Medical Device Link.

Originally Published IVD Technology May 2004

Detection Technologies

Biosensors, transducers, and other electronics are being incorporated in a way that allows them to be produced via conventional integrated circuit manufacturing processes. As part of such a process, a technician inspects a preproduction prototype IC wafer at a test station.

Mounting an effective defense against bioterrorism and biowarfare agents will require fast, reliable environmental sensors that provide early detection of chemical and biological threats. Accurate characterization of electronic sensor components can help speed up development.

Many types of biosensors, transducers, and detection systems are being investigated for bioterrorism defense applications, and also for more-general use.¹ Some of the designs being investigated generate an electrical output, which is intended to speed up measurements and meet ease-of-use criteria. Proper testing of the electrical portions of these sensors is needed to qualify the designs for further development.

Current Biosensor Designs

This article focuses on the characterization of components and assemblies used in electronic biosensors. These components are designed to detect, record, convert, process, and transmit information regarding a physiological change or process; utilize biological materials to monitor the presence of various chemicals in a substance (e.g., an analyte); or combine an electrical interface (i.e., a transducer) with the biologically sensitive and selective element.

The transducer portion of a biosensor converts a biorecognition event into a measurable signal that correlates with the quantity or presence of the chemical or biological target of interest (see Figure 1). In this case, the target is one that could cause physiological damage. Electronic component and assembly testing are needed to meet various performance criteria. These criteria include: speed, ease of use (i.e., testing, calibration, and maintenance) by nontechnical personnel; specificity to the target analyte; sensitivity; accuracy, resolution, and repeatability; dynamic range (i.e., high analyte concentrations will not degrade sensor usability); environmentally robust (i.e., relatively insensitive to temperature, electrical noise, physical shock, vibration, etc.); reasonable lifetime; adaptability; safety; and integrity (i.e., when used by personnel, equipment, and with various analytes).

Current immunochemical sensors can accurately recognize protein-based agents, and electrochemical sensors can identify DNA. However, most of these sensors require significant sample preparation. For toxic airborne gases, inorganic sensor materials such as tin oxide, gold, platinum, and semiconductors are used for in-line measurements. These chemical sensors can react quickly to their selected targets and provide a convenient electrical output. However, there is no way to detect the presence of a biopathogen with a simple enzyme-based biosensor, or to sense the DNA inside the pathogen.

For field use, it is highly desirable to have a biosensor that can provide a high-level electrical signal for direct readout on a visual indicator. An ideal biodefense sensor would have electrical characteristics similar to those of a simple glucose sensor (See Figure 2).²

Investigative Biosensor Designs

Figure 1. Representation of a generic biosensor (Click to enlarge).

For fast detection with a readable electrical output, researchers are trying to develop biosensors that combine various processes, such as pathogen cell lysis, debris removal, DNA cleanup, polymerase chain reaction (PCR), and fluorescence detection. Work is going on to combine these and similar processes with semiconductor devices to create compact portable units for field use.

One such effort involves the development of an oligonucleotide sensor and nucleic acid reaction to indicate the presence of a pathogen.³ Other work is being done with surface plasmon resonance (SPR) to detect biological molecules such as protein and DNA.

In addition, living cells that can react functionally to the presence of both biological and chemical threat agents are being placed on chips. Lastly, scientists are developing liquid-crystal detectors that can selectively identify pathogens by the way they alter light-transmission characteristics.⁴

Depending on the analyte and bioreceptor, the transducer portion of a biosensor could utilize one of several mechanisms. The transducer could rely on amperometric technology, which involves devices that detect changes in current. They measure currents generated when electrons are exchanged between a biological system and an electrode (see Figure 2). Alternately, it may use a potentiometric mechanism which involves reactions that cause a change in voltage (i.e., potential at constant current) between electrodes and this change can be detected or measured. 

A transducer that relies on resistive or conductive mechanisms will detect changes in conductivity or resistivity between two electrodes. A capacitive transducer relies on a biorecognition reaction, which causes a change in the dielectric constant of the medium in the vicinity of the bioreceptor. Such a capacitance measurement method can be used as a transducer.

A transducer can also utilize a piezoelectric mechanism. In a piezoelectric material there is a coupling between its mechanical and electrical properties. The coupling can be used to create an electrical oscillator, the frequency of which can be varied and measured by varying a mass applied to its surface. In the case of a biosensor, that mass can change due to the reaction taking place on the surface.Thermal transducers measure changes in temperature.

Biosensors with optical mechanisms correlate changes in concentration, mass, or number of molecules to direct changes in the characteristics of light. For this process to work, one of the reactants or products of the biorecognition reaction has to be linked to colorimetric, fluorescent, or luminescent indicators. An optical fiber might perform this linkage by guiding light signals from the source to the detector.

Device Characterization

Biosensor development programs generally aim to overcome design limitations in current biosensor systems. For example, one of the challenges involved in biosensor design is achieving a stable, reproducible interface between the biological affinity elements and an inorganic transducer element in the sensor. 

The desire to miniaturize biosensors for handheld portability and still achieve adequate sensitivity imposes significant technical challenges in the coupling of biomolecules to transducer surfaces. Therefore, fast and accurate electrical characterization of biosensor devices and mechanisms in the development lab is essential for qualifying new designs.

Because of the complexity in extracting cell and tissue signatures of agent activity and response, it is often desirable to conduct direct current-voltage (I-V) characterization on key components of the biosensor. I-V characterization requires only a small fraction of the time needed for most types of functional testing, but is a powerful predictor of full-fledged operation. For example, I-V data can be used to study anomalies, locate maximum or minimum curve slopes, and perform reliability analyses. Depending on design specifics, I-V characterization is often suitable for sensors based on amperometric, potentiometric, conductive, resistive, and thermal principles.⁵

Figure 2. Electrical characteristics of a prototype electrochemical transdermal sensor based on a glucose oxidase reaction with blood plasma from capillaries.2 (Click to enlarge).

Usually I-V testing applies a voltage or current to the device under test (DUT) and measures its response to that stimulus (see Figure 2.) Temperature measurements may also be taken. The test procedures may involve the probing of integrated circuits to apply the stimulus to certain connection pads and measure the response on others.

Depending on the DUT, signal levels may be quite low. In this case, highly sensitive source and measurement instruments and test techniques that minimize external sources of error are necessary. When an optical mechanism is involved, I-V characterization may also involve simultaneous measurements of the wavelength or intensity of a light output with a photodetector. This is called L-I-V testing.

Characterizing Device Performance

In many cases, biosensors used by medical practitioners, military personnel, and public safety forces will be part of a portable system. This places restrictions on the operational power requirements of the sensors and may dictate the level of voltage or current output that can be provided to measurement circuitry. In battery-operated systems, sensor output current can range from nanoamps to milliamps, and voltage from nanovolts to volts. Different measurement techniques and tools are required for signal levels at the opposite ends of such wide ranges.

Voltage Instruments. Characterizing biosensor devices at voltage levels greater than 100 µV should be relatively easy. A sampling data-acquisition system based on a PC plug-in board may provide adequate resolution, as would many programmable digital multimeters (DMMs) and self-contained data loggers. For example, most laboratory-grade DMMs provide enough range and resolution to make voltage measurements from 1 µV to 1000 V. For the PC-based data-acquisition board solution, you can measure over a variety of voltage levels, depending on the resolution of the analog-to-digital (A/D) convertor and its gain (see Table I).

For measurements below 1 µV, a nanovoltmeter should be considered instead of a data acquisition board or DMM solution. For example, the Model 2182 by Keithley Instruments Inc. (Cleveland) is a low-noise digital nanovoltmeter with an A/D resolution in the 24-bit range. This type of instrument is optimized for accurate low-level voltage measurements, even when the signal is approaching the theoretical (lower) limit associated with sensors that have a low output impedance. Sensors with a low output impedance require a voltmeter with a high input impedance to avoid measurement errors. Although the input impedance of a nanovoltmeter is similar to a DMM, it has much lower voltage noise and drift. This gives it much better voltage sensitivity, and it may be able to read down to 1 nV. 

Noise in Voltage Measurements. Noise is often a problem for tools used to measure voltage. Significant errors can be generated by noise that originates in the sensor, the measuring instrument, and sources external to the test circuit. External sources include electromagnetic fields (EMFs), measuring-circuit ground loops, and thermal EMFs. Johnson noise is another form of thermal noise that occurs in every electrical component. Johnson noise establishes the ultimate limitation of the measurable signal level. For accurate measurements, noise sources should be minimized as much as possible.⁷

In some test environments, electrical noise is difficult to avoid. When that happens, measurement compensation techniques are needed. These are included in most benchtop DMMs and nanovoltmeters to minimize electrical noise from ac lines and from random noise sources. These techniques are less likely to be available in PC-based data-acquisition systems, but knowledgeable users can program a system for signal averaging to help reduce random noise, and use longer A/D integration periods to minimize ac line noise.

Table I. Voltage measurement resolution and maximum ranges for different A/D convertor resolutions. (Source: Data Acquisition and Control Handbook.6) (Click to enlarge)

Current Measurements. Different types of amperometric devices require different characterization approaches. Electrical currents can be measured with data-acquisition systems, but the method of acquisition selected will depend on the current level and number of required measurement channels. Otherwise, I-V characterization of a device using a current loop is uncomplicated as long as the current source output voltage is high enough to overcome any test lead resistance. The corollary is that current loops are ideal when there is an appreciable distance between the signal source and the instrumentation.

It may be desirable to qualify the output of a biosensor assembly in terms of engineering units, such as concentration of the target analyte in milligrams per deciliter. In the case of an amperometric output, the current is an indicator of the phenomenon actually being measured. Therefore, it is desirable to have instrumentation that makes it easy to do the conversion. This requires that the instrument be able to perform internal calculation and scaling features that convert current readings to appropriate engineering units. For example, a transducer utilizing a 4–20-mA current loop might be calibrated for a concentration of zero mg/dl at 4 mA and a full-scale concentration of 100 mg/dl at 20 mA.

Figure 3. Current measurement using a dropping resistor and voltmeter (Click to enlarge)

In the early stages of product development, it may not be necessary to do this conversion when the significance of the transducer’s output current level is already understood. In this case, the current reading is the parameter of interest and a voltage stimulus may serve as a proxy for the biological event during I-V characterization. Under these circumstances, the current signal can be relatively high, but may require care in selecting a dropping resistor for the measurement. The current passes though the dropping resistor at the input stage of a data-acquisition system, and voltage is measured across the dropping resistor to determine the current level. This type of scaling is a common feature of data-acquisition systems.

However, unlike voltage measurements, current measurements may be subject to “voltage burden” errors. Voltage burden is defined as the voltage drop across the input of an ammeter when it is inserted into a circuit. The dropping resistor (R) and A/D voltage input constitute an ammeter and the current flow can be calculated from the voltage drop across the resistor (see Figure 3).² Noise effects, similar to those for voltage measurements, must also be considered and are exacerbated by voltage burden.

Figure 4. SourceMeter instrument configured as a constant- current source and voltmeter that measures the DUT response. Its compliance function makes sure that voltage does not exceed a preset level (Click to enlarge)

The resistance value will normally be selected to provide a voltage drop corresponding to the full input range of the A/D board when the maximum anticipated current flows (see Figure 3). For example, a 20-mA current produces a 10-V drop across a 500-ž resistor. Note that the sensor (i.e., current source) must be capable of developing a minimum output potential of around 10–11 V in order to achieve the full voltage drop across the resistor. This may not be a problem in the lab, but if a portable biosensor circuit is powered by only 6 V, it cannot drive more than 12 mA through the resistor (6 V/ 500 ž; see Figure 3). Furthermore, the resistor must have an adequate power rating (I2R) for the current and resistance values.

Issues such as these can be avoided by performing current measurements with benchtop instruments that use either a small dropping resistance or a feedback ammeter circuit. In the former case, highly sensitive voltage measurements allow a small dropping resistance value. A feedback ammeter avoids this problem because its input circuitry consists of an operational amplifier, which has very low input resistance and a voltage burden that typically ranges from about 10 µV to 1 mV. This type of current-measurement circuit is typically used in a complete biosensor instrument.

DMMs typically use dropping-resistor ammeter circuits, whereas picoammeters and electrometers use feedback ammeter circuits. In either case, making connections to the signal source is straightforward, and they are suitable for measuring currents up to several amperes. At the other end of the scale, DMMs can measure currents down to about 10 nA, and picoammeters can measure currents as low as 10 fA. Another advantage of these instruments is their built-in signal-conditioning circuitry. Some DMMs can provide voltage, current, resistance, and temperature (thermocouple or thermistor) measurements on up to 200 channels.

Source-Measure Instruments. In I-V characterization, the integration of a DC source and measuring instrument can be problematic because of intricate triggering issues. Such issues can often be avoided by using a tightly integrated source-measure unit (also referred to as an SMU or a SourceMeter instrument). These high-precision instruments can act as either a voltage or current source with sweep, pulse, and compliance-limit capabilities, and simultaneously measure I and V parameters (see Figure 4). Typical resolutions are in the range of microvolts and picoamperes.

The bipolar voltage and current sources of these instruments are controlled by a microprocessor, which makes I-V characterization much more efficient and simplifies instrumentation setup. When a SourceMeter instrument is used, many different test sequences can be stored in its program memory and executed with a simple trigger signal. Test data can be stored in a buffer memory until an I-V sweep is completed, and then downloaded to a PC for processing and analysis. 

Cabling, Conductors, and Capacitance

The connections between the instrumentation and the DUT are important parts of a measurement system. Understanding and managing the limitations of these connections is crucial for accurate measurements. Noise sources, cable length, and cable capacitance can affect the quality of any measurement, but the lower the signal level the more important these issues become. To minimize problems, the measurement circuit, its cables, and its connectors should be matched to the test signals. In addition, cables and test leads should be carefully routed and mounted. 

Cabling. When evaluating a cable for the measurement application, several issues should be considered. The electrical noise present in the test environment should be quantified. Noise can be defined as any undesirable signal that is impressed upon a signal of interest. 

Sources of electromagnetic noise include ac power lines, motors and generators, transformers, fluorescent lights, cathode-ray tube displays, computers, radio transmitters, etc. Depending on the nature of the signal and the noise, it may not be possible to separate them once the signal has been acquired at the instrumentation input terminals.

To the extent possible, test leads and cables should be routed to minimize their exposure to noise. The leads should then be mounted rigidly in place so that they cannot move and cause the generation of spurious voltage in the presence of electromagnetic fields.

In addition, the distance between the signal source and measurement system terminals should be measured. Wire has an electrical resistance that is dependent on its composition, length, and diameter. Resistance increases with increasing length and decreasing wire diameter. This resistance is a component of the total cable effects that become part of the analog input of a measurement circuit (see Figure 3). High cable resistance in conjunction with low A/D input resistance can result in a significant voltage drop through the interconnect wiring, resulting in measurement errors.

Lastly, it should be determined whether the data-acquisition channel has a single-ended or differential input. Single-ended signals (i.e., those referenced to ground) can be transmitted with two wires or with a shielded cable where the shield is tied to ground. For differential signals, at least two wires are needed to transmit the signal, which consists of a signal high and a signal low, neither of which is referenced to ground. Two individual conductors are sufficient to transmit the signal, but a twisted pair or shielded twisted pair provides greater noise immunity.

Conductors. The conductors used in shielded or unshielded cable can be made of solid or stranded wire. Solid wire results in minimum signal attenuation, but stranded conductors provide more flexibility and may be easier to route and mount. Conductors may consist of bare copper that is either plated with silver or tinned with solder. Connector and conductor materials should match to minimize resistance and thermally generated EMFs.

For the highest signal integrity, cables with shielded conductors should be used. Shielding reduces electromagnetic noise picked up by signal leads. It is also helpful in reducing electromagnetic radiation from conductors carrying high-frequency signals. Shielding is constructed with different types of wire braid or a combination of wire braid and foil. Multilayer or multibraid shields are more effective than a single layer in attenuating signal pickup and radiation. However, this tends to make cables more stiff and difficult to route and mount.

Several points should be considered when selecting shielded cable. For instance, higher-frequency noise is difficult to attenuate and requires more-elaborate shielding. 

Simple spiral wire-wrap foil is the least effective type of shielding. Tight braiding, double braiding, or braiding plus foil offer more-effective shielding. Lastly, caustic atmospheres, moisture, etc., can reduce the effectiveness of shielding. In some cases, these contaminants can leach into a cable and degrade the shielding far beneath the outer insulating jacket. Testing in such environments should be avoided, if possible.

Capacitance. For many DUTs, the output signal can be modeled as a voltage source in series with a resistance. Similarly, an analog instrument input can be modeled as a meter in parallel with an input resistance (see Figure 3). During a measurement, the instrument input absorbs a small bias current that the source must be able to supply. The interconnecting cable is an essential part of this circuit and can introduce resistance, capacitance, and inductive effects that depend on length, gauge, composition, routing, and the physical environment.

For high-speed DUTs, rapidly changing signals, circuit inductance, and capacitance can be serious obstacles to measurement speed, even if signal source and instrument impedances are properly matched. Often, spurious capacitance is more of a problem than inductance. Signals originating from a high-impedance source take longer to stabilize at the instrument input because the limited current level of the signal requires more time to charge the cable capacitance. In that case, taking a measurement before the signal has settled leads to erroneous readings.

Conclusion

Jonathan Tucker is a lead industry consultant at Keithley Instruments Inc. (Cleveland). He can be contacted at jtucker@keithley.com.

Qualifying sensors and components for biodetection systems and analytical instruments can be simplified in the early stages of development by using I-V characterization techniques. In many cases, these same techniques can be carried over to production testing. Instrument manufacturers may provide valuable information for use when applying these techniques to a wide range of sensor and component types and in the selection of the best measuring instruments for research and development as well as for production testing.

References

1. R Park, “Developing Diagnostics for the War on Bioterrorism,” IVD Technology 9, no. 3, (2003): 29–34.

2. AP Kretz and D Styblo, “Toward Continuous Blood Glucose Monitoring,” Medical Device & Diagnostic Industry 25, no. 6 (2003): 78–83.

3. T Hianik et al., “Amperometric Detection of DNA Hybridization on a Gold Surface Depends on the Orientation of Oligonucleotide Chains,” Bioelectrochemistry 53, no. 2 (2001): 199–204.

4. “A Twist of Fate—Electrochemistry on Display,” Kent State Magazine (Kent, OH: Kent State University, 2003).

5. ND Popovich, “Mediated Electrochemical Detection of Nucleic Acids for Drug Discovery and Clinical Diagnostics,” IVD Technology 7, no. 3 (2001): 36–42.

6. A Armutat et al., Data Acquisition and Control Handbook, 1st edition (Cleveland: Keithley Instruments Inc., 2001): sect. 5, p 4.

7. J Tucker et al., Low Level Measurements, 5th edition (Cleveland: Keithley Instruments Inc., 1998): sect. 1, pp 9–15. 

Copyright ©2004 IVD Technology