Sensors have a transducer element that converts one stimulus or form of energy (input) into another form of energy (output signal).  Inputs may be physical (e.g., temperature, pressure) or chemical (e.g., vapors, liquids).  A sensor behaves reversibly, i.e., it should return to its original state after the sensing event (stimulus) has occurred.  Sensors should not be confused with assays, which are irreversible, nor with actuators, which convert energy into motion (e.g., muscles).

One of our areas of interest is chemical vapor sensing, where the sensing event is the interaction of a vapor with a transducer element.  The ideal chemical vapor sensor is small, lightweight, inexpensive, sensitive (able to detect vapors down to extremely low concentrations, e.g., parts per trillion, ppt), selective (able to distinguish vapors of interest from chemically similar background ‘interferent’ vapors), and should respond to analytes quickly and reversibly with no long-term deterioration in response after multiple sensing events.

Sensors may function according to a lock-and-key model, or according to a cross-reactive array model.  Array systems are ‘cross-reactive’ when one component in the array responds to multiple analytes, and a particular analyte causes a number of components in the array to respond (see above). This is in contrast to lock-and-key sensor systems where each component is specific to a given analyte or class of analytes.  Cross-reactive array-based vapor sensor systems include surface acoustic wave (SAW) sensors (giving changes in acoustic frequency), carbon-polymer composite sensors (giving changes in volume and in resistivity), metal oxide sensors (giving changes in resistivity), conducting organic polymers such as polypyrroles (giving changes in resistivity), and chemically responsive dyes (giving on-off intensity changes or wavelength changes in absorption or fluorescence emission).

In any array-based sensor system, it is necessary to determine the optimum number and composition of components in the array. Too few components may result in a lack of diversity and an inability to distinguish one vapor from another, while too many components may result in redundancy, i.e., two array components may give the same response to a given vapor.  Hence one of these components is not contributing new information about the vapor, and is redundant in the array. It is also necessary to consider the target application of the sensor system; an array for detecting a wide range of chemically diverse vapors (e.g., toxic industrial chemicals) will have markedly different components to an array used for distinguishing a set of chemically similar vapors (e.g., phosphonate pesticides and nerve agents). An array must also have the ability to distinguish vapors of interest from interferent vapors present in the background environment.