respectively. However, this sensor retains 77% of its activity after a period of 7 weeks, compared to 36% and 58% for the 0.0% and 0.5% w/w DEAE-dextran containing sensors, respectively. An increase in the sensitivity observed for the sensors containing 0.0% and 2.0% after the 40th day is unexpected, and cannot be explained at this time.
4. Conclusions
The combination of an activated conductive porous carbon rod with the diethylaminoethyl-dextran stabilized enzymes was used for the construction of highly stable glucose oxidase and horseradish peroxidase based sensors. The combined use of this novel active carbon and the polyelectrolyte stabilized enzymes result in the construction of biosensors with extended operational stability, good reproducibility and fast response times, offering a novel methodology for the construction of a wide variety of sensors based on enzymes stabilized with electrolytes. Further experiments are aiming in the use of a variety of negatively and positively charged electrolytes for the development of operationally stable biosensors employing less stable enzymes.
Fig. 7. Influence of the KHP concentration on the HRP-based sensor in (䊐) 1.0 mM, (䊊) 5.0 mM, (왕) 10.0 mM and (䉫) 20.0 mM KHP, pH: 5.0 at 0.0 mV vs Ag/AgCl.
V.G. Gavalas et al. / Biosensors & Bioelectronics 13 (1998) 1205–1211
1211
Fig. 8. Effect of the DEAE-dextran, to the long-term stability of the HRP sensor with (䊐) 0%, (䊊) 0.5% and (왕) 2.0% DEAE-dextran respectively. The electrodes were tested and then stored in buffer solution at 227 K. The y-axis represents the sensitivity of the sensor obtained from the calibration curve of the sensor to peroxide concentration from 4.4 ⫻ 10−5 to 3.1 ⫻ 10−4 M, in a 10 mM phthalate buffer.
References Appleton, A., Gibson, T.D., Woodward, J.R., 1997. High temperature stabilization of immobilized glucose oxidase: potential applications in biosensors. Sensors and Actuators B 43, 65–69. Cass, A.E.G. (Ed.), 1990. Biosensors. A Practical Approach. Oxford University Press, New York. Cso¨regi, E., 1994. Construction and miniaturization of peroxidase based amperometric biosensors. Doctoral thesis. Lund University. De La Guardia, M., 1995. Biochemical sensors: the state of art. Mikrochim. Acta 120, 243–255. Emr, S.A., Yacynych, A.M., 1995. Use of polymer films in amperometric biosensors. Electroanal. 7 (10), 913–923. Gorton, L., Jo¨nsson-Pettersson, G., Cso¨regi, E., Johansson, K., Dominguez, E., Marko-Varga, G., 1992. Amperometric biosensors based on an apparent direct electron transfer between electrodes and immobilized peroxidases. Analyst 117, 1235–1241.
Gorton, L., 1995. Carbon paste electrodes modified with enzymes, tissues, and cells. Electroanal. 7 (1), 23–45. Khan, G.F., Wernet, W., 1997. Design of enzyme electrodes for extended use and storage life. Anal. Chem. 69, 2682–2687. Lutz, M., Burestedt, E., Emne´us, J., Lide´n, H., Gobhadi, S., Gorton, L., Marko-Varga, G., 1995. Effect of different additives on a tyrosinase based carbon paste electrode. Anal. Chim. Acta 305, 8–17. Vamvakaki, M., Chaniotakis, N.A., 1996. Solid-contact ion-selective electrode with stable internal electrode. Anal. Chim. Acta 320, 53–61. Wang, J., Naser, N., 1994. Improved performance of carbon paste amperometric biosensors through the incorporation of fumed silica. Electroanal. 6, 571–575. Yabuki, S., Mizutani, F., Katsura, T., 1994. Choline-sensing electrode based on polyethylene glycol-modified enzyme and mediator. Sensors and Actuators B 20, 159–162.