Our group’s interest in this area began about 30 years ago by originating the basic concept of the use of molecular or nanowires to create superior sensitivity in sensors. Tim wrote about the early phases of the molecular wire concept in his review in 1998 (“The Molecular Wire Approach to Sensory Signal Amplification” Acc. Chem. Res. 1998, 31, 201-207). This concept can be applied to both the transport of charges (electrical conduction) or energy (exciton migration). Chemiresistivity refers to the development of materials that change their electrical resistance in response to a chemical or biochemical stimulus. Our early work focused on conducting polymers and we pioneered integrated designs of molecular recognition elements that can profoundly affect the electrical transport through a conducting polymer. Similarly, we created a number of sensory systems based on excitonic transport and in the process of doing these studies we conducted detailed analyses about the physics of transport such as dipolar vs. exchange (FRET vs. orbital mediated) mechanisms. Our deep understanding of these processes is still relevant to our use of semiconducting polymers in optical sensors.
Carbon nanotubes (CNTs) are also very useful for the creation of chemical sensors and we have recently published a comprehensive review of this area (“Carbon Nanotube Chemoresistive Sensors” Chem. Rev. 2019, 119, 599-663). Similar to our approaches with conducting polymers, we have strived to create systems that make use of clear chemical principles. In a CNT sensor there are three different types of mechanisms for creating signals.
Chemiresistors based on single walled carbon nanotubes (SWCNTs) can transduce chemicals (analytes) by three general mechanisms.
Examples of systems we have designed include:
DNA hybridization for the creation of simple high sensitivity genomic sensors by localized enzyme triggered metal deposition. In this work we made “giga-polymers” that were alternating macromonomers of CNTs and DNA (i.e. (-SWCNT-DNA-)n) and deposited these materials between electrodes. The trick here was to ensure that the DNA would only react with the ends of the SWCNTs and we used surfactants, in effect, as protecting groups to conduct this polymerization in water (Weizmann, Y.; Chenoweth, D. M.; Swager, T. M. “Addressable Terminally-Linked DNA-CNT Nanowires” J. Am. Chem. Soc. 2010, 132, 14009–14011). The DNA represented a high resistivity linkage in the polymer and with exposure to a target DNA we can pull down and localize an HRP enzyme that is modified with complementary DNA to the target DNA. The HRP can be used to deposit Ag metal (a biological version of the Tollens reaction) at the junctions between the SWCNTs and connect them into a circuit. This scheme produced very sensitive resistivity tests for DNA. (Weizmann, Y.; Chenoweth, D. M.; Swager, T. M. “DNA-CNT Nanowire Networks for DNA Detection” J. Am. Chem. Soc. 2011, 133, 3238–3241.)
Illustration of DNA binding and sequestering of HRP can be used to deposit silver metal to make an electrical circuit for resistive ultra-trace detection of DNA.
We appreciate the recent innovations in catalysis and synthesis that occupy many of our brilliant colleagues in chemistry. In the area of organometallic transition metal chemistry many catalytic processes involve oxidative addition and reductive elimination. These changes in the nature of the metals can be used to create sensors wherein the metals can be used to interact directly with carbon nanotubes. In this case, we simply run reactions in ionic liquids over the top of SWCNT networks. As an example, we made use of the oxidative Heck reaction developed by the Stahl group at the University of Wisconsin to detect vapors of toxic monomers like ethyl acrylate. (Schröder, V.; Swager, T. M. “Translating Catalysis to Chemiresistive Sensing” J. Am. Chem. Soc. 2018, 140, 10721-10725.). Note that the response is highly selective.
The oxidative Heck reaction is run in an ionic liquid solvent over the top a network of SWCNTs to create selective sensors for reactive toxic alkenes like ethylacrylate.
Other dynamic processes can also be used to make sensors. In particular, chemically biasing equilibria has been used to create sensors for CO (Liu, S. F.; Lin, S.; Swager, T. M. “An Organocobalt-Carbon Nanotube Chemiresistive Carbon Monoxide Detector” ACS Sensors 2016, 1, 354-357). As shown, the CO takes the binding site where the amine was attached to the cobalt in this compound and the free amine donates electrons to the SWCNT, quenching the holes (carriers).
CO liberation of amines from binding cobalt canter interact with SWCNTs to cause a selective sensor reponse.
Chemiresitive sensors are powerful in that they are the simplest type of element to integrate into electrical devices. Many of our methods make use of complex arrays with many sensor elements and we can use these devices with proper orthogonality to create robust electronic nose devices. Chemiresistors also have exceedingly low power requirements and can be put into devices that can be powered and read by digital radiofrequency stimulation with a smart phone. We have developed these passive RFID methods for the detection of oxygen in food packaging (see below, Zhu, R.; Desroches, M.; Yoon, B.; Swager, T. M. “Wireless Oxygen Sensors Enabled by Fe(II)- Polymer Wrapped Carbon Nanotubes” ACS Sensors, 2017, 2, 1044-1050) and for wearable sensors that can detect chemical weapons.
RFID dosimeters for O2 are formed using Fe+2 ions that initially quench the SWCNT conductivity and with oxidation by air create Fe+3 centers that restore the carrier density.
Concept of Smartphone readable RFID chemiresistors for wearable badges for the detection of toxic chemicals.
The spatial and temporal information from a passive RFID tag read by a smart phone is a powerful concept in sensing. For a real-time video showing this method click below.