Increasingly, an electronic tongue, or E-tongue (a sensor device for recognition and quantitative component analysis of taste and flavor) is used to measure, identify, classify, and discriminate between liquid or solid samples in the food and beverage industry. E-tongues provide automatic analysis and recognition, and after identification is complete, the sensor array is able to estimate the concentration of the samples and the sample’s inherent properties.
Employed for quality control during processing and storage of such products as water, wine, coffee, milk, and juice, these taste sensors can be used to analyze flavors in beverages and to quantify the “spiciness” or “bitterness” of teas, other drinks, and dissolved compounds.
The use of taste sensors has expanded to the control of aging processes for such products as cheese and whiskey, and into the medical industry where e-tongue and e-nose technologies are used in the diagnosis of breath, sweat, urine, and skin odor.
Sensing methods involved in e-tongue applications include:
- Potentiometer sensors
- Conductivity sensing
- Voltamperometry (information about a substance is obtained by measuring the current as the potential is varied)
- Optical sensors
The heart of the e-tongue’s success is the quality of its sensors, whether individual or multiple devices, to recognize patterns. Artificial taste sensors have polyvinyl chloride (PVC)/lipid membranes that interact with a target solution such as beverages, caffeine, and more. The membrane potential of the lipid membrane changes, which is the sensor output, or measurement. Investigating potential change results in measuring the “taste” provided by the output of the chemical substances. With a sensor array, multiple sensors provide output and can form a unique substance fingerprint. (For more details see the Technology Zone article “The Five Senses of Sensors—Taste”)
A recently published paper by Universitat Autonoma de Barcelona, Spain describes work performed with electronic tongue systems using electrochemical sensors. The paper discusses specific cases from the Sensors and Biosensors Group involving work with chemical and biosensors, and the use of sensors with reduced selectivity, grouped in arrays with cross-response characteristics. Tools were used in the processing of data for complete analysis, and a number of substances were determined in a single, direct operation. This work was performed on the basis of automated detection principles in such fields as food, beverages, and pharmaceuticals, where alternatives to the human expert are in great demand based on the need for increased accuracy.
In their study, the preferred chemometric tool for the processing of this particular data is ANN, artificial neural networks, recognized as powerful nonlinear modelers for quantitative and qualitative applications. The study uses a group of sensors with cross-response as the sensing method in the taste buds of animals, and uses ANNs as parallel-information processing tools with roots in the animal nervous system.
In their work, the researchers defined the electronic tongue as an analytical instrument comprising an array of non-specific, poorly-selective chemical sensors with cross-sensitivity to different compounds in a solution, coupled with the tool for data processing. They also offered a definition from a recent IUPAC report, “a multisensory system, which consists of a number of low-selective sensors and uses advanced mathematical procedures for signal processing based on pattern recognition and/or multivariate data analysis, Artificial Neural Networks (ANNs).”
Within these applications, sensors are used not only to find the right mix, but also for consistency within batches and to identify any contaminants that find their way into the process. The applications typically involve:
- Process monitoring
- Shelf-life analysis
- Freshness evaluation
- Authenticity verification
- Quality control
Not only are the sensors being used more efficiently in industrial settings, they are increasingly being used to replace or support food expert panels in taste-test settings. The highly sensitive sensors are able to correlate with data provided by human-based panels.
Accuracy is key
Improvements in sensor accuracy and their ability to measure samples and taste, and provide analysis and assessment, are making rapid gains. Sensor accuracy is critical in food processing equipment, as is the ability to operate in rugged and corrosive environments. In food and beverage manufacturing settings, frequent cleaning using high-pressure, high-temperature water and strong detergents is potentially damaging to sensors and account for early sensor failure. Replacing sensors is costly in terms of both dollars and time lost due to line shutdowns that affect the bottom line.
Sensors themselves are not enough to address the challenges of food and beverage environments. In this application group, it is also packaging that is extremely important. For example, the Omron E3ZM Compact Photoelectric Sensor with Stainless Steel Housing (Fig. 1), designed for greater hygiene, is used based on the ability of its materials to work within a harsh and demanding environment. The housing makes the E3ZM suitable for the cleaning conditions that exist in food-processing machinery. It is suitable for high temperatures, high pressure, and jet water spray-cleaning applications.
Designed with few indentations in the shape, less dust and water can be collected. It has model and lot numbers imprinted with a laser marker instead of labels to prevent foreign matter from contaminating the food products.
Figure 1: The Omron E3ZM’s strong housing provides resistance against detergents, disinfectants and jet liquid flow. (Courtesy of Omron)
Several types of sensors are involved within food and beverage manufacturing. Texas Instruments’ pressure, load, and force sensors typically combine a Wheatstone bridge-resistor configuration with a mechanical structure. As the mechanical structure is stressed, the bridge output changes in proportion to the pressure, load, or force applied to the sensor.
Using its Webench online tool and configurable sensor AFE (analog front-end) ICs, Texas Instruments allows an engineer to select a sensor, design a solution, and download the result. Suitable to interface with pH sensors, the Texas Instruments’ LMP7312 (Fig. 2) AFE provides a high rate of accuracy, sensitivity in the 5 mV/V to 15 mV/V range with a bridge impedance between 3 KΩ and 10 KΩ, and typical overall signal accuracy of 0.5% to 5%.
Figure 2: The TI LMP7312 is used for food, beverage, and industrial applications. Hardware/system monitors measure temperature, voltage, and currents. Some products provide outputs for processor clocking or fan control to regulate temperature.
Advances in E-tongue taste sensors and in pressure, position, and temperature sensing continue to take place throughout the food and beverage industry. New packaging materials designed to withstand the extremely harsh environment encountered by food and beverage sensors also continue to advance. Ultimately, the higher level of accuracy, reduction in contaminants, and positive impact on the bottom line will ensure their rapid adoption.