▎ 摘　　要

Over the past decade, advantages of graphene in high-performance gas sensing have been demonstrated, especially for single- or few-layered graphene wherein the theoretical and technical advances are mature. Owing to the complexity of multi-layered graphene (MLG) sensors and the increasing demand for practical applications, there is an urgent need to comprehensively understand the correlation between MLG and its derivatives for developing next-generation gas sensors. Herein, theoretical and empirical strategies for obtaining better gas sensors are developed. These approaches can be divided into three categories: 1) building devices with Fermi level near the Dirac point (E-F,E-Dirac), 2) enhancing the adsorption probability f(x) and driving force (gap between as-prepared and saturated Fermi levels), and 3) accelerating mobility. A device employing p-type reduced graphene oxide (rGO) decorated with n-type indium oxide and phenylenediamine (GIP) was designed and fabricated by adopting approaches 1 and 2 (E-F,E-Dirac) and f(x) enhancement). The resulting hole-compensated GIP displayed a remarkable response to formaldehyde (HCHO), which was 66.3 times higher than rGO, with faster response/recovery. GIP also exhibited higher selectivity for HCHO than for ammonia and trimethylamine. We believe that the classification will untangle the complex role of graphene in sensing, helping to design next-generation advanced gas sensors.