Over recent decades, carbon-based chemical sensor technologies have advanced significantly. Nevertheless, significant opportunities persist for enhancing analyte recognition capabilities, particularly in complex environments. Conventional monovariable sensors exhibit inherent limitations, such as susceptibility to interference from coexisting analytes, which results in response overlap. Although sensor arrays, through modification of multiple sensing materials, offer a potential solution for analyte recognition, their practical applications are constrained by intricate material modification processes. In this context, multivariable chemical sensors have emerged as a promising alternative, enabling the generation of multiple outputs to construct a comprehensive sensing space for analyte recognition, while utilizing a single sensing material. Among various carbon-based materials, carbon nanotubes (CNTs) and graphene have emerged as ideal candidates for constructing high-performance chemical sensors, owing to their well-established batch fabrication processes, superior electrical properties, and outstanding sensing capabilities. This review examines the progress of carbon-based multivariable chemical sensors, focusing on CNTs/graphene as sensing materials and field-effect transistors as transducers for analyte recognition. The discussion encompasses fundamental aspects of these sensors, including sensing materials, sensor architectures, performance metrics, pattern recognition algorithms, and multivariable sensing mechanism. Furthermore, the review highlights innovative multivariable extraction schemes and their practical applications when integrated with advanced pattern recognition algorithms.
