Medical Device Link.

IVD Technology Magazine
IVDT Article Index

Planar electrochemical biosensors for critical-care applications

Hans Ludi

The diagnostic instruments currently used in most critical-care environments to measure blood gases, electrolytes, and metabolites (especially glucose) employ an electrode technology first commercialized in the 1970s. Such systems typically require a significant amount of maintenance: they make use of reagents and electrodes that must be replaced regularly, and have a number of moving parts that must be serviced periodically. Over the past 10 years, since the introduction of membrane-covered electrodes that do not need to be exchanged regularly, the maintenance requirements for such electrode-based systems have diminished. Nevertheless, required maintenance is still problematic because all parts and reagents must be serviced individually.

Figure 1. A planar electrochemical glucose sensor.

Technological change in this area has been driven in part by the successful development of home-use glucose monitors. The key component of such monitors is a flat, membrane-covered unit called a planar sensor (see Figure 1). Over the past few years, planar sensors have been brought into use for measuring blood gases and electrolytes, and have also been introduced in portable cartridge-based critical-care systems. However, integrating a complete menu of critical-care analytes in a single cartridge remains a major technical challenge. The change from three-dimensional electrodes to planar sensors requires the use of specialty chemicals to meet the new functional requirements of planarization and miniaturization. Because planar sensors also have a shorter use-life than traditional sensors (from a single use to use for fewer than 1000 samples), it will also be necessary for manufacturers to develop techniques to achieve higher production volumes.

For product developers, the main consequence of this shift will be the need for closer links among the activities related to sensor design, the creation of sensor membrane chemistries, and the development of production processes. This article outlines the performance characteristics and design requirements related to the use of planar sensors in diagnostic test systems, and discusses some of the functional requirements and processing technologies needed to bring this new technology into the diagnostic marketplace.

Inherent Performance Characteristics

For many years, researchers have believed that planar biosensors could offer performance characteristics that would make them well-suited for applications in critical care, point-of-care, bedside, and home-use test systems. Compared to sensors using three-dimensional electrodes, planar sensors are inherently smaller, more responsive, and more versatile. When coupled with biomolecules, they can be formed into biosensors with high sensitivity and specificity. And by adapting manufacturing processes from the semiconductor industry, they can be produced at unit costs low enough to make their use cost-effective. The key characteristics that make planar biosensors attractive to product developers are the following:

Small Size. To achieve the portability necessary for developing true handheld, bedside, and home-use testing systems, miniaturization is a must. Product developers desire the flexibility to match the format of their instruments to customer needs. In some cases, personal convenience calls for a handheld, battery-powered unit. In others, the fact that handheld units can easily be dropped or misplaced is considered a disadvantage, making a benchtop unit more desirable. The small size of planar sensors—typically less than 0.5 mm²—is an advantage in many such applications, offering product developers a platform that can be readily adapted to a variety of formats (see Figure 2). The small size of planar sensors makes them the technology of choice for any test system that employs replaceable cartridges.

Time-to-Result. Reductions in the size of test instruments may increase the number of settings where they can be used, but few users would consider adopting such systems if they were not can also fast. Time-to-result is thus a key criterion for measuring the acceptability of any such test system. Because planar sensors require smaller sample volumes than traditional analyzers, they can also have a faster reaction time. However, the reaction time of a sensor is generally determined by the speed at which the sample diffuses, and the speed of the molecular interactions between the analyte and the sensor's recognition element.

Versatility. A number of manufacturers are now developing benchtop diagnostic systems that promise to offer quantitative testing for many analytes. Like their larger clinical laboratory counterparts, these instruments will inevitably compete to some degree on the basis of their menu of available tests. Using planar sensor technologies could enable manufacturers to expand their test menus. Researchers have described the application of planar sensors for a wide variety of analytes, including chemistries, immunoassays, and nucleic acid assays.^1,2 At present, the majority of commercially available planar sensors are intended to analyze blood gases, electrolytes, and metabolites. The greatest influence on the future versatility of planar sensors is likely to come from the development of optical transducer technologies, such as combinations that make use of fluorescent or chemiluminescent recognition elements with planar waveguides, laser diodes, and charge-coupled device cameras.

Sensitivity and Specificity. Without question, these factors are the sine qua non of any diagnostic test system. A system that cannot offer sensitivity and specificity at least comparable to existing tests is not only unlikely to gain market acceptance, it is also unlikely to be considered reimbursable or to gain FDA approval. Planar sensors using high-sensitivity transducers and detectors have been shown to achieve the sensitivity and specificity required in diagnostic test systems.^1,2 The current generation of planar sensors can reliably detect a wide range of natural compounds (ions, metabolites, enzymes, antibodies, molecular receptors, etc.) with sensitivity and specificity suitable for diagnostic testing.

Ease of Manufacturing. For some years, researchers have believed that it would be an easy matter to adapt the batch processing techniques developed by the semiconductor industry for use in the production of planar sensors. Once accomplished, such a shift was expected to make possible the production of low-cost sensors in large quantities. In reality, adapting chip manufacturing techniques has proven to be a greater challenge than anticipated. One stumbling block has been the incompatibility of such techniques with the processing requirements of unstable biological compounds. Another has been the relatively high cost of implementing batch processing at this stage in the technology's evolution. Proponents continue to believe that these difficulties will be overcome, and that the initial promise of low-cost, easily manufactured sensors will eventually be fulfilled.

Cost-Effectiveness. To customers, cost of ownership is what counts. Low sensor costs may not necessarily translate into a low cost of ownership, and may therefore be relatively unimportant for the market acceptance of a new test system. The actual cost of sensor components is probably of greater importance to manufacturers, who must evaluate the cost of adopting this technology relative to the cost of using an alternative technology to obtain test results. As advances in planar sensor design and manufacturing reduce the costs associated with using them, manufacturers will undoubtedly become increasingly interested in adopting them into their next-generation test systems.

Figure 2. A planar electrochemical glucose sensor. Size is demonstrated by a comparison with a U.S. dime. The sensor is packaged for use in a blood gas, electrolyte, or metabolite system.

System Requirements

Although planar sensors inherently possess a number of characteristics that could make them useful in diagnostic applications, these traits are far from the only requirements that such sensors must meet. FDA and other regulatory agencies worldwide will have some influence over the eventual shape of any diagnostic test system that incorporates planar sensors. Laboratorians and other users also impose performance requirements for such systems, and these requirements can vary widely according to the type of user, the settings in which the system is designed to be used, and the actual types of tests the system will run. Taken together, these additional system requirements pose a considerable challenge for product developers who want to make use of planar sensors. Following are some of the key areas in which regulatory or customer requirements may be important.

Precision and Accuracy. Laboratorians are accustomed to the precision and accuracy delivered by today's highly sophisticated laboratory equipment, and they are unlikely to tolerate any reduction of these parameters. The same standards are also being applied to test systems designed for use outside the clinical laboratory, meaning that even units designed for home use will be required to perform as precisely and accurately as much larger systems. This requirement will challenge developers to design systems that can be operated reliably without the intervention of trained laboratorians.

Use- and Shelf Life. The shelf life of planar sensors can be a key factor in achieving efficiencies in manufacturing, distribution, and sales. Even with just-in-time inventory delivery, professional customers will consider products that have a short shelf life to be unacceptable. This factor is even more important for test consumables designed for the over-the-counter or prescription home-use market. For product developers, the challenge is to design components so as to maintain the integrity of minute amounts of chemical or biological compounds for several months, preferably at ambient temperatures. The compounds must also remain stable in use, during which the patient sample is often brought to 37°C to allow a measurement close to body temperature.

Ease of Use. Independent of the measurement technology used to derive a test result, customers insist that their diagnostic systems be easy to use. Such a requirement is true for both professional users and laypersons, though with significantly different definitions of what constitutes an easy system. From the manufacturer's perspective, the goal is to reduce the complexity of the system (as categorized by the Centers for Disease Control and Prevention [CDC] or FDA) to as low a level as possible. A waived-complexity unit, for example, would be suitable for use by untrained laypersons, giving the manufacturer the broadest possible market for sales of the device. Few diagnostic systems will achieve that status, but product developers will still be challenged to make their devices easy to use. In large measure, this factor is determined by the user interface, which is what customers must deal with when they operate the system. To the extent that planar sensors can help to reduce the complexity of the user interface (e.g., through automated calibration or quality control), they can contribute to making systems easy to use. In many respects, however, effecting ease of use is outside the scope of sensor development.

Whole Blood. Customers of all types desire systems that can use whole blood. This capability eliminates the need for extensive sample preparation, and therefore also the potential for preanalytical error. For planar sensors, however, achieving this goal has proven to be one of the most challenging tasks of all. This is largely because of the relatively wide spectrum of pH, ionic strength, viscosity, water content, and interfering substances that can be present in a whole blood patient sample. All of these factors can affect the response of the biomolecules used in the sensor, thereby reducing the accuracy and performance of the system.

Sample Handling. According to the test being performed, laboratorians must often use different sample-collection devices or employ different sample-preparation methods. The goal in this area is to reduce complexity by eliminating steps or using automated methods to compensate for variations. For instance, anticoagulants used in some sampling devices can interfere significantly with the measurement of calcium and glucose, and it has proven a major challenge to design sensors that can accommodate all the various types and concentrations of anticoagulant found in patient samples.

Regulatory Requirements. FDA's quality system regulation specifies design control and good manufacturing practices requirements that must be met by all diagnostic systems marketed in the United States. Product approval in Europe is no less rigorous. These quality requirements apply both to the manufacture of the sensor components and to the diagnostic system as a whole. Compliance with such requirements is more than a matter of mere inspection or testing, and requires manufacturers to develop specific systems for ensuring the quality of their products. Like all other diagnostic equipment, test systems that incorporate planar sensors must meet these regulatory requirements.

Connectivity. Primed by the explosion of information technologies over the past decade, customers are increasingly interested in making sure that new diagnostic systems possess a full menu of connectivity options. Such a requirement is being placed not only on sophisticated clinical laboratory systems, but also on personal, handheld monitors. Ultimately, users want test results to be available whenever and wherever they are needed. Remote access to analyzers (e.g., by service technicians or central laboratory personnel) is also considered a highly desirable feature, because it helps users to ensure proper calibration, use, and functioning of the equipment. According to some manufacturers, equipment is also being developed to permit the automated inventory and reorder of consumables used by such test systems. While essential to the marketability of future diagnostic systems, connectivity features are largely outside the scope or influence of sensor development.

Figure 3. An amperometric lactate biosensor using a cover membrane with limited diffusion for lactate, but unrestricted diffusion of oxygen. The membrane at the same time acts as an interference-rejecting membrane. To measure the presence and level of interferences (mostly present as a combination of metabolites and drugs) a correcting electrode is used. Custom inks are used to apply the enzyme.

Functional Requirements

For each planar sensor that is to be employed in a diagnostic system, researchers must translate the performance characteristics and system requirements described above into a set of functional requirements. Functional requirements are used by scientists and engineers in developing the technical solutions that will enable a specific analyte to be measured by a sensor.

Following are some of the functional requirements created as the basis for developing planar sensors for blood gases, electrolytes, and metabolites. Also provided is a short description of the technical solution formed to fulfill the functional requirements. All of the components described below use a ceramic wafer as their substrate; most of the sensors consist of two or more polymeric membranes.³ The sensors were designed to run 750 whole blood samples or to last 28 days (whichever comes first).

Amperometric Oxygen Sensors. In this sensor, the internal electrolyte must provide the necessary conductivity for current flow and must remain stable under constant polarization. Nafion (polymeric perfluorinated ionomer) was found to show sufficiently fast wet-up (time required for the sensor to be ready to report results) and maintained chemical integrity for more than 30 days even when polarized to ­800 mV at 37°C (see Figure 4).

Figure 4. For this planar oxygen sensor, it was necessary to develop two main components: a cover membrane that allows for both rate-limiting oxygen diffusion and fast water vapor diffusion; and high-purity metal inks to avoid such electrochemical reactions as the deposition of materials on the electrodes, which are polarized constantly at –800 mV. The resulting sensor has a use-life of more than 30 days.

To maintain the integrity of the electrochemical cell over an extended period of use, the cover membrane of this sensor must have low oxygen permeability, resulting in low current levels. To achieve a fast wet-up time, on the other hand, the membrane must allow for rapid transport of water vapor. Wet-up time requirements are very stringent for single-use sensors, but wet-up times of less than 20 minutes are very desirable even for multiuse sensors. These two requirements—restricted oxygen permeability and rapid water vapor transport—are not necessarily compatible. The desired properties were achieved simultaneously only through the use of a patented acrylate copolymer.

Potentiometric Ion Sensors. In this sensor, the solid-state internal contact must provide conductivity for ions as well as a stable reference potential (see Figure 5). Changes in interface potential can cause offset potential drifts that must be reduced to a level that will not affect sensor performance. As with all such critical-care systems, regular calibration is still required to ensure that system performance meets specifications. A copolymer of methacrylamidopropyltrimethylammonium chloride and methyl methacrylate can be used.

Figure 5. In this planar potassium sensor, a solid internal approach is employed to fulfill the function of an internal fill solution in a conventional three-dimensional electrode. In this planar format, the volume of the methacrylamidopropyltrimethylammonium chloride (MAPTAC) hydrogel layer is approximately 0.2 µl, compared with volumes greater than 100 µl used in conventional electrodes. This design has achieved stable off-set potentials (<0.01 mV/min drift) over a period of more than 30 days.

Amperometric Enzyme Sensors. Researchers have shown that the addition of polyvinylalcohol or carbohydrates can stabilize biological compounds such as enzymes. Interference-rejecting membranes such as cellulose acetate and FC 61 from Dow Chemical (Midland, MI) have been used in the development of glucose sensors. They are now also routinely used for planar biosensors to meet the performance characteristics of specificity, use-life, and shelf life (see Figure 3).

Manufacturing Processes

In the development of planar sensors, selection of the appropriate manufacturing process is of the utmost importance (see Table I). The first decision that must be made is to select the substrate for the sensor (e.g., silicon, ceramic, glass, plastic). There is no "right" choice. Instead, manufacturers should consider the type of analyte to be measured, whether the test system requires single-use or multiuse sensors, the strengths and weaknesses of various processing alternatives, the costs associated with those alternatives, and all other system requirements.

The functional utility of a planar sensor can be described by the slope of its calibration curve, a graph that describes how all of the sensor's attributes interact to measure a known concentration of the target analyte (see Figure 6). Reproducibly manufacturing sensors with a specified slope while also increasing manufacturing yields is a major challenge. Figure 7 illustrates the slopes of calibration curves measured over the 60-day use-life of 1400 preproduction multiuse planar glucose sensors with a design similar to that of the lactate sensor shown in Figure 3. Each data point represents the slope of the calibration curve for an individual sensor; the line represents an arbitrary mathematical fit to the data points. Taken as a whole, the figure displays a wide distribution of slope among different production lots and over the 60 days of use-life.

Figure 6. Planar electrochemical sensor development. Singulated chips are placed in a module for performance testing.

Nevertheless, most of the 1400 sensors tested met all performance requirements over their use-life. This result suggests that if a minimal slope can be achieved, sensors can meet their performance requirements. The satisfactory performance of these sensors was supported by the fact that they were regularly calibrated and their quality control values were subjected to trend analysis. When such sensors are adopted for use in a blood gas, electrolyte, or metabolite test system, calibration must be performed in order for the system to meet regulatory requirements.

Figure 7. Slopes for a glucose sensor over a 60-day use-life, showing results from reproducibility tests of more than 1400 preproduction sensors.

Figure 8. Bias to reference results for multiple lots of preproduction glucose sensors tested over a 60-day use-life.

Figure 8 shows the bias to reference of the same 1400 preproduction glucose sensors. Bias to reference is a tool for determining the accuracy of a test system. Using the same sample, it establishes the difference between the test values measured by the system and by the reference method (in the case of glucose, the CDC hexokinase method). Each data point represents a single determination of bias to reference; the graphed line is a linear fit through the data points. The graph makes it apparent that not all sensors met their bias-to-reference specifications at all times. To correct such a problem, the manufacturer would need to optimize its manufacturing processes and implement lot acceptance procedures.

Conclusion

The examples provided in this article suggest that planar electrochemical sensors offer useful performance characteristics and can meet the requirements for diagnostic systems established by customers and regulatory agencies. In the near future, the adoption of such miniaturized planar sensors should make it possible to design critical-care testing systems that employ easy-use cartridge-type consumables.

Such systems have the potential to simplify sample handling and test operation, and make it possible for manufacturers to offer a wide range of diagnostic tests on a single platform. However, market acceptance of these new systems is likely to be driven by such factors as reduced cost of ownership and, ultimately, their impact on the quality of medical decisions.

References

1. FW Scheller, F Schubert, and J Fedrowitz, eds., Frontiers in Biosensorics, vol. 1: Fundamental Aspects (Basel: Birkhaeuser, 1997).

2. FW Scheller, F Schubert, and J Fedrowitz, eds., Frontiers in Biosensorics, vol. 2: Practical Applications (Basel: Birkhaeuser, 1997).

3. P D'Orazio, et al., "Planar (Bio)sensors for Critical-Care Diagnostics," Clinical Chemistry 43 (1997): 1804.

Hans Ludi is director of critical-care R&D at Bayer Business Group Diagnostics (Medfield, MA).

Photos courtesy Bayer Business Group Diagnostics

Back to the IVDT Sept. 99 table of contents