Hysteresis spoils the show

The apparent carbon nanotube sensor refreshing obtained with a gate voltage pulse (GRP) was determined to be a measurement artifact attributable to device hysteresis. Pulsed measurements, too fast to induce hysteresis, were used to determine that the gate bias pulses induced a hysteresis that compensated for the ammonia’s effect on the sensor.

In the figure, the red trace shows the normal sense and recovery from a short exposure to ammonia. The blue trace results when the gate bias is pulsed after the ammonia is removed. This appears to show that the sensor has recovered to (nearly) its original pre-ammonia state. When a very brief gate pulse of opposite polarity (the hysteresis reset pulse, HRP) is applied after the GRP, the sensor is then restored to its normal post ammonia recovery trajectory. Subsequent fast pulsed measurements were able to determine that hysteresis effects were making the sensor appear to refresh with a GRP.

Unfortunately, the highly desirable ability to electronically reset these chemical sensors (using a low power, non-thermal approach) appears to be a measurement artifact at least in this particular type of sensor. Perhaps other sensor devices or geometries will allow the realization of electronically refreshable chemical sensors.

The researchers presented their results in the journal Nanotechnology.

About the Author

Matthew Ervin is a Research Chemist in the US Army Research Laboratory’s Nanoelectronics Team. He is currently interested in CNT and graphene chemical sensors and supercapacitors. Andrew Dorsey and Natalie Salaets were summer students (from the University of Maryland) working at the US Army Research Laboratory.