Analytical and Bioanalytical Chemistry

, Volume 382, Issue 2, pp 297–302

Amperometric detection in TMB/HRP-based assays


  • Pablo Fanjul-Bolado
    • Departamento de Química Física y AnalíticaUniversidad de Oviedo
  • María Begoña González-García
    • Departamento de Química Física y AnalíticaUniversidad de Oviedo
    • Departamento de Química Física y AnalíticaUniversidad de Oviedo
Special Issue Paper

DOI: 10.1007/s00216-005-3084-9

Cite this article as:
Fanjul-Bolado, P., González-García, M.B. & Costa-García, A. Anal Bioanal Chem (2005) 382: 297. doi:10.1007/s00216-005-3084-9


3,3′,5,5′-Tetramethylbenzidine (TMB) is the most commonly used chromogen for horseradish peroxidase (HRP) and so its performance as an electrochemical substrate was evaluated. Measurements of HRP activity in solution were carried out by using an amperometric detector coupled to a flow injection analysis (FIA) system. The enzymatic product was easily detected at a potential of +0.1 V (vs. Ag-pseudoreference electrode) at a bare screen-printed electrode placed in a homemade electrochemical flow cell. A high flow rate (4.3 mL min−1) of 0.5 M H2SO4 was used to obtain repeatable signals and a short analysis time. The detection limit achieved after 15 min of incubation was 2×10−14 M of HRP. The applicability of the amperometric detector to ELISAs was demonstrated by using a commercially available kit for the quantification of interleukin-6 (IL-6) without modifying the kit manufacturer’s protocol or the reagents for this test.


Horseradish peroxidase3,3′,5,5′-TetramethylbenzidineScreen-printed carbon electrodeFlow injection analysisELISA


3,3′,5,5′-Tetramethylbenzidine (TMB) was introduced by Bos et al. in 1981 as a chromogenic substrate for horseradish peroxidase (HRP)-based detection systems [1]. TMB is neither mutagenic nor carcinogenic [1] and is more sensitive than traditional HRP substrates like O-phenylenediamine (OPD) and 2,2′-azino-bis-(3-ethylbenzthiazoline-6-sulfonic acid) (ABTS) [2]. TMB presented initial problems related to its stability and poor solubility in aqueous buffer solutions; however, several formulations have been proposed to enhance the solubility and to increase the sensitivity of this substrate [35]. Therefore, nowadays TMB is the most used chromogenic substrate in HRP-based commercially available ELISA test kits.

Oxidation of aromatic amines by peroxidase has been studied for many years [6]. Thus, the mechanism of oxidation of TMB by HRP is a well-known process described by Josephy et al. [7]. Oxidation of TMB by HRP/H2O2 first generates a blue-colored complex product, which turns yellow after the addition of sulfuric acid to the reaction media. This yellow product has been identified as a two-electron oxidation product (diimine), which is stable in acid solutions. It has a maximum absorbance peak at 450 nm and it is also electroactive thus allowing an electrochemical detection.

There have been a great number of works about the electrochemical behavior of aromatic amines [810] and benzidine and its derivatives [11, 12]. Moreover, taking into account that TMB is the most sensitive chromogen for HRP, its use as an electrochemical substrate has already been widely exploited [1321]. An electrochemical approach for the detection of Helicobacter Pylori specific IgG antibody has been reported [14]. It is a differential pulse voltammetric enzyme-linked immunoassay in which a small three-electrode system was directly inserted in the well for the voltammetric detection. Another three-electrode configuration to fit to the 96-well microplate format has been employed as an amperometric detector for ELISAs [15]. However, these devices do not automate the detection system and demand a high increase in the analysis time due to the necessary cleaning step before dipping the electrodes into each new well. These disadvantages can be solved with the use of an FIA system in which the flow carrier is continuously cleaning the surface of the electrode. Moreover, an FIA system acts as an automatable linkage between the batch immunoassay and the detector, which is a great advantage compared to previous protocols. This approach has been optimized [16] and applied by Palleschi et al. [17, 18] who used an electrochemical thin-layer flow cell for liquid chromatography including a glassy carbon disk as working electrode. However, this kind of cell is expensive and suffers from fouling of the working electrode when continuously used. Its area and its sensitivity decrease with working, and a tedious procedure for polishing and renewing the electrode surface is required.

With the aim to avoid the above mentioned problems, we proposed an FIA system coupled with a simple homemade electrochemical flow cell, containing a screen-printed carbon electrode (SPCE). Versatile, low-cost, and mass-produced SPCEs are inserted into our electrochemical cell and they can be easily replaced when required. In this paper the feasibility of using our amperometric detector (in a flow system) of TMB (Ox) for ELISAs was demonstrated by applying it to the determination of interleukin 6 (IL-6). Interleukin 6 is a multifunctional protein produced by lymphoid and non-lymphoid cells, and by normal and transformed cells, including fibroblasts, hepatocytes, myelomas, astrogliomas, and glioblastomas. Interleukin 6 can be grouped in a family of cytokines that have growth factor activities and stimulate the growth of myeloma/hybridoma/plasmacytoma cells. The various activities of IL-6 suggest that this factor has a major role in the mediation of the inflammatory and immune responses initiated by infection or injury. Elevated IL-6 levels have been reported to be associated with a variety of diseases including autoimmune diseases such as arthritis and Castleman’s disease, psoriasis, myelomas, lymphomas, leukemias, and ovarian cancers. Some reviews on IL-6 have been reported [2224].


Reagents and solutions

Horseradish peroxidase Type VI-A (EC, specific activity 987 U mg−1) and bovine serum albumin (fraction V, BSA) were purchased from Sigma Chemical Company (Spain). HRP stock solution (1 mg mL−1) was made in H2O and was stored at 4°C. The blocking of the microtiter wells was performed by using a 10 mM saturated (8 g L−1 of NaCl (Merck), 2 g L−1 of KCl (Merck)) phosphate buffer solution (PBS) pH 7.4 with 3% of BSA. R&D Systems Inc. provided stabilized TMB and H2O2 stock solutions and the IL-6 ELISA test kit. This comprised a microtiter plate coated with anti-IL 6 monoclonal antibody, polyclonal anti-IL 6 HRP conjugate, and a washing solution. IL-6 standard (1.5 ng) was supplied lyophilized; reconstitution and dilution of this standard were made with an animal serum termed diluent RD6F. Sulfuric acid (95–97%), ortho-phosphoric acid (85%), and acetic acid glacial (100%) were provided by Merck. All other chemicals employed were of analytical reagent grade. Ultrapure water obtained with a Milli-R 3 plus/Milli Q plus 185 purification system from Millipore Ibérica S.A (Spain) was used throughout this work.

Apparatus and materials

Staircase cyclic voltammetry and amperometric measurements were performed with an Autolab PGSTAT 12 (Eco Chimie B.V., The Netherlands) potentiostat interfaced to an AMD K-6 266 MHz computer system and controlled by Autolab GPES 4.8 software (version for Windows 98).

Amperometric flow injection analysis was carried out with a simple system that mainly comprised a 12-cylinder Perimax Spetec peristaltic pump (Spetec GmbH, Germany) together with a manual six-port rotary valve, Model 1106 (Omnifit Ltd., UK), and a homemade wall-jet flow cell (Fig. 1) where the SPCEs were placed. The flow cell includes two pieces of methacrylate screwed together; one of these has inlet and outlet flow channels forming an angle of 30°. The SPCEs were fixed in between and were easily connected to the potentiostat through the edge connector. A syringe was used to aspirate the samples into the system by filling a loop of 50 μL that was further discharged into the flow system by means of the injection valve.
Fig. 1

Flow cell equipped with a screen-printed carbon electrode (SPCE)

Screen-printed carbon electrodes and a specific connector were purchased from Alderon Biosciences (Durham, NC, USA). These sensors have been described elsewhere [25]. A model Tecan Sunrise Remote/Touch screen microplate reader was used for ELISA with spectrophotometric detection. Microtiter plates (flat bottom, high binding) were purchased from Costar (Cambridge, MA 02140, USA). An eight-channel micropipette and an MS1 minishaker (IKA-Werke GmbH Co. KG, Germany) with a microtiter adapter and a Sanyo incubator were also employed.

Analytical procedures

Flow injection analysis

The working carrier (H2SO4) was flowed through the electrochemical cell by a peristaltic pump, and the absence of bubbles was checked. When a constant baseline current was reached, solutions were injected into the flow stream via the automatic valve loop (50 μL) and the fiagram was recorded at a fixed potential.

Enzyme determination

The measurement of the HRP activity was performed as follows. A 200-μL aliquot of the blocking buffer was added to each well and incubated overnight at 4°C. After washing four times with an AcOH/AcO pH=5.0 buffer solution and drying, 50 μL of TMB solution and 50 μL of H2O2 were mixed with 8 μL of an HRP solution in each well. These were then incubated for 15 min at room temperature and under constant shaking. The enzymatic reaction is terminated by addition of 100 μL of 1 M H2SO4. This results in a pH shift to below unity and a complete conversion of the blue product to its yellow diimine form, which can then be quantified by injecting it into the FIA system.


An ELISA test kit for IL-6 measurement was kindly provided by VITRO SA (Madrid, Spain) and was used in accordance with the manufacturer’s instructions (R&D Systems, Inc.): this procedure is illustrated in Fig. 2. IL-6 standard solutions (100-μL volumes) were incubated 2 h at room temperature in each well (containing the immobilized murine monoclonal antibody against IL-6). A 200-μL volume of the antibody–enzyme (HRP) conjugate solution was then incubated in the wells at room temperature for 2 h. Between the above mentioned steps, the plate was emptied and washed four times with a washing buffer. After drying, 200 μL of substrate mixture TMB/H2O2 was pipetted into each well. The enzymatic reaction took place for 20 min at room temperature and was then stopped by adding 50 μL of a 1 M H2SO4 solution. Finally, 50 μL of the product was injected into the FIA system coupled with the amperometric detection.
Fig. 2

Scheme of the IL-6 ELISA procedure

Results and discussion

Cyclic voltammetry

Cyclic voltammetric studies were carried out at an untreated SPCE in a batch system and at an acid pH wherein the enzymatic product, yellow diimine, is stable [7, 8]. Figure 3 displays a staircase cyclic voltammogram of the substrate mixture, TMB/H2O2 in a 0.5 M H2SO4 solution, when a potential scan ranging from −0.25 to +0.7 V (vs. Ag-pseudoreference electrode) was applied. It is characterized by a well-defined reversible electrode signal with a formal potential (E1/2) of 0.452 V (vs. Ag-pseudoreference electrode). In this study, the peak/semi-peak potential separation is 30 mV, around the expected Nernstian value for a two-electron reversible process, as was evidenced in earlier polarographic studies [8].
Fig. 3

Cyclic voltammogram of the substrate mixture (TMB + H2O2) in 0.5 M H2SO4. v=50 mV s−1

When the TMB/H2O2substrate mixture is incubated in the presence of HRP prior to the addition of the stopping solution (1 M H2SO4) the electrochemical oxidation of TMB (shown in Fig. 3) is accomplished enzymatically. Therefore, the enzymatic reaction product, TMB (Ox), can be detected through its reduction at the surface of the SPCE, according to the scheme shown in Fig. 4.
Fig. 4

Mechanistic scheme of the electrode process

Scan rate was varied from 2.5 mV s−1 to 100 mV s−1 and cyclic voltammograms were recorded. A linear relationship between the peak current and the square root of the scan rate [I (μA)=1.63v1/2 (mV s−1)1/2−0.154; r2=0.9999] demonstrated that the process is diffusion-controlled. In contrast with these results, when a gold disk electrode was used as working electrode, the electrochemical process of the enzymatic product is adsorption-controlled [14].

Amperometric detection

Amperometric FIA hydrodynamic voltammetry for TMB (Ox) was carried out in the range from −0.2 V to +0.2 V to fix the optimum detector potential. A working potential of +0.1 V was selected for the measurement of HRP activity in order to obtain highly repeatable signals. Moreover, at this low working potential the current background was near to zero, no substrate oxidation occurred, and an adequate stability of the electrode was achieved.

The flow rate of the carrier stream was varied between 1.2 and 6.2 mL min−1 in order to study the influence that this parameter had over the reduction signal of the enzymatic product, TMB (Ox). Peak currents increase with increasing flow rate as is reported in Table 1. Moreover, the use of a high flow rate decreases the analysis time and is also important in the cleaning of the electrode surface. This surface could suffer from fouling by the adsorption of products from the electrochemical process. This fact constitutes a drawback of voltammetry on solid electrodes; however, it is minimized when the electrode is used in flow systems. In Table 1 the relative standard deviation (RSD) of five successive measurements, corresponding to 4.5×10−12 M HRP determinations, is also reported at different flow rates. A flow rate of 4.3 mL min−1 was used for the remainder of the work because higher peak currents and lower RSD values were obtained.
Table 1

Flow rate, number of measurements, cathodic peak current, and relative standard deviation for 4.5×10−12 M HRP determinations recorded in 0.5 M of H2SO4 (E=+0.1 V)

Flow rate (mL min−1)


I (nA)

RSD (%)

















Taking into account that the stopping solution for the HRP and ELISA assays is sulfuric acid, the working carrier stream was varied between 0.1 H2SO4 and 1 M H2SO4 with the aim of being quite similar to the nature of samples injected in the FIA system. In order to obtain the best baseline stability, a flow carrier of 0.5 M H2SO4 was employed in all subsequent studies.

The homemade electrochemical flow cell wherein is inserted the SPCE presents two flow channels: one perpendicularly fitted to the electrode strip and the other forming an angle of 30°. The wall-jet configuration was chosen for the measurement of the analytical signals. The repeatability of the peak current was estimated on a 4.5×10−13 M HRP determination with one as well as with five different strips. The intra-electrode relative standard deviation (RSD) was 3% (n=15, Fig. 5); meanwhile the inter-electrode value was around 7% (n=5).
Fig. 5

Successive signals corresponding to the determination of 4.5×10−13 M HRP with the same electrode strip

HRP determination

Under the selected enzyme-catalyzed reaction conditions reported in the “Experimental” section, previously optimized amperometric detection was used to measure the height of the reduction peak of the enzymatic product, TMB (Ox). This showed a good linear relationship with the concentration of HRP in solution in the range from 4.4×10−14 to 8.7×10−12 M. The corresponding equation is:
$$ \begin{aligned} - I\,({\text{nA}}) = 8 \times 10^{13} [{\text{HRP}}](M) + 4.95 & \\ r^2 = 0.9994\quad n = 8 & \\ \end{aligned} $$
The relative standard deviation (RSD) for seven consecutive measurements of 4.5×10−13 M of HRP was 4.2% with 39.68 nA average cathodic peak current. The detection limit is 2×10−14 M (calculated as the concentration corresponding to three times the standard deviation of the estimate), which appears promising for enzyme immunoassays. The analytical signals corresponding to the calibration curve for HRP are shown in Fig. 6. Working under previously optimized conditions, no baseline drift or baseline noise were observed on the background current. The first four analytical signals were recorded in the absence of HRP showing a RSD value of 2.5% with a mean reduction peak current of −7.16 nA. The lower limit of linearity range corresponding to 4.4×10−14 M HRP gave a peak current of −10.45 nA. TMB appears to be suitable for use as an electrochemical substrate for HRP determination owing to its limited spontaneous oxidation in the presence of hydrogen peroxide.
Fig. 6

Amperometric responses of the FIA system corresponding to the HRP calibration curve (A). Amplified graph of the lower signals recorded (B). (1) blank, (2) 4.4×10−14, (3) 8.7×10−14, (4) 1.7×10−13, (5) 4.4×10−13, (6) 8.7×10−13, (7) 1.7×10−12, (8) 4.4×10−12, (9) 8.7×10−12 M HRP

Screen-printed carbon electrodes did not show any surface contamination after successive injections of the enzymatic product into the flow system, which allowed the use of the same strip for several days. Thus, it is possible to avoid either the application of cleaning steps or the replacement of the electrode system after each experiment. Furthermore, when the change of the SPCE is necessary (a high baseline noise is registered), the procedure is as simple as inserting a new board into the flow cell. This process is easier than that required for renewing of a common amperometric detector with a glassy carbon as working electrode.

ELISA determination

The IL-6 assay was carried out by following the sandwich procedure described under “Experimental” and illustrated in Fig. 2. To evaluate the performance of the amperometric detector for conventional ELISA, the absorbance measurements (recognized as a standard method) were used as a reference parameter. The correlation was determined by the spectrophotometric and amperometric signals obtained in the range from 3.12 to 300 pg mL−1 of IL-6 (it is the range of standards reported in the kit to generate the standard curve).

Figure 7 shows a linear correlation between the amperometric signal and the optical density values in the range required. The linearity coefficient was determined to be r2=0.9997. Analytical signals used to determine the correlation were obtained after subtracting the average zero standard optical density or peak current for each measurement corresponding to the range (3.12–300 pg mL−1) assayed. With regards to repeatability, an RSD intra-assay lower than 7% was always found.
Fig. 7

Correlation between amperometric signal and absorbance values with a linearity coefficient r2=0.9997 (n=7)

These results show that the amperometric detector is available in the end point measurement of this ELISA using HRP as enzymatic label and TMB as substrate.

Compared to the conventional optical detector, an amperometric detector has advantages such as simplicity of the instrument and low production cost. However, for the routine application of the amperometric detector to ELISAs, further improvement of the analysis time is required. In this work each measurement was taken about 40 s but an array of electrodes the same as in the case of the optical detector should be considered.


The coupling of amperometry to SPCE makes the generation of cheap and simple prototypes for the sensitive detection of TMB/HRP-based assays possible. A homemade electrochemical flow cell which allows the easy replacement of the electrode system was proposed. When this flow cell is joined to an FIA system, repeatable signals were obtained and no baseline drift or baseline noise were observed on the background current

The practical feasibility of using this approach was demonstrated through the sensitive quantification of HRP in solution. Moreover, its applicability in a conventional ELISA was demonstrated through its correlation with the spectrophotometric standard method of detection. Work is in progress to decrease the analysis time through the development of an array of electrodes as detector.


This work has been supported by the Project BIO2003-06008-C03-01.

Copyright information

© Springer-Verlag 2005