Anal. Chem. 1994,66, 290-294
Substrate-Purging Enzyme Electrodes. Peroxidase/ Catalase Electrodes for H202with an Improved Upper Sensing Limit Tetsu Tatsuma,'*t Tsuyoshi Watanabe, Sayuri Tatsuma, and Tadashl Watanabe' Institute of Industrial Science, University of Tokyo, Roppongi, Minato-ku, Tokyo 106, Japan
A substratepurging enzyme electrode carrying a sensing enzyme and a substrate-purging catalyst on its surface is proposed and was examined theoretically and experimentally. Since the substrate-purging catalyst partially consumes the substrate without consuming a cosubstrate, the sensitivity of the enzyme electrode is lowered while the maximum response remains unchanged leading to a raising of the upper sensing limit. Thus, by coating it with a cross-linked film of catalase and albumin, the upper sensing limit of a peroxidaseincorporatedpolypyrrole membrane electrodeas an H203 sensor was raised by -2 orders of magnitude. Enzyme electrodes serve as sensitive sensors for biologically important analytes.' Enzyme electrodes satisfy, at the same time, sensitivity, selectivity, and operational simplicity. Repetitious usage of enzyme is an additional advantage of enzyme electrodes. These electrodes, however, have an upper sensing limit ingeneral. In a recent paper, we fabricated a horseradish peroxidase-incorporated polypyrrole (HRP/PPy) membrane electrode as a reagentless sensor for H202.2 This electrode senses H202 to M without any mediator in the solution. This lower sensing limit is the best among electrochemical H202 sensors, to the best of our knowledge. The reactions of peroxidase in the HRP/PPy electrode are represented as follows:
-
+ H20, H R P compound I + H,O H R P compound I + e- + H+ H R P compound I1 H R P compound I1 + e- + H+ ferric H R P + H 2 0 ferric H R P
-
(1) (2) (3)
where compounds I and I1 are oxygen complexes of HRP. Compounds I and I1 are reduced by accepting electrons from PPy, yielding a reduction current response. At lower H202 concentrations, reaction 1 determines the overall reaction rate so that the response is proportional to the H202 concentration. At higher H202 concentrations, however, reaction 1 no longer determines the overall reaction rate but reaction 2 or 3 determines it, and hence, the response is saturated. The response in the saturated region (the maximum response) depends on the proton concentration so that the upper sensing limit of H202 can be controlled by changing P H , though ~ this is not of practical value because a change in pH will change
'
Present address: Department of Applied Chemistry, Faculty of Technology, Tokyo University of Agriculture and Technology, Koganei, Tokyo 184, Japan. (1) Scheller, F.; Schubert, F. Biosensors; Elsevier: Amsterdam, 1992. (2) Tatsuma, T.; Gondaira, M.; Watanabe, T. Anal. Chem. 1992,64,1183-1187.
290
Analytical Chemistry, Vol. 66, No. 2, January 75, 1994
the properties of H R P and PPy. When the response is controlled by mass transfer, use of a membrane that is more permeable to protons than to H202 will raise the upper limit. Lowering the H202 concentration in the vicinity of the enzyme electrode while retaining the proton concentration is an additional means of raising the upper limit. To this end, we introduce a catalyst that partially decomposes H202 without consuming protons onto the enzyme electrode. We call such a catalyst a substrate-purging catalyst, and such an electrode a substrate-purging enzyme electrode. In this configuration the lower sensing limit will be raised together with the upper limit, because the noise level will not change. Here we employ catalase (CAT) as an HzO2-purging catalyst for the HRP/PPy electrodes. The reaction of catalase proceeds as follows:
--
+H20 ferric CAT + 0, + H,O
ferric CAT + H202 CAT compound I CAT compound I
+ H202
(4)
(5)
CAT consumes H202 without consuming a proton CAT may exhibit peroxidative activity (reactions 1-3) so that thecatalase layer, which contains no electron donor, is separated from the HRP/PPy layer to avoid the electron transfer from PPy to CAT as shown in Figure 1A. The HRP/PPy layer as an H202-sensing layer is electrochemically deposited on a tin oxide electrode followed by deposition of a cross-linked CAT/ bovine serum albumin (BSA) membrane as a substratepurging layer. To avoid significant changes in characteristics of the film such as film thickness and permeability for the substrate, the CAT content in the CAT/BSA film is limited to 20% or less. In what follows, we propose the concept and explain the mechanism of the substrate-purging enzyme electrode via theoretical evaluation of the steady-state responses. Further, we demonstrate experimentally the effectivenessof the concept proposed for a bienzyme electrode with H R P and CAT. The response of the CAT/HRP electrodes will be analyzed in terms of the theoretical response. THEORY First we discuss the mechanism and response of the substrate-purging enzyme electrode theoretically. Figure 1A schematically illustrates a substrate-purging enzyme electrode with a bilayer configuration. The underlayer is the sensing layer containing a "sensing enzyme". The sensing enzyme reacts in a ping-pong mechanism: E +S,
ki
E'
+ PI
(6)
0003-2700/94/036&0290$04.50/0
0 1994 American Chemical Society
as follows: Layer
k3
c +s, - C + P 3
Layer
Distance
Distance
Flguro 1. (A) Schematic configuration of a substrate-purging enzyme electrode consisting of a sensing layer and a substrate-purging layer. (B, C)Substrate concentration profiles for the substratepurgingenzyme electrode under diffusion-controlled (B) and enzymatic reactioncontrolled (C)conditions.
E’
+ S,
-+ ki
E
P2
(7)
where E, S, and P are the sensing enzyme, substrate, and product, respectively. Substrates, is the analyteof the enzyme electrode and SZis the cosubstrate. The analyte SImay be sensed on the basis of electrochemical (amperometric or potentiometric) detection of an increase in P1 or PZor a decrease in S2. Direct or indirect charge transfer from E’ to the base electrode may yield an amperometric response. When the S1 concentration in the vicinity of the enzyme is so low that reaction 1 determines the enzymatic reaction rate, the response of the enzyme electrode is proportional to the S1 concentration. We refer to such a concentration region as a “linear region”3 hereafter. On the other hand, when the S1 concentration is so high that reaction 2 determines the reaction rate, the response no longer depends on the S1 concentration. Such a concentration region is called a “saturated r e g i ~ n ” . The ~ response in the saturated region is the maximum response. Thus, the upper sensing limit is determined by the sensitivity in the linear region and the maximum current. In this paper, the upper sensing limit CSLis defined as follows:
where Rma, is the maximum response and s is the sensitivity in the linear region. The upper layer, “substrate-purging layer”, contains a catalyst consuming the substrate S1 without consuming the cosubstrate S2. The reaction of the catalyst C can be written ( 3 ) Tatsuma, T . ; Watanabc, T . A n a l . Chem. 1992, 64, 625-630.
(9)
The concentration of substrate SIin the sensing layer is thus lowered by the substrate-purging layer, while the Sz concentration is not changed. Because of this, the sensitivity in the linear region is lowered without lowering the maximum response so that the upper sensing limit is raised (see eq 8). For practical enzyme electrodes, the noise level may not be changed by the introduction of the catalyst so that the lower sensing limit determined by the noise level and the sensitivity will be shifted together with the upper limit. In addition, it must be noted here that the apparent reaction rate of the substrate-purging catalyst in the film must be rectilinear with the substrate concentration up to a concentration expected as an upper limit. Here we formulate the response of substrate-purging enzyme electrodes for two cases, namely, the enzymatic reaction-controlled and diffusion-controlled cases. It is assumed that a change in the concentration of the substratepurging catalyst causes no changes in the diffusion coefficient of SZin a film and in the film thickness. Diffusion-ControlledCase. The concentration profile for S1 at the steady state under the diffusion-controlled condition is schematically illustrated in Figure 1B. In this case, the total activity of the sensing enzyme in the sensing layer is so high that the SI concentration at the interface of the sensing layer and the substrate-purging layer is negligible before the bulk concentration. A differential equation for the mass balance in the substrate-purging layer is as follows:
where Ds, CS, and CC are the diffusion coefficient of S1, concentration of SI, and concentration of the substrate-purging catalyst C , respectively, and k3 is the reaction rate constant for the catalytic reaction 9. The boundary conditions are Cs(x=0) = 0
(11)
Cs(x=d) = KsCs*
(12)
where x is the distance from the sensing layer/substratepurging layer interface and d is the thickness of the substratepurging layer. Ks and Cs* are the partition coefficient of S1 for the substrate-purging layer and the bulk concentration of SI,respectively. In the linear region, the enzymatic reaction rate is controlled by SIso that the supply of SZis not limited by diffusion3 and that the SZconcentration in both layers is not polarized. Equations 10-12 yield Cs as a function of distance x:
where the constant a is given by a = (k3Cc/Ds)1/z
(14) The sensitivity under the diffusion-controlled condition is proportional to the gradient of CSat x = 0, hence the ratio of the sensitivity in the linear region, s, to that for the case Ana!yticaI Chemistry, Vol. 66,No. 2, Janwty 15, 1994
291
where CC = 0, SO, is thus given by s/so =
[dCs/dx] (x=O) -- 2ad [dCs(Cc=0)/dx](x=O) ead - e-ad
11
I
(15)
Since the upper limit of the measurable concentration range CSLis defined by eq 8, the ratio of CSLto that for the case where CC = 0, CSLO,is
Enzymatic Reaction-Controlled Case. Figure 1C schematically illustrates a steady-state concentration profile for S1 in the vicinity of the electrode surface under the enzymatic reaction-controlled condition. In this case, the total activity of the sensing enzyme in the sensing layer is so low that the distribution of SI concentration in the sensing layer is not polarized. A differential equation for the mass balance in the substrate-purging layer is given by eq 10 again. The boundary conditions are
Cs(x=d) = KsCs*
(18)
Equations 10, 17, and 18 yield the following equation correlating CS and x:
where the constant a was given by eq 14. The response in this case is proportional to the SI concentration a t x = 0 so that the relative sensitivity S/SO is
Thus, the relative upper limit of the measurable concentration range is given by
EXPERIMENTAL SECTION Materials. Sn02- (9000 A thick, F-doped) coated glass plates as the base electrode were obtained from Nippon Sheet Glass. HRP (Grade 11) from Boehringer, CAT (from bovine liver) from Sigma, and BSA from Wako were used without further purification. Glutaraldehyde was purchased from Sigma. Electrode Preparation. An SnO2-coated glass plate (1 .O cm2) was pretreated with sulfuric acid that had been heated by dilution (1:l) for a few minutes. HRP/PPy electrodes were prepared by electropolymerization in an electrolyte solution ( 5 mL) containing 0.06 M KC1,0.05 M pyrrole, and 0.6 g L-I H R P a t room temperaturee2 Reference and counter electrodes were Ag/AgCl/KC1(0.06 M) and platinum black, respectively. Electropolymerization was done in a galvanostatic mode (current density, 0.1 mA cm-2), which yielded films with more uniform thickness than in a potentiostatic mode. The electrode potential was +800 to +900 mV during polymerization. An HRP/PPy electrode thus obtained was 292
Analyficel Chemistry, Vol. 66, No. 2, January 15, 1994
3 2
1
0
1
2
Log(Relattve Concentration)
Flgure 2. Theoretical calibration curves for a substratapurgingenzyme under a diffusion-controlled condition. The response and the substrate concentration are Indicated as relative values. Parameters: (a@ = 0 (a), 1 (b), 2 (c), 5 (d), or 10 (e); d = 10-3 cm; Ks = 1.
coated with a cross-linked CAT/BSA film by casting a 1O-kL aliquot of a citrate buffer (0.1 M, p H 5.9) containing CAT ( x g L-')/BSA (30 - x g L-I), and glutaraldehyde (ca. 10%). Measurement of the Sensor Performance. Electrochemical measurements were performed in a 0.1 M citrate-buffered solution (pH 5.9) a t 30 "C. Reference and counter electrodes were Ag/AgCl/KCl(satd) and platinum black, respectively. The electrode potential was +I70 mV vs Ag/AgCI. The solution was kept stirred, unless otherwise noted. Calibration curves were obtained by stepwise injections of an H202 standard solution with Eppendorf pipets.
RESULTS AND DISCUSSION TheoreticalResults for the Diffusion-ControlledCase. The responses of the substrate-purging enzyme electrode with the bilayer configuration under the diffusion-controlled condition were simulated on the basis of equations derived above, as shown in Figure 2. The diffusion coefficient of S2 and the film thickness d are assumed to be independent of CC here, otherwise the maximum response changes. As mentioned above, [dCs/dx](x=O) is proportional to the response. The value of (ad)*,which is proportional to CC,is changed stepwise from 0 to 10. Film thickness d = cm and KS = 1. The sensitivity in the linear region decreases with increasing ad because of the decreasing concentration gradient of SI a t x = 0. On the other hand, the maximumresponse is independent of cud because the rate of reaction 7, which determines the maximum response, is not affected by the reaction of the substrate-purging catalyst. The upper sensing limit is thus shifted toward higher concentrations by the presence of the substrate-purging catalyst. It is seen that the effect of the substrate-purging catalyst is negligible for (ad)Z< 1 under the present conditions. Thus, the concentration of the substrate-purging catalyst in the film (CC), the activity of the catalyst (k3), the diffusion coefficient of S1 in the substratepurging film (&), and the film thickness (d)should be selected or controlled to obtain the expected measurable concentration region.
I
a b c
e
d
I
4 -
8
8
0
8
3 -
8
e
e
4
0
0
0
0 0
A
0
A 0
0
2 0
I
I
I
I
I
-4
-3
-2
-1
-Y
-2
-1
0
1
2
-6
-5
Log(Re1ative Concentration)
Flgure9. Theoreticalcalibrationcurves for a substratepurgingenzyme under an enzymatic reaction-controiled condition. The response and the substrate concentration are indicated as relative values. Parameters: = 0 (a), 1 (b), 2 (c), 5 (d), or 10 (e); Ks = 1.
(cue
Theoretical Results for the Enzymatic Reaction-Controlled Case. Figure 3 depicts the simulated calibration curves under the enzymatic reaction-controlled condition. As mentioned above, Cs(x=O) is proportional to the response. The value of (ad)2is changed stepwise from 0 to 10 and Ks = 1. It is seen that the effect of the substrate-purging catalyst is more pronounced here than under the diffusion-controlled conditions. In the present case, the response at CC = 0 was assumed to be controlled by the reaction of the sensing enzyme. In practice, however, a change in CCmight change the diffusion coefficient of SZand/or the film thickness d, and a decrease in the diffusion coefficient and increase in d would lead the system to a mass-transfer-controlled condition. It should be noted that this may result in a change of the maximum response. Response Properties of the HRP/PpY Electrodes. An HRP/PPy film electrode coated with a cross-linked BSA film exhibited a cathodic current response to Hz02. The cathodic current was generated on the basis of positive charge transfer from H202 to PPy via H R P (eqs 1-3) as was the HRP/PPy electrode without the BSA film.2 Since the amount of H R P on the electrode surface depends on the thickness of the HRP/PPy film, the response of electrodes with thinner HRP/PPy films is expected to be limited by the enzymatic reaction and that of electrodes with thicker films to be limited by the mass transfer. Thus, the response-H2Oz concentration profiles were examined at various film thicknesses, which can be controlled by the electropolymerization charge. Figure 4 shows the experimental response+oncentration profiles for the HRP/PPy electrodes with an overlying BSA film (10 pL cast). These results are similar to those obtained for the HRP/PPy electrodes without a BSA film.z The response of the electrode increased with the polymerizationcharge until it was saturated at 5 mC cm-z. It is thus clear that the response is controlled by the enzymatic reaction at 1 2 mC cm-z while controlled by mass transfer at 25 mC cm-2.
Log([H,O,I/M)
Flgure 4. Calibration curves for the HRP/PPy electrodes coated with a cross-linked BSA film, In a 0.1 M cltrate-buffered solution (pH 5.9) at 30 OC: electrode potential, +170 mV vs AgIAgCI; PPy electropcb lymerization charge, 2 (O),5 (A),and 10 ( 0 )mC Cm-*.
0 l -6
-5
-4
I
-3
-2
A -1
Log([H,O,I/M)
Flgure 5. Calibration curves for the CAT/BSA-HRP/PPy electrodes ina 0.1 M cltrate-buffered solutkn(pH 5.9)at 30 O C : electrodepotential, +170 mV vs Ag/AgCl; PPy electropolymerlzatkm charge, 5 mC cm-*; content of CAT (CAT/total protein), 0 (O),2 (0),5 (A),10 (A),and 20 (e)wt % (total, 0.3 mg cm-*).
Experimental Results for the Diffusion-Controlled Case. The responsesof the CAT/BSA-HRP/PPy bilayer electrodes were examined under the diffusion-controlled condition. The polymerization charge for the HRP/PPy film was 5 mC cm-z. Figure 5 shows the dependencies of the cathodic current increase on the HzOz concentration for various CAT contents (weight percent of CAT in total protein) in the HzOz-purging layer. To avoid significant changes in the film thickness and diffusion coefficients of the substrates in the film, the CAT content in the film is limited to 20% or less. As expected, the sensitivity of the electrode in the linear region decreased with an increase in thecontent of CAT, while the maximum response was decreased to a much lesser extent. Thus, the upper sensing limit was improved by 1-1.5 orders. The observed decrease Analytical Chemistry, Vol. 66, No. 2, January 15, 1994
293
in the maximum current at 210% CAT may be caused by an increased film thickness and/or decreased diffusion coefficient of proton. The effect of the CAT content on the improvement of the upper limit is more significant at lower CAT contents and almost negligible at extremely higher contents. This tendency is, however, not in line with the theoretical expectation described above (Figure 2). Although this could be explained in terms of the increased diffusion coefficient of H202 in the film at high CAT contents, it may not be likely because the diffusion coefficient of proton does not seem to change that much. For the moment we speculate that this is caused by an 02-mediated response to H202. Molecular oxygen is generated as a result of the H202 disproportionation reaction catalyzed by CAT (reactions 4 and 5). The 0 2 generated can be reduced at PPy-coated S n 0 2 electrodes at +170 mV vs Ag/AgCl, though thereaction rate may not be so fast. Indeed, base cathodic currents observed at the present electrodes under air were higher than those under nitrogen by 20-60 nA cm-2. In view of this, we examined the response of CAT/BSA filmcarrying electrodes without H R P to H202. A PPy film containing BSA (BSA/PPy) was deposited on an Sn02 electrode by electropolymerization (5 mC cm-2) in a manner similar to that for the HRP/PPy films, and a BSA/PPy electrode thus obtained was coated with a CAT/BSA film (10%in CAT content). This CAT/BSA-BSA/PPy electrode responded to H202, though the sensitivity was 1 order of magnitude lower than that of the CAT/BSA-HRP/PPy electrode. A larger permeability of the HRP/PPy film than the BSA/PPy film by -1 order of magnitude would be required to explain the difference in the sensitivity. Experimental Results for the Enzymatic Reaction-Controlled Case. The responses of the CAT/BSA-HRP/PPy bilayer electrodes were also examined under the enzymatic reaction-controlled condition. The polymerization charge for the HRP/PPy film was 2 mC cm-2. Figure 6 shows the dependencies of the cathodic current increase on the H202 concentration for various CAT contents in the H202-purging layer. The improvement of the upper sensing limit was again verified. However, addition of CAT was less effective in the present case than in the diffusion-controlled case, contrary to the theoretical expectation. This may also be caused by the 02-mediated response. The difference in the sensitivity between the CAT/BSA-BSA/PPy electrode prepared with a polymerization charge of 5 mC cm-2 and that prepared with 2 mC cm-2 was smaller than the difference in the sensitivity between the cross-linked BSA film-coated HRP/PPy electrode
-
(4) Tatsuma, T.; Okawa, Y . ; Watanabe, T. Anal. Chem. 1989, 61, 2352-2355.
294
Analytical Chemistry, Vol. 66, No. 2, January 15, 1994
5------7 4
I
1
-6
-5
-4
-3
-2
-1
Log([H,O,IIM) Figure 8. Calibration curves for the CAT/BSA-HRP/PPy electrodes in a 0.1 M citrate-bufferedsolution (pH 5.9)at 30 OC: electrode potential, +170 mV vs Ag/AgCl; PPy electropolymerizatloncharge, 2 mC cm-*; content of CAT (CATItotal protein), 0 (O),2 (0),5 (A),10 (A),and 20 (e)w i % (total, 0.3 mg cm-2).
prepared with 5 mC cm-2 and that prepared with 2 mC cm-2. Therefore, the response originating in the supposed 0 2 mediation is more significant under the enzymatic reactioncontrolled condition ( 2 mC cm-2). Perspectives and Limitations. The concept of a substratepurging enzyme electrode for improving the upper sensing limit is thus demonstrated and its effectiveness is supported theoretically and experimentally. CAT as a substrate-purging catalyst can be used not only in the present HRP/PPy electrode but in mediated peroxidase electrode^.^ Though the present method can be applied in principle to other enzyme electrodes, it has some limitations. The most crucial problem is the availability of a substrate-purging catalyst suited to a given system. As mentioned above, the reaction rate of the catalyst in the film must be rectilinear with the substrate concentration up to a concentration desired as an upper limit. Further, if the catalyst needs its own cosubstrate, reactions in the system may be complicated. However, since a high selectivity is not always a prerequisite for the substrate-purging catalyst in practical systems, artificial catalyst can be used in place of a biological enzyme in those cases. Received for review June 21, 1993. 1993."
Accepted November 5,
Abstract published i n Advance ACS Abstracts, December 15, 1993.
